Electricity Market Reform An International Perspective

Electricity Market Reform: An International Perspective Elsevier Internet Homepage-http://www.elsevier.com Consult th...

Author: Woodrow Clark | Ted Bradshaw

118 downloads 1713 Views 7MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Electricity Market Reform: An International Perspective

Elsevier Internet Homepage-http://www.elsevier.com Consult the Elsevier homepage for full catalogue information on all books, journals and electronic products and services Elsevier Titles of Related Interest DICTIONARY OF ENERGY Cleveland & Morris 2005; ISBN: 0-08-044578-0 www.elsevier.com/locate/doe AGILE ENERGY SYSTEMS Clark & Bradshaw 2004; ISBN: 0-08-044448-2 www.elsevier.com/locate/isbn/0080444482 RESHAPING EUROPEAN GAS AND ELECTRICITY INDUSTRIES Finon & Midttun 2004; ISBN: 0-08-044550-0 www.elsevier.com/locate/isbn/0080445500 Related Journals The following titles can all be found at: http://www.sciencedirect.com The Electricity Journal Applied Energy Climate Policy Energy Energy Conversion & Management Energy Policy Renewable Energy Renewable & Sustainable Energy Reviews Solar Energy Progress in Nuclear Energy Refocus To Contact the Publisher Elsevier welcomes enquiries concerning publishing proposals: books, journal special issues, conference proceedings, etc. All formats and media can be considered. Should you have a publishing proposal you wish to discuss, please contact, without obligation, the Publisher responsible for Elsevier’s energy publishing program: Henri van Dorssen Publisher, Energy Elsevier Limited The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK Tel.: ⫹ 44 1865 84 3931 Fax: ⫹ 44 1865 84 3920 E-mail: [email protected] General enquiries including placing orders, should be directed to Elsevier’s Regional Sales Offices – please access the Elsevier internet homepage for full contact details.

Electricity Market Reform: An International Perspective Edited by Fereidoon P. Sioshansi and Wolfgang Pfaffenberger

AMSTERDAM – BOSTON – HEIDELBERG – LONDON – NEW YORK – OXFORD PARIS – SAN DIEGO – SAN FRANCISCO – SINGAPORE – SYDNEY – TOKYO

ELSEVIER B.V. Radarweg 29 P.O. Box 211, 1000 AE Amsterdam The Netherlands

ELSEVIER Inc. 525 B Street, Suite 1900 San Diego, CA 92101-4495 USA

ELSEVIER Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB UK

ELSEVIER Ltd 84 Theobald’s Road London WC1X 8RR UK

© 2006 Elsevier Ltd. All rights reserved. This work is protected under copyright by Elsevier Ltd., and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier’s Rights Department in Oxford, UK: phone (⫹44) 1865 843830, fax (⫹44) 1865 853333, e-mail: [email protected]. Requests may also be completed on-line via the Elsevier homepage (http://www.elsevier.com/locate/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone (⫹1) (978) 7508400, fax: (⫹1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (⫹44) 20 7631 5555; fax: (⫹44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier’s Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. First edition 2006 ISBN-13: 978-0-08045-030-8 ISBN-10: 0-080-45030-X Printed in Great Britain. 06 07 08 09 10

10 9 8 7 6 5 4 3 2 1

Contents Contributors Foreword: The Market Versus Regulation Stephen Littlechild Introduction to Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies Paul L. Joskow PART I: What’s Wrong With the Status Quo? 1.

Why Restructure Electricity Markets? Fereidoon P. Sioshansi and Wolfgang Pfaffenberger

2.

Sector-Specific Market Power Regulation versus General Competition Law: Criteria for Judging Competitive versus Regulated Markets Günter Knieps

PART II: Trailblazers

vii xvii 1

33 35

49

75

3.

Chile: Where It All Started Ricardo Raineri

4.

Electricity Liberalization in Britain And The Evolution of Market Design David Newbery

109

5.

The Nordic Electricity Market: Robust By Design? Eirik S. Amundsen, Lars Bergman and Nils-Henrik M. von der Fehr

145

PART III: Evolving markets 6.

The Electricity Industry in Australia: Problems Along the Way to a National Electricity Market Alan Moran

77

171

173

7.

Restructuring the New Zealand Electricity Sector 1984–2005 Geoff Bertram

203

8.

Energy Policy and Investment in the German Power Market Gert Brunekreeft and Dierk Bauknecht

235

9.

Competition in the Continental European Electricity Market: Despair or Work in Progress? Reinhard Haas, Jean-Michel Glachant, Nenad Keseric and Yannick Perez

265

v

vi

Contents

PART IV: North America, New world, New Challenges

317

10.

California Electricity Restructuring, The Crisis, and Its Aftermath James L. Sweeney

319

11.

Texas: The Most Robust Competitive Market in North America Parviz Adib and Jay Zarnikau

383

12.

Electricity Restructuring in Canada Michael J. Trebilcock and Roy Hrab

419

13.

The PJM Market Joseph Bowring

451

14.

Independent System Operators in The USA: History, Lessons Learned, and Prospects Richard O’Neill, Udi Helman, Benjamin F. Hobbs and Ross Baldick

15.

Competitive Retail Power Markets and Default Service: The US Experience Taff Tschamler

479 529

PART V: Other Markets

563

16.

The Case of Brazil: Reform By Trial And Error? João Lizardo R. Hermes de Araújo

565

17.

Understanding The Argentinean and Colombian Electricity Markets Isaac Dyner, Santiago Arango and Erik R. Larsen

595

18.

A New Stage of Electricity Liberalization in Japan: Issues and Expectations Mika Goto and Masayuki Yajima

617

Index

645

Contributors

Parviz ADIB is Director of Wholesale Market Oversight at the Public Utility Commission of Texas where he is engaged in designing competitive electricity market rules, providing policy recommendations to the Commissioners, and supervising wholesale market monitoring functions. Prior to joining the Commission, he taught graduate and undergraduate courses in The University of Texas at Austin. He, along with a couple of other Commission Staff, was the recipient of Center for the Advancement of Energy Markets’ Unsung Heroes Award for leadership by a civil servant who has made a significant difference in gas and electric competition in 2005. Dr. Adib’s main research interests include energy and public policy, efficient operation of restructured electricity markets, and effective market monitoring mechanism. He has published papers and prepared numerous presentations and policy recommendations on these topics, notably covering Restructured Electricity Markets. Dr. Adib completed his BA and MS studies in Economics from Tehran University and holds a PhD from The University of Texas at Austin. Eirik S. AMUNDSEN is Professor of Economics at University of Bergen, Norway and at Institute of Food and Resource Economics, The Royal Veterinary and Agricultural University, Copen hagen, Denmark. He has served as Nordic Research Professor affiliated with The Nordic Energy Research Programme, and as a Scientific Adviser to SNF, Norway and SNS, Sweden. He has contributed a variety of articles in theoretical and applied economic journals related to energy economics, environmental economics, and resource economics. In the last few years his research has focused on electricity markets. He has his university degrees from University of Bergen, University of Copenhagen, Institut Français du Pétrole (ENSPM) and Stanford University. He is Dr.ès.sc.écon. (dr. d’Etat) with a Prize-awarded thesis from University of Paris, Panthéon-Assas. Santiago ARANGO is at the Universidad Nacional de Colombia and is about to finish the PhD in the System Dynamics Group, University of Bergen, Norway. He is exploring the dynamics of deregulated electricity markets with experimental economics and system dynamics. From 1998 until 2002 he has done consulting work for governmental and private organization in Colombia, attached to the Energy Institute, Universidad Nacional de Colombia. João Lizardo R.H. de ARAÚJO is Professor of Energy Economics at the Institute of Economics, Federal University of Rio de Janeiro, presently loaned to Eletrobrás to head the Centre for Electric Power Research (CEPEL). Most of his 40-year academic career has been at UFRJ, in the Coordination for Graduate Programmes in Engineering – COPPE (1970–1994) and in the Institute of Economics (1994-present). He has taken a number of positions, including Head of Computing Department at the Institute of Mathematics, Head of Systems and Optimization Programme and of the Interdisciplinary Energy Programme, both at COPPE, and Research Director at the Institute of Economics. Prof. Araújo has also been a Visiting Scholar at the Imperial College, at Lawrence Berkeley Laboratory, and at SPRU. vii

viii

Contributors

Prof. Araújo’s main fields of interest include system optimization, modeling and regulatory issues. He has published numerous papers on energy regulation and energy modeling, as well as on technological innovation. His latest book on energy regulation issues is “Diálogos da Energia: Reflexões sobre a última década 1994–2004” (7 Letras, Rio de Janeiro 2005), with Prof. Oliveira, also of IE/UFRJ. Other relevant publications include “Brazilian Energy Policy: Changing Course?” (with A. de Oliveira). Prof. Araújo has a Degree in Electronic Engineering from Instituto Tecnológico de Aeronáutica, a DEA and a Dr. Sp. from Université de Toulouse. Ross BALDICK is Professor of Electrical and Computer Engineering at The University of Texas at Austin. From 1991 to 1992 he was a Post-Doctoral Fellow at the Lawrence Berkeley Laboratory. In 1992 and 1993 he was an Assistant Professor at Worcester Polytechnic Institute. Dr. Baldick has published over 40 refereed journal articles and has research interests in a number of areas in electric power. His current research involves optimization and economic theory applied to electric power system operations, the public policy and technical issues associated with electric transmission under deregulation, and the robustness of the electricity system to terrorist interdiction. In 1994, Dr. Baldick received a National Science Foundation Young Investigator Award. He is Editor of IEEE Transactions on Power Systems and the Chairman of the System Economics Sub-committee of the Power Systems Analysis, Computation, and Economics Committee of the IEEE Power Engineering Society. He received his BSc and BE Degrees from the University of Sydney, Australia and his MS and PhD from the University of California, Berkeley. Lars BERGMAN is Professor of Economics and President of the Stockholm School of Economics. He is also a Member of the Royal Swedish Academy of Engineering Sciences and Chairman of the Swedish Association for Energy Economics. His research has been focused on general equilibrium modeling, environmental economics and policy, and the economics of electricity markets. During more than a decade he has had an active role in the management of both the Nordic Energy Research Program and an Energy Market Research Program at the Stockholm School of Economics. He is one of the authors of A European Market for Electricity (1999). He has a PhD in Economics and an MSc from the Stockholm School of Economics. Geoff BERTRAM is Senior Lecturer in Economics at Victoria University of Wellington, New Zealand. His primary research interests are in the economics of regulation and industry restructuring, with particular reference to energy and infrastructure sectors; and the economics of development in small island states. Prior to joining the faculty of Victoria University he was employed as a Research Officer at the Institute of Economics and Statistics at Oxford University, writing an economic history of Peru. Dr. Bertram has published numerous papers on energy sector restructuring, including most recently “Price–Cost Margins and Profit Rates in New Zealand Electricity Distribution Networks Since 1994: The Cost of Light-Handed Regulation”, Journal of Regulatory Economics 27(3): 281–307, May 2005; and “Deregulation and Monopoly Profits in New Zealand’s Gas and Electricity Sectors”, Energy Studies Review 12(3): 208–227, Spring 2004. Dr Bertram holds a BA Honors Degree from Victoria University of Wellington, and an MPhil and DPhil from the University of Oxford, UK. He is on the editorial advisory boards of a number of journals, including Environment and Development Economics and Asia Pacific Viewpoint, and was formerly on the editorial board of World Development.

Contributors

ix

Gert BRUNEKREEFT is a Researcher at the Tilburg Law and Economics Center (TILEC) at Tilburg University in the Netherlands. Before that he held research positions in Applied Economics at the University of Cambridge, where he was part of the Cambridge-MIT-Institute (CMI) Electricity Project and before that at the Institute of Transport Economics at Freiburg University in Germany. Dr. Brunekreeft’s main research interests are in industrial economics, regulation, and competition policy of network industries, especially electricity markets. His book Regulation and Competition Policy in the Electricity Market published in 2003 analyses developments on the German electricity market. He has published widely in a variety of energy-related academic journals, including Utilities Policy, Oxford Review of Economic Policy, and Energy Journal. Dr. Brunekreeft holds a Degree in Economics from the University of Groningen in the Netherlands and a PhD and Habilitation from Freiburg University in Germany, both in economics. At the moment he is Associate Editor for the Journal of Network Industries. Dierk BAUKNECHT is Research Fellow with the Oeko-Institut, a German research institute in the area of applied environmental and energy research. He currently works on power plant investment models, network regulation, decentralized power generation, innovation research and transformation of energy systems. Previously, he was market analyst for Germany and modeling manager with a UK-based power market consultancy. His main research interests include the integration of distributed generation into power networks and markets. He has published a number of papers both on this topic and on the development of competition in the German electricity market, mainly in industry journals and the Financial Times. Bauknecht graduated in Political Science at the Free University of Berlin and holds an MSc in Science and Technology Policy from SPRU at the University of Sussex, UK. He is currently completing his PhD on Network Regulation and Distributed Generation. Isaac DYNER is Professor of Operational Research and Energy at the Energy Institute, Universidad Nacional de Colombia. He has been a Visiting Professor of the British Academy and Academic Visitor at Imperial College London, City University, London Business School, Warwick University and several Latin American universities. He has been Editor of the Energy Journal Energetica and Member of the Colombian Council for Energy Research. He has consulted with companies and governments in the areas of strategy, deregulation, energy, and modeling. He has published over 150 research papers in the areas of energy, operational research, and system dynamics. He has published in scientific journals including, among others, Journal of Operational Research, Energy Policy, Utilities Policy, International Journal of Global Energy Issues, Futures, International Journal of Operational Research and System Dynamics Review as well as three books and two chapter in books published by Wiley and Risk Books. He obtained his PhD from London Business School and Master of Sciences from both University of Warwick and Southampton University, UK. Nils-Henrik VON DER FEHR is Professor of Economics at the University of Oslo, where is engaged in research in the fields of competition policy, regulation and energy and environmental economics. Prof. von der Fehr’s main research interests include the design of electricity markets, competitive wholesale auctions, and investment incentives. He has published numerous papers on such topics. He has acted as advisor on regulatory matters to governments around the world, including Australia, Brazil, the Netherlands, Norway, and Sweden. He

x

Contributors

was a Member of the Dutch Electricity Market Surveillance Committee and recently chaired a commission on the developments of the Nordic electricity market. He has a Degree in Economics and Mathematics from the University of Oslo and holds a PhD in Economics from the same university. Jean-Michel GLACHANT is Professor of Economics at University Paris XI, Head of its Economics Department, Member of the Board of the International Society for New Institutional Economics (ISNIE), and Director of the New Institutional Economics Research Center at La Sorbonne University 1991–2000. Since 1995 he has managed a series of research projects on utility privatization and regulation, electricity markets competition, and restructuring. He his or has been Advisor for the French energy regulator, the French government, the European Commission (Energy or Competition), and coordinator of the EU-funded project SESSA on European electricity reforms. He is Founding Member of the new “European Energy Institute”. The two most recent books he edited are Competition in European Electricity Markets: A Cross-Country Comparison, Edward Elgar, 2003 and The Economics of Contracts: Theories and Applications, Cambridge University Press, 2002. He has a Degree in Economics from La Sorbonne University in Paris and holds a PhD in Economics from the same university. Mika GOTO is Research Economist at Socio-Economic Research Center, Central Research Institute of Electric Power Industry (CRIEPI). She was a Visiting Researcher at the Institute of Energy Economics at the University of Cologne and the National Regulatory Research Institute at the Ohio State University. Her main areas of interest include structural reforms, productivity and efficiency analysis as well as cost structure of electric power utilities. She is engaged in research of the economies of scale and productivity growth of electric power industry. Dr. Goto obtained her Doctoral Degree in Economics from Nagoya University, Japan. Reinhard HAAS is Associate Professor of Energy Economics at Vienna University of Technology in Austria. Since 2001 he has served as the Vice-Head of the Institute of Power Systems and Energy Economics. He has been engaged in research projects for the European Commission and others. His current research includes dissemination strategies for renewables; sustainable energy systems, and liberalization versus regulation of energy markets. He has published numerous articles in peer-reviewed international journals. He has studied industrial economics in mechanical engineering. He holds a PhD in Energy Economics from Vienna University of Technology. Udi HELMAN is an Economist at the Federal Energy Regulatory Commission (FERC). He has worked extensively on US electricity market design issues, including auction markets for wholesale energy, ancillary services and installed capacity, transmission usage pricing and market rules for transmission property rights. His major projects have included the initial design and redesign of the Independent System Operator (ISO) New England markets, the Standard Market Design initiative and on the development of the Midwest ISO markets. Dr. Helman has authored or co-authored a number of articles and book chapters on aspects of electricity regulatory reform. These include “Market power monitoring and mitigation in the US wholesale power markets”, Energy (May–June 2006); “A Primer on ComplementarityBased Equilibrium Modeling for Electric Power Markets”, in D.W. Bunn (ed.), Modeling Prices in Competitive Electricity Markets (Wiley, 2004); and “A Joint Energy and Transmission Rights Auction: Proposal and Properties”, IEEE Transactions on Power Systems (November 2002).

Contributors

xi

Dr. Helman holds BA and MA Degrees from the University of Toronto and a PhD from The Johns Hopkins University in Energy Economics and Systems Analysis. Benjamin F. HOBBS is Professor in the Department of Geography and Environmental Engineering of The Johns Hopkins University. He also holds a joint appointment with the Department of Applied Mathematics and Statistics. Previously, he was a Professor of Systems Engineering and Civil Engineering at Case Western Reserve University. He has also been a Member of the Research Staffs of Brookhaven and Oak Ridge National Laboratories. Dr. Hobbs’ research interests are in electric sector policy and planning, and ecosystem management. He is Member of the California ISO Market Surveillance Committee, Scientific Advisor to the ENC Policy Studies Unit, and Member of the Public Interest Advisory Committee of the Gas Technology Institute. His recent books are Energy Decisions and the Environment, A Guide to the Use of Multicriteria Methods (with P. Meier, Kluwer, 2000) and The Next Generation of Electric Power Unit Commitment Models (edited with M. Rothkopf, R. O’Neill, and H. Chao, Kluwer, 2001). His PhD is in Environmental System Engineering from Cornell University. Roy HRAB is Policy Advisor at the Ontario Energy Board where he is engaged in regulatory policy development. Prior to joining the Board he worked at the Institute for Competitiveness and Prosperity and for the Government of Ontario’s Panel on the Role of Government. Mr. Hrab’s current research interests include study of restructured electricity markets and transmission planning. He has published papers and articles (with Michael J. Trebilcock) on the electricity restructuring experience in Ontario, Canada, notably “Electricity Restructuring in Ontario”, in The Energy Journal, 2005. Roy has earned BA and MA Degrees in Economics from the University of Toronto. Paul L. JOSKOW is Elizabeth and James Killian Professor of Economics and Management at MIT and Director of the MIT Center for Energy and Environmental Policy Research. He received his PhD in Economics from Yale University in 1972 and has been on the MIT faculty since then. At MIT he is engaged in teaching and research in the areas of industrial organization, energy and environmental economics, and government regulation of industry. Prof. Joskow has published six books and over 100 papers on topics in these areas. He began doing research and writing on competitive electricity markets over 20 years ago and was co-author (with Richard Schmalensee) of Markets for Power published by MIT Press in 1983. He is Fellow of the Econometric Society and a Fellow of the American Academy of Arts and Sciences. Nenad KESERIC holds a Degree in Electrical Engineering (Power Engineering and Electrical Energetic Systems) at Cacak College of Engineering, University of Technology Kragujevac, Serbia and Montenegro. He is working on a PhD at the Energy Economics Group at Vienna University of Technology. His major focus of work is cross-border electricity trading, congestion management, energy modeling, and price forecasting. He has consulted for regulatory commission and private companies. Günter KNIEPS is Professor of Economics at the University of Freiburg, Germany, where he is engaged in research in the field of network economics. Prior to joining the Faculty of

xii

Contributors

Economics in Freiburg he held a Chair of Microeconomics at the University of Groningen, the Netherlands. He is Member of the Scientific Council of the Ministry of Economics and Labor as well as of the Ministry of Transportation and Housing. Dr. Knieps’ main research interests include study of network economics, de-regulation, competition policy, industrial economics, sector studies on telecommunications, transportation, and energy. He has published widely in academic and professional journals. Dr. Knieps has diplomas in Economics and Mathematics from the University of Bonn, Germany, and a PhD in Mathematical Economics from the University of Bonn. Erik LARSEN is Professor of Management, University of Italian Switzerland, Lugano, Switzerland. Erik previously held appointments at Cass Business School, University of Bologna, London Business School and Copenhagen Business School. During the period 1996–1998 he was an EU Marie Curie Fellow at the University of Bologna. His teaching and research is in the areas of strategy and deregulation of utility companies. His research has been published in journals such as American Metric Journal, European Journal of Operational Research, Energy Policy, Utility Policy, and Management Science. He has consulted for private companies, international organizations and governments in the area of deregulation. He obtained his PhD from the Institute of Economics, Copenhagen Business School and his MSc from the Technical University of Denmark. Stephen LITTLECHILD is Emeritus Professor at the University of Birmingham and Senior Research Associate at the Judge Business School, University of Cambridge. He has been a Member of Ofgem’s Panel of Economic Advisers since 1999. He was the UK’s Director-General of Electricity Supply and Head of the Office of Electricity Regulation (OFFER) from 1989 to 1998. Previously he was Professor of Commerce and Head of Department of Industrial Economics and Business Studies at the University of Birmingham from 1975 to 1989. He was a Member of the Monopolies and Mergers Commission from 1983 to 1989. Prof. Littlechild is an international consultant on privatization, regulation, and competition, especially in the electricity and telecommunications sectors. Since 1999, he has been a policy advisor to governments, regulators, and companies in Chile, Mexico, Philippines, India, Australia, New Zealand, Thailand, Brazil, Poland, Romania, Canada, Saudi Arabia, Russia, China, Argentina, and Colombia, as well as the UK. His recent research interests in electricity include competition in retail supply, transmission regulation, and negotiated settlements. He has a Bachelor of Commerce Degree from the University of Birmingham, a PhD from the University of Texas at Austin, and Honorary Doctorate Degrees from the Universities of Birmingham and East Anglia. Alan MORAN is the Director of the Deregulation Unit at the Institute of Public Affairs in Melbourne, Australia. He was previously a Senior Official in the Australian Federal and Victorian Governments responsible for regulatory review and energy policy matters. He is one of Australia’s best-known commentators on the energy industry and policy having published over 30 major papers covering specific aspects of the industry. He has authored four books, three on environmental economics, and published dozens articles and submissions on privatization, energy, and other economic policy matters. He has Degrees in Economics from the London School of Economics, Management from the University of Salford and a PhD from the University of Liverpool.

Contributors

xiii

David NEWBERY is Professor of Applied Economics at the University of Cambridge, where he is Research Director of the Electricity Policy Research Group. He has been an academic at the Faculty of Economics since 1966, with sabbaticals at Yale, Princeton, Stanford and Berkeley, as well as a spell as Division Chief at the World Bank. He has been an Economic Advisor to Ofgem, Ofwat, and a Former Member of the Competition Commission, and Chairman of the Dutch Electricity Market Surveillance Committee. He was President of the European Economic Association in 1996, was awarded the Frisch Medal of the Econometric Society, the Harry Johnson Prize of the Canadian Economic Association in 1993, and the recipient of the IAEE 2002 Outstanding Contributions to the Profession of Energy Economics Award. Dr. Newbery has managed a series of research projects on utility privatization and regulation, electricity restructuring, road congestion and road pricing, tax reform in Hungary and Greece. He has published over 100 academic journal articles. His two most recent books are A European Market for Electricity?, 1999, and Privatization, Restructuring and Regulation of Network Utilities, MIT Press, 2000. He was the Guest Editor of the special issue of The Energy Journal on European electricity liberalization in 2005. Dr. Newbery was educated at Trinity College, Cambridge with BA Degrees in Mathematics and Economics. He has a PhD in Economics and an ScD Degree from the University of Cambridge. He is Fellow of the Econometric Society and of the British Academy. Richard P. O’NEILL is the Chief Economic Advisor at the Federal Energy Regulatory Commission (FERC). Previously, he was the Chief Economist and Director of the Office of Economic Policy, the Director of the Commission’s Office of Pipeline and Producer Regulation at FERC, and prior to that, he worked at the Energy Information Administration. His work has focused on open access, restructuring, competition, performance-based incentive regulation, and market design. Prior to that he was on the Computer Science and Business Faculty of Louisiana State University and the University of Maryland. He has worked with several countries, states, the World Bank, energy companies and computer companies in the development of mathematical software, energy modeling, forecasting, regulation, privatization, restructuring and market design. His published work has appeared in academic and professional journals and books in the areas of applied mathematics, optimization, operations research, management science, computer science, energy, electrical engineering, economics, and law. He has a BS in Chemical Engineering, an MBA and a Doctorate in Operations Research from the University of Maryland. Yannick PEREZ is Associate Professor of Economics at the University of Paris-Sud 11 since 2003 where he is engaged in research in the field of Electricity Market Reforms, Industrial Organisation and Institutional Economics within the Groupe Réseaux Jean-Monnet at the ADIS laboratory. Prior to joining the faculty, he studied at University Paris Panthéon-Sorbonne. Dr. Perez’s main research interests include study of restructured electricity markets, benchmarking studies including study of company methodologies and strategies. He has published widely on these topics. Dr. Perez has Degrees in Economics from the University of Paris, Sorbonne and Ecole Normale Supérieure de Fontenay St-Cloud, and holds a PhD from the University Paris, Sorbonne. Wolfgang PFAFFENBERGER is Professor of Economics at the International University Bremen and Director of Bremer Energy Institute, an interdisciplinary research center focusing on energy system analysis and energy markets.

xiv

Contributors

He is especially interested in the economics of competition with reference to network industries and the electricity and natural gas industries in particular. He has published a textbook on electricity economics (in German) and co-authored a textbook on energy economics (in German) and written numerous studies and articles on macroeconomic implications of energy and environmental policy, on various aspects of market opening in the electricity supply industry (ESI) and on the special problems of the ESI in transformation countries. Dr. Pfaffenberger holds a Diploma in Economics from Freie Universität Berlin where he also got his PhD. He is Honorary Doctor of the Faculty of Economics of the State University of Novosibirsk in Russia. Ricardo RAINERI is Professor of Economics at the Engineering School of the Pontificia Universidad Católica de Chile, where he is engaged in research on industrial organization and regulation of the Energy Industry. Dr. Raineri’s main areas of interest include economics of competition, regulation, markets’ restructuring, and pricing. He has widely published in a variety of energy and related academic journals and authored two books. Also, he has been consultant for private companies and for the government, member of arbitrage commissions, and has been expert witness at the Chilean Congress. Dr. Raineri has Degrees in Business Engineering and Master in Economics from Pontificia Universidad Católica de Chile. He obtained a MA and a PhD in Economics from the University of Minnesota, under the supervision of Prof. Edward C. Prescott, 2004 Nobel Prize Laureate. Fereidoon SIOSHANSI is President of Menlo Energy Economics, a consulting firm specializing on the economics of the electric power sector. His professional experience includes working at Southern California Edison Company, the Electric Power Research Institute, National Economic Research Associates and Henwood Energy Services, Inc., now Global Energy Decisions. Dr. Sioshansi’s main research and consulting interests include electric power restructuring and market liberalization, global climate change, environmental and energy policy, sustainability, demand and price forecasting, integrated resource planning, energy efficiency, and renewable energy technology. He has authored numerous reports, books, book chapters and articles. He is the Editor and Publisher of EEnergy Informer, a monthly newsletter on the electric power industry. He is a frequent contributor to The Electricity Journal and Energy Policy and is on the editorial board of Utilities Policy. He has Degrees in Engineering and Economics, including an MS and PhD in Economics from Purdue University. James L. SWEENEY is Professor of Management Science and Engineering at Stanford University; Senior Fellow of the Stanford Institute for Economic Policy Research; Senior Fellow of the Hoover Institution on War, Revolution and Peace; and Senior Fellow of the Stanford Institute for International Studies. His professional activities focus on economic policy and analysis, particularly in energy, natural resources, and the environment. At Stanford he has served as Department Chair, Director of the Energy Modeling Forum, Chairman of the Institute for Energy Studies, and Director of the Center for Economic Policy Research (now the Stanford Institute for Economic Policy Research). In the early 1970s he was Director of the Office of Energy Systems Modeling and Forecasting of the US Federal Energy Administration. He was a Founding Member of the International Association for Energy Economics and Co-editor of the Journal Resource and Energy Economics. He is Fellow of the

Contributors

xv

California Council on Science and Technology and a Member of Governor Schwarzenegger’s Council of Economic Advisors. He holds a BS Degree from MIT in Electrical Engineering and a PhD from Stanford University in Engineering Economic Systems. Michael TREBILCOCK is the Director of the Law and Economics Program at the University of Toronto Faculty of Law. His research and teaching interests over the past 36 years relate to the fields of competition policy, contract theory, international trade regulation, and economic and social regulation, all areas in which he has published widely. His numerous academic honors include the 1998 Molson Prize for contributions to the social sciences in Canada. Over the past 20 years, he has consulted widely to government and the private sector on matters of competition policy and economic and social regulation. Prof. Trebilcock was Research Director to the Electricity Market Design Committee during 1998, appointed by the Government of Ontario to design the detailed rules for the introduction of wholesale and retail competition in the ESI in Ontario in 2002. He holds an LLB Degree from the University of Canterbury, New Zealand, and an LLM Degree from the University of Adelaide, South Australia. Taff TSCHAMLER is Senior Principal at KEMA and Director of the firm’s North American retail energy practice. His primary areas of expertise include business planning, market analysis, policy assessment and financial and risk analysis of retail energy markets. He has advised numerous North American utilities, competitive energy suppliers, private equity firms and large energy buyers on market and regulatory strategy. He has published in leading industry periodicals such as The Electricity Journal and Power Economics. Prior to joining KEMA he was with Decision Analysis Corporation and ICF Consulting. Mr. Tschamler holds a Masters Degree in Public Policy from the College of William and Mary and a BA in Economics with distinction from the University of Maine. Masayuki YAJIMA is Executive Research Economist at Central Research Institute of Electric Power Industry (CRIEPI) in Japan where he is engaged in research on the electric power industry. Dr. Yajima’s research is concentrated in areas of regulatory reform, management strategy of utility companies to cope with competition as well as energy security policy. Dr. Yajima has published many books on these topics, including Deregulatory Reforms of the Electricity Supply Industry, Liberalization of Electricity Markets, Electric Restructuring, Big-Bang in Electricity Markets in the World, Energy Security, and Reconsideration of Electric Restructuring. Dr. Yajima holds a Masters Degree in Public Administration as well as a PhD in Public Administration from International Christian University in Tokyo. Jay ZARNIKAU is President of Frontier Associates engaged in retail market strategies, utility pricing, demand forecasting, and energy policy. He formerly served as a Vice President at Planergy, and prior to that, he was with The University of Texas at Austin Center for Energy Studies, and worked at the Public Utility Commission of Texas. His publications include articles in The Energy Journal, Resource and Energy Economics, Energy Economics, IEEE Transactions on Power Systems, Energy Policy, Energy, and The Electricity Journal. His past papers have estimated demand response in the ERCOT market, market concentration in ERCOT, and examined cogeneration policies in Texas. Dr. Zarnikau has a PhD Degree in Economics from The University of Texas at Austin, where he also teaches Applied Statistics as a part-time visiting professor.

This page intentionally left blank

Foreword: The Market versus Regulation STEPHEN LITTLECHILD University of Birmingham, UK; Judge Business School, University of Cambridge, Cambridge, UK

From its beginnings in Chile and the UK in the 1980s, electricity reform has spread almost throughout the world, especially in the more developed countries. The objective has generally been to increase efficiency, reduce costs, and improve quality of service. The reforms have varied in extent and scope and detail but in general they have sought to achieve this objective by relying less on public enterprise and regulated monopoly, and more on market mechanisms such as private ownership and competition. With up to two decades of experience now in hand, discussion has increasingly gone beyond theory to look at practice. Many books and papers have analyzed experience in individual countries, especially the UK and USA. There have been some comparative studies within wider geographical areas, such as Europe. But the present volume is perhaps the first, certainly the most extensive and up to date, systematic comparison of experience on a worldwide basis with the aim of identifying what works and what does not and why. The editors have invited me to focus this Foreword on the role of the market versus regulation. Each chapter in the book contains a wealth of relevant information, analysis, and stimulating interpretation, but it would be impossible to discuss and evaluate all the chapters here. Paul Joskow’s Introduction in any case does much of this job. It is a comprehensive, authoritative, indeed magisterial survey of experience worldwide. It reflects his incomparable research record and practical involvement in reform, and exhibits considered judgment. I cannot hope to better it, and I agree with it almost entirely, although we may differ on a few aspects of policy. I shall begin by briefly reiterating the aims, criteria, textbook model, and main lessons to be learned. These lessons seem to me deserving of general acceptance. Then I want to discuss four major topics – competition in the wholesale and retail markets, the nature of network regulation, and the regulation of transmission networks – where we observe not only variation in practice but also differences of view in the literature. Here I want to argue that there is a greater role for the market vis-à-vis regulation than is often accepted.1 The Aims and Assessment of Reform Proponents of electricity reform have had many and diverse aims, not always mutually consistent. The Introduction suggests that “the over-riding reform goal has been to create new 1

For simplicity of exposition, this Foreword does not include references to the literature. Relevant references can be found in my Beesley lecture which covers some of the same topics in more depth “Beyond Regulation”, Beesley Lecture Series XV, October 4, 2005 (revised version October 12, 2005) at http: //www. electricitypolicy.org.uk/pubs/misc/Beesley.pdf, forthcoming in Colin Robinson (ed.) (2007), Government and Utility Regulation (provisional title), Cheltenham: Edward Elgar. I am grateful to Paul Joskow for helpful comments on this Foreword, to Jim Sweeney for clarifying some points on California, and more generally to many other colleagues acknowledged in the Beesley lecture.

xvii

xviii

Electricity Market Reform

governance arrangements that provide long-term benefits to consumers”. These benefits are to be realized by creating competitive wholesale and retail markets to improve efficiency and responsiveness to customer preferences, by incentive regulation of privatized transmission and distribution networks to improve their efficiency and facilitate competition across them – and, I would add, by reducing the role of government and political influence generally. This fairly summarizes the aims that many of us had in the UK. I believe it also reflects the main aims in those other countries that embarked on thorough-going reform of the electricity sector. (In general this aim was not limited to the electricity sector, of course.) In assessing performance, it is necessary to adopt a “comparative governance” approach. Observed performance should be compared against clearly defined alternative institutional arrangements, recognizing that “ideal” textbook performance is never achievable in reality. In this respect, the “new institutional economics” is an important advance on the prevalent economic analysis of, say, 30 years ago. But economists have sometimes been reluctant to abandon the ideal theoretical benchmark (what Demsetz called the “nirvana fallacy”). Even now, public choice considerations do not always inform the analysis of regulatory alternatives. The “Textbook Model” for Restructuring and Competition What I have elsewhere called the “standard model” for electricity reform is well spelled out in the 10 components of the “textbook architecture” for restructuring and competition set out in the Introduction. In summary, they are: ●

●

● ●

●

●

●

● ●

●

Privatization to enhance performance and reduce the ability of the state to use these enterprises to pursue costly political agendas. Vertical separation of competitive and regulated monopoly sectors to facilitate competition and regulation. Horizontal restructuring to create an adequate number of competing generators and suppliers. Designation of an independent system operator to maintain network stability and facilitate competition. Creation of voluntary energy and ancillary services markets and trading arrangements, including contract markets and real-time balancing of the system. Application of regulatory rules to promote access to the transmission network and incentivize efficient location and interconnection of new generation facilities. Unbundling of retail tariffs and rules to enable access to the distribution networks in order to promote competition at the retail level. Specification of arrangements for supplying customers until retail competition is in place. Creation of independent regulatory agencies with adequate information, staff and powers, and duties to implement incentive regulation and promote competition. Provision of transition mechanisms that anticipate and respond to problems and support the transition rather than hinder it.

I am tempted to add a final component: Do nothing more. At least, the need to avoid excessive government and regulatory involvement is one of the lessons to be learned. The Importance of Following the Textbook Model Where the “textbook model” has been largely followed it has been broadly successful; for example, in the UK, Argentina, the Nordic countries, Victoria, and Texas. Where it has not been followed, there have been problems. Departures from the textbook model include sins of omission and sins of commission, and in some cases both.

Foreword: The Market versus Regulation

xix

Examples of the sins of omission include Belgium, Italy, and especially France, where relatively little has been done to restructure to create competitive markets (and in France’s case to privatize). Opening up to competitive markets seems to have been very slow in Japan. In Germany and New Zealand there were initial failures to recognize the need for a sector regulator. Examples of the sins of commission include Chile (excessive restrictions on generation and competition) and California (inappropriate restrictions on contracting and retail pricing and undue government involvement). Examples of both types of sins include most other North American jurisdictions, where there has often been insufficient restructuring coupled with excessive retail (and in some cases wholesale) price controls. In Ontario these problems have been compounded by undue government involvement. Similarly, continued state ownership in some states in Australia has been coupled with excessive retail price controls. In many of these cases, competition has been less effective, and prices to customers have been correspondingly higher, than would have been the case had the textbook model been adopted in full. In other cases, by contrast, prices have been artificially held below market levels, which has been the cause of different problems. The inability or unwillingness of governments to secure and defend market prices that cover reasonable costs has often precluded the full application of the textbook model. This has been the case in many developing countries, particularly in Asia.

California The problems in California have been much analyzed, not least in this volume. But since they are cited around the world as a reason for not engaging in electricity reform, it is worth a few words here to explain why that is the wrong lesson to draw, and why experience in California should not be a deterrent to electricity reform elsewhere. Commentators in this volume are right to explain that the California problems were not primarily a failure of the wholesale market. Generators wanted to invest but were slowed down by delays in regulatory approvals (which did not reflect environmental restrictions) and by regulatory uncertainty until restructuring policy was clarified. There may have been some exploitation of generator market power (to an extent that is subject to debate). But this was exacerbated, if not caused, by the regulatory framework. The main mistake was to prohibit or discourage the incumbent utilities (as retail suppliers to the majority of customers) from entering long-term contracts with the generators. This almost invited increases in spot market prices. Another problem was the obligation on the utilities – once they had covered their stranded costs and the price caps had expired – to pass directly through to customers the prices obtaining in the wholesale market. This happened in San Diego, and when prices rose sharply in the wholesale market the extent of customer protests in San Diego led directly to the intervention of the Governor of California. A third problem was the inflexible retail price caps that led to bankruptcy or near-bankruptcy for the other two utilities when wholesale prices rose. These caps did not incorporate the costs of entering long-term contracts to hedge against such price rises because the utilities had been discouraged from entering these. The refusal to raise the caps exacerbated the financial problems of the main utilities and reduced any role that reductions in demand could have played in ameliorating the electricity crisis. The main problem, in short, was one of inappropriate regulation, and was not attributable to privatization or competitive markets per se. In partial defence of the regulatory commission, it

xx


has to be admitted that the main components of the policy in question (no contracts and wholesale spot price pass through) were ones that had been advocated by leading US economists and advisers at the time. It is now generally accepted that neither of these components is appropriate, and that suitable contracting is to be encouraged. Of course, the situation in California was much exacerbated by political involvement. The then-Governor refused to allow any retail price increases in the rest of the state and declared an emergency situation. The California Legislature required that the Department of Water Resources purchase electricity on behalf of customers, which it did via large quantities of very long-term contracts at what turned out to be excessive prices. The Legislature also suspended retail competition (direct access) in order that the costs of these contracts could be passed on to retail customers. The California Public Utility Commission voted to implement this (with a strongly dissenting minority). I return to some of the retail competition issues later. But the point to emphasize here is that the proper lesson of California is to avoid over-regulation, not to avoid electricity reform. Competition in the Wholesale Generation Market International experience supports the argument for dealing with potential market power ex ante rather than ex post, and for doing so structurally rather than by restrictions on conduct. At the time of electricity privatization in the UK there was an awareness of the need for several generators from the outset, but this was outweighed by the desire to privatize the nuclear stations as well. After the nuclear stations were withdrawn, a concern to meet the privatization timetable precluded remedying this mistake. It was not straightforward or costless to remedy the situation later. Argentina, Australia, and some other countries learned the lesson and initiated a more extensive and beneficial structural separation. Much of continental Europe has not yet done so. Stronger transmission links between countries are now seen as the solution. It remains to be seen whether this will suffice to alleviate market power and political influence within each country. Most US jurisdictions have instituted market monitoring and mitigation protocols. Much thought and expertise has gone into these. But sometimes they seem rather prescriptive and severe. The Introduction reflects a concern that they may have constrained prices from rising to competitive levels, especially for peaking plant, thereby creating adverse investment incentives in the longer term. Some market monitoring protocols may be a regulatory over-reaction to early price spikes. Arguments by some economists that observed prices reflected market power because they exceeded short-run marginal cost were not helpful here. It turns out that the prices in question were often barely above the cost of staying in the market, and generally below the longrun cost of entering the market. This illustrates the disadvantages of using an inappropriate theoretical benchmark. Is it possible that concerns about market power may have been exacerbated by the wholesale market design? A characteristic of an electricity pool is that the price becomes a kind of public good, imposed on all parties uniformly without their explicit agreement (other than their acceptance of the process). This may have some advantages but the price can also become a matter of public concern. Attention and complaints naturally focus on it. An advantage of bilateral trading (as in the UK) is that there is no single imposed price and all trades are voluntarily agreed between the parties. It is encouraging that many pools are


xxi

now allowing bilateral trading outside the pool, and it will be interesting to see how both forms of market develop. Adequacy of Generation Investment A different concern is now emerging in some countries, under the heading of security of supply or resource adequacy. Can competitive markets stimulate adequate investment in new generation? Some policy-makers in particular fear that they may be held responsible for any deficit. Whether these concerns are justified is another matter. The market has already brought forth extensive investment in many different countries. There is no reason to expect that it will not be equally forthcoming in future, provided that it is not discouraged by inappropriate regulatory or government policies. As the Introduction explains, such potentially discouraging policies include macroeconomic and political instability; the uncertainty, complexity, and discouraging effects of investment by state-owned entities (including possible or threatened investment); the similar effects of some programs to stimulate uneconomic renewable sources; the costs and uncertainties of continuous market redesign; and actions by regulators or independent standard operators (ISOs) that unduly limit or depress market prices. Where such policies are discouraging generation, the remedy is surely to cease or moderate them rather than to conclude that the market does not work. Many chapters in this volume document significant investment in new generation. Are some of the authors unduly optimistic about the future? It is true that the UK has benefited from the ability to call on mothballed generation plant (new as well as older plant). It is surely an achievement of privatization and competition to have induced generating companies to consider mothballing as an economic option where it was previously ruled out. There seems no reason why other countries should not follow suit where it is economic to do so. Importantly, however, mothballing is not at the expense of new plant: there has also been and still is significant declared intention to construct new plant in the UK and no doubt elsewhere. There may be less interest in new generation in some countries if prices do not justify it at the moment. But low prices are not inherent in a competitive market, unless the market rules and monitoring guidelines so dictate. There is a danger that what emerges from the market are the prices that the regulator or market monitor deem appropriate, rather than what a competitive market would require in order to meet demand. Again, the solution may be to reconsider the nature and scope of the market rules and the monitoring guidelines. As noted earlier, bilateral trading also offers advantages here. Some countries or jurisdictions where a wholesale market already exists are implementing or considering a greater role for regulation; for example, by imposing forward contracting obligations or administratively determined capacity payments. This is disappointing: I regard it as a step forward to have abolished such a capacity mechanism in the UK. Regard must be had to the way such arrangements would work in practice rather than in theory. Market participants are quite capable of manipulating any such regulation to suit their own ends. It is also necessary to be realistic about regulation. Are regulators really capable of forecasting future demand and determining what levels of contracting obligations or capacity payments are needed to meet this demand most economically? And is this what in practice will influence their decisions? Introducing more regulation involves creating a greater role for political considerations, which are invariably influenced by a variety of factors, often of

xxii


a short-term nature. Past political concerns have included the levels of costs and prices and the extent of competition; today the concerns include resource adequacy and the environment; what will be the concerns tomorrow?

Competition in the Retail Supply Market We now understand more clearly the problems that emerge for retail competition if an inadequate regulatory framework is put in place. Defects include no or inadequate unbundling and ring-fencing of distribution networks and retail supply, inadequate provision for equal access to networks on non-discriminatory terms, cross-subsidization of one service or supplier by another, unclear allocation of costs and setting of price controls, inadequate provision for revenues to cover default services, use of default services or price controls to achieve objectives other than a robust retail market, and so on. A failure to recognize these problems has often proved critical in practice, sometimes fatal. Some might suggest that a lack of retail competition does not matter. They see costs but not many benefits. The Introduction notes that the potential benefits of retail competition include lower prices (from lower costs of energy purchasing and retailing, admittedly perhaps offset by increases in some other retailing costs), and new value added services such as risk management, demand management, and energy services. No one disputes that these benefits are likely to outweigh the costs for large and medium industrial consumers. Failure to provide for effective retail competition in this sector of the market is generally agreed to be a serious error, and policy is coming to reflect this internationally. However, some commentators are not convinced that this is the case for residential and small commercial customers. Their concern is that retail competition for such customers introduces additional costs and higher profit margins, and therefore is only feasible if regulated or default prices are increased to more than cover these additional costs and margins. These commentators prefer, and consider that customers would prefer, a regulated alternative to a competitive market. I take a different view. Although regulation may seem to offer lower costs and prices in the short term, I believe that the market will offer better value in the longer term when one considers how regulation will actually operate. Let us first identify those markets where residential retail competition has so far developed relatively well, note some characteristics of those markets, then examine what forms of regulation are realistically available.

International Experience of Retail Competition in Residential Electricity Markets Table 1 lists the main residential markets that are now open. Some of them exhibit significant switching away from the incumbent supplier, but others do not. Of those residential markets that opened about 6 years ago (in the period 1998–2000) the proportions of residential customers with non-incumbent suppliers are now 43% in UK; 29% in Sweden, and 24% in Norway but only about 11% in Finland and 5% in Germany; 26% in New Zealand but seldom over 7% in North America. In a few US states some high proportions were observed initially, but in only a few territories in those states. Of those residential markets that opened just over 3 years ago (in January 2002) the proportions are already 24% in Texas and 33% in Victoria though only 11% in New South Wales. In Ontario (which opened in May 2002) 20% of customers had signed with another supplier

xxiii

Foreword: The Market versus Regulation Table 1. Residential customer switching in international electricity markets.2 (proportion (%) of residential customers served by non-incumbent supplier). Market Markets opened 1998–2000 UK Sweden Norway Finland Germany New Zealand Alberta California Maine MPS BHE & CMP Maryland Potomac Electric Other utilities (3) Massachusetts New Jersey New York Ohio First Energy (3) Cincinnati Other utilities (4) Pennsylvania Duquesne Light PECO Energy Other utilities (4) Markets opened January 2002 Texas Ontario (open May 2002) New South Wales Victoria

After approximately 3 years

After 5–6 years

34 18 15 5 4 18?

43 29 24 11 5 26

2 2

7 1

36 0

7 0

15 0 3 0 4

6 0 3 0 6

40 2 0

45 3 0

35 18 1–7

23 2 1

19 (now 24) 23 (in September 2002, then 0) 9 (now 11) 24 (now 33)

by the day the market opened, but within a few months a price cut and subsidy by the state government led to the disappearance of the market.3 With the exception of most North American markets, the proportions of customers switching are growing steadily over time; in North America (apart from Texas) the proportions are generally static or declining. The high switching markets – notably the UK, Norway and Sweden, New Zealand, Victoria, and Texas – exhibit many other forms of competition as well. They are generally characterized by considerable entry and exit of suppliers; by growth and decline of individual suppliers; by 2

My Beesley lecture gives sources and further discussion. Whether any competitive contracts remained valid is unclear. Competitive offers (for 3- and 5-year fixed-price contracts) are now beginning to reappear in Ontario. 3

xxiv


mergers and takeovers; by a variety of marketing techniques; by active competition on price; and by an increasing variety of non-price services and product variations.4 Of particular interest is the increasing range of contractual terms. In addition to the traditional standard tariff that is variable at the utility’s (or regulator’s) discretion, the Nordic markets offer a wide range of products including fixed-price contracts varying from 3 months up to 5 years, and spot price-related contracts including a variety of optional hedges. Up to 40% of residential customers have chosen such contracts. They have exhibited a wide range of different preferences, which have also responded to changing market conditions and evolved over time. These retail electricity markets seem increasingly indistinguishable from other competitive markets such as banking, insurance and mortgages, other fuels including petrol (gasoline) and heating oil, telecommunications, food and housing, and indeed many consumer goods and services generally. These markets too involve costs to operate, some of which could no doubt be reduced if there were a single regulated supplier. Consumers there too may be “sticky”, with some reluctant to change from their traditional supplier. From time to time there may be concerns about some aspects of these markets, and a variety of restrictions may be applied to suppliers in them. But in general economists do not consider that it would be better to replace competition in these markets by a regulated outcome. Have we somehow discovered the one product in the whole of the consumer market for which regulation is better than competition?

Alternative Mechanisms or Regulating Residential Markets Consider now what form the regulation of residential electricity markets might take. The policy might be “a regime where the distribution company procures power competitively and resells it at cost”. But this sounds disconcertingly like the ideal theoretical benchmark. If nationalized industries or regulated public utilities could and would do this there would be no need for privatization and competition in the first place. Socialism and communism would work better than markets. What are the practical regulatory alternatives to retail competition? They include the following: ●

●

●

●

4

The regulator specifying ex ante the quantities and timing and prices of energy purchases that are to be made, and the terms on which this energy should be sold to customers, or approving a detailed process for doing this. The regulator approving or disapproving the above items ex post, and consequently allowing or disallowing the costs and revenues involved. The regulator using benchmarks based on purchasing by other comparable suppliers to use either ex ante or ex post in the above schemes. The regulator setting price caps for specified periods of time based on assumptions about the costs of hedging variations in energy prices.

In the UK, these variations include bundled offers notably dual fuel, credits in the form of airmiles, loyalty points with specific retailers or shopping cards (Nectar), contributions to charities and deserving customer groups, green tariffs, energy efficiency packages, insurance cover, discounts for self-reading meters, the Staywarm scheme offering unmetered electricity for a fixed monthly fee, discounts for various prepayment meter schemes, discounts for a range of home services and financial products, tariffs with no standing charges, single billing for up to six utility and other services, and so on.


●

xxv

The regulator specifying or approving ex ante the basis on which contracts for supply are to be put out to tender, and the prices at which this supply is to be priced to customers over time.

All these possibilities are “workable”, and all have probably been tried at one time or another. I myself have been involved in several of them. Some have greater merit than others. Fixedprice caps worked acceptably in the UK for a transitional period of 4 years. Price cap adjustment mechanisms have so far not been inconsistent with competition in Texas. Competitive tendering seems to have secured very good prices for customers in New Jersey and Maine, at least in the short term. But all these alternatives have disadvantages too. For example, it is easier to envisage a comparator for an incumbent supplier in Holland than in France or Italy. And even with several suppliers in areas of similar size the variations between suppliers (e.g. in customer mix, consumption patterns, weather conditions and other factors) should not be underestimated.

The Decision-Making Process Let me focus here on two aspects of the institutional comparison between regulation and the competitive market. The first aspect is the decision-making process, whether by the regulator or by market participants generally, and the information available for this purpose. I refer here not to information about what the utility companies are doing but about what the future holds for wholesale prices. Whoever is purchasing electricity, or prescribing or approving its purchase ex ante or ex post, has to decide when and what to buy. Is it better to buy ahead or on the day? In the former case, is it better to buy a week ahead, or 3 months or a year or 3 years or 15 years ahead? What should the portfolio of purchased contracts look like? Should they be fixed-price or indexed contracts, and if indexed to what input price or other parameters? There are even more far-reaching decisions to be made. Is it better to integrate vertically into generation rather than buy on the wholesale market? If so, to what extent is it prudent to integrate and what kinds of generation plant is it best to buy? There is a comparable set of decisions on the selling side, with analogous required information. Should the electricity be sold at a price that varies according to wholesale market conditions? Or should there be fixed prices and if so when should they be set and for what periods? A tender for 2 years might imply fixing the price for 2 years – but is 2 years the right period, and does this depend on views about the future level of prices? British Gas has just offered an optional fixed price for 5 years, but in Norway suppliers are now changing tariffs as frequently as monthly to reflect changing wholesale market conditions. If a single product is to be offered, which is the right one? If, instead, customers should be given a choice, how to discover what customers want and to predict what they will choose, and how to adapt to their perceptions and preferences that evolve over time? These are decisions that have to be taken in any market, and they are increasingly important in the electricity market. Economists normally argue that it is better to allow many players to take such decisions in a competitive market than to give a regulator monopoly power over such decisions. It is not that private suppliers invariably get the decisions right or that regulators are particularly unintelligent or unwise. Rather, competition is a discovery process that itself tends to identify and encourage those individuals and organizations that prove better at purchasing inputs and at understanding and providing what customers want. Competition also tends to eliminate those decision-makers that prove less able in these respects. There is no

xxvi


comparably effective discovery process for regulation. Do not the principles of competitive markets apply to electricity too? Retail Competition and Public Choice The other aspect of the institutional comparison between competition and regulation concerns how regulation actually operates. Here, both statutory and public choice considerations must be brought to bear. Is a regulator in practice able to purchase power competitively and sell it at cost? The duties of a regulatory body will typically require it to have regard to a wide range of considerations, directly or indirectly including some reference to the public interest, and generally referring to the long term as well as the short-term interests of customers. These duties would apply to the regulator in assessing what kinds of contracts the utility should purchase, and at what prices, and on what basis the power should be sold on to customers. In addition, politicians, the media, interest groups and not least the government will from time to time draw regulators’ attention to the importance of this or that consideration. Yesterday the main consideration may have been coal or promoting competition or reducing price, today it may be renewables or security of supply, tomorrow it might be demand management or nuclear. Similar considerations apply to decisions about the terms on which a regulatory body should authorize electricity to be sold to customers. The same parties will ask: Is it really appropriate to increase consumer prices just at this moment? And is it really appropriate to offer products that distinguish between this and that type of customer? These are the kinds of broader- and longer-term considerations that seem to me crucial in assessing the case for retail competition. If the aim is to facilitate certain kinds of political or non-market objectives, then regulation may have advantages over retail competition. But decisions of regulators and governments as an alternative to retail competition can be very costly to customers: witness the tens of billions of dollars associated with the coal contracts in the UK, the renewable and nuclear purchasing policies that first prompted reform in California, and the more recent insistence on expensive long-term contracts that nearly ended reform there. If the aim of electricity reform is to create governance arrangements that provide long-term benefits to consumers, then retail competition down to the residential level should be an integral component of the textbook model. The Nature of Network Regulation Incentive regulation has been extensively developed and applied in the UK and many other jurisdictions. It has led to increases in efficiency, reductions in prices, and improvements rather than reductions in reliability. In the UK, at least, there has been more capital investment than in the previous regime, not less. Fears that the mechanism is inconsistent with adequate security of supply have proved groundless. In general, consumers have benefited and so have investors. In contrast, it seems that consumers in Germany and New Zealand have suffered from the absence of such regulation. US regulatory commissions have not shown much interest in effective incentive regulation. They do not seem to have gone beyond accepting voluntary price freezes. It has been said that US regulators have no power to impose an incentive price cap. It may be that the US regulatory framework, with the obligation on the regulator to prove in course of litigation that a particular expense is unnecessary or a particular investment is not required, is not conducive to the approach.


xxvii

This makes it particularly important to look at alternative mechanisms for achieving mutually advantageous outcomes. In reality, negotiated settlements between utilities and interested parties, subsequently approved by regulators, have been an important feature of US regulation. I understand that Paul Joskow noted this in his thesis many years ago. But economists have almost completely neglected this practice. Recent research shows that the informal or negotiated settlement process differs fundamentally from the formal litigation process and has significantly different outcomes. For example, Federal Energy Regulatory Commission (FERC) has accepted settlements for the majority of gas pipeline cases. The typical outcomes have been rate moratoria that FERC itself could not impose. In Florida, the Office of Public Counsel (the consumer advocate) has gone further. It has negotiated with the utilities over three quarters of the electricity rate reductions that have been achieved over the last quarter century. This amounted to nearly $4 billions for consumers. For their part, the utilities gained more flexibility in accounting procedures, and also more attractive fixed-term revenue-sharing arrangements that provided greater incentives to efficiency than conventional rate of return regulation. The possibility of negotiated settlements instead of regulatory-determined price controls has some appeal even in jurisdictions where there is no barrier to incentive regulation via price caps. The advantages are at least two-fold. 1. First, the outcomes would reflect the preferences of the parties themselves, rather than those of the regulator. This could be particularly important in issues such as investment for reliability of supply. Certainly the Electricity Consumers’ Committee in the North of Scotland took a different view from the UK regulator in 1995, siding with the utility’s argument for more investment despite the higher price. On appeal, the Monopolies and Mergers Commission endorsed the view of the utility and the consumer committee on this issue. 2. The second advantage is that negotiated settlement potentially introduces greater variety of outcomes; for example, different kinds and durations of price caps. The regulator is effectively constrained to a uniform approach at any time. Negotiated settlements therefore offer more scope for learning from experience. The details remain to be worked out in each jurisdiction. There might need to be a regulatory backup, or at least an appropriate regulatory context. However, the principle of negotiated settlements seems worth further exploration, including in those jurisdictions that have already been keen to implement incentive regulation.

The Regulation of Transmission Networks The Introduction describes the relative lack of transmission investment in several countries, especially the USA and Europe; notes the adverse effects this has on congestion, competition, reliability, and cost; contrasts the comprehensive arrangements in the UK and Argentina with the less considered approaches in many other countries; suggests that relying primarily on market-based merchant transmission is likely to lead to inefficient transmission investment; and commends the progress being made in certain other countries. I am not well able to assess all these issues, but most of the above diagnosis is consistent with my understanding of the situation in many countries. However, I have been especially concerned about the analysis of “market versus regulation” in the context of transmission.

xxviii


The case for merchant transmission seems to have depended on the theoretical argument that, under certain conditions, it will lead to optimal investment. The case against merchant transmission is that these conditions are unlikely to be met. In simple terms, the main concern is that, because of market power and other market failures, merchant transmission is likely to be “too little, too late”, if indeed it happens at all. However, as noted earlier, the relevant benchmark is not some theoretical optimum. It is necessary to consider whether these theoretical market failures are serious in practice. It is also necessary to compare each proposed solution against the institutional arrangements that would be adopted instead. Suppose the alternative to merchant transmission is the conventional decision-making process by a transmission company and/or a regulator, with costs assigned to other parties (transmission users) regardless of outcome. There is no obvious reason why a regulated entity will be better able to predict demand. And public choice theory, backed up by considerable evidence, suggests that there will be commercial and political pressures to build excessive transmission capacity, or “too much, too soon”. In other words, there is a possibility of regulatory failure. It is therefore necessary to compare the alternatives as they would work in practice. Which is likely to exhibit the more serious failure?

Merchant and Regulated Interconnectors in Australia Experience in Australia has been an eye-opener. Both merchant and regulated transmission lines have been built. All are interconnectors between electricity regions rather than expansions within a single system. However, they clearly illustrate some of the important factors at work. Two merchant lines have been built. They seem to have underestimated the speed and extent to which new generation in the high price regions would reduce the price differentials between the regions. This means they overestimated the economic benefits and profitability of the lines. Their market power has in practice been negligible. Far from being “too little, too late” they both appear to have been “too much, too soon”. Their investors have had to foot the bill for the misjudgments of the value of the interconnectors. They have learned their lesson and settled for a regulated income. One regulated line has been built broadly in parallel to the first merchant line. Arguably it misjudged the demand even more severely: it was about five times the size of the merchant line and correspondingly much more uneconomic. Transmission users rather than investors are having to foot this bill. Separately, another transmission company and the regulatory body would have built another regulated line, duplicating the second merchant line and being wholly redundant, if the courts had not stopped them from doing so. What has been or would be the experience of transmission expansions within an electricity system, sometimes called “intensive network upgrades”, deserves further research. But in the case of interconnectors between electricity regions in Australia, there seems no doubt that regulatory failure has been more serious than market failure.

The Public Contest Method in Argentina An alternative to both regulated and merchant transmission is the Public Contest method used in Argentina. Transmission expansions have to be proposed, approved, and financed by users themselves. Construction is then put out to tender. The users typically comprise generators, distribution companies, and large industrial consumers. This arrangement was carefully designed to avoid the inefficient over-expansion that characterized the pre-privatization


xxix

era in Argentina. It was envisaged, correctly, that if users who benefited from an expansion had to pay for it as well, they would give more careful consideration to whether it really was worthwhile. Although the approach has received some unjustified criticism, it has actually worked rather well. The main concern was that users initially turned down a long-expected Fourth Line bringing power to Buenos Aires. But on closer examination the line turns out to have been uneconomic. Nowadays it is cheaper to transport gas to Buenos Aires and to generate power there. The Public Contest method has a good record. It enabled a variety of economic expansions to go ahead, halved the cost of building transmission lines, greatly increased the productivity of the transmission system, and resisted the political demands for uneconomic expansions in outlying regions.

Analyzing Institutional Arrangements for Transmission My argument is not that merchant transmission or the Public Contest method is necessarily the best approach in all circumstances. Nor do I propose no role for regulation. Rather, the economic analysis needs to be more consistent with the comparative institutional approach. The proposed rules about performance assessment (see Introduction) need to be applied in designing transmission (and other) arrangements ex ante as well as in assessing them ex post. Lessons also need to be learned from actual experiences. The examples I have cited suggest that the concerns about merchant transmission and user-determined expansions may not be as great in practice as feared in theory. And the alternative of regulated transmission has its own disadvantages that need to be taken into account. The chosen institutional mechanism must detect and prevent those transmission expansions that are excessive or uneconomic as well as discover and implement those that are needed and economic. There is indeed a great deal of work to be done on the creation of effective institutional arrangements to achieve this.

Conclusion To be successful, electricity reform must reflect a good understanding of what makes markets work well and what prevents this happening. Experience and analysis of many electricity markets over the last two decades is leading to broad agreement on the basic prerequisites. There is still scope for debate on the relative roles of regulation and the market. I have argued that regulation can usefully be reduced and the role of the market increased, not only at the wholesale and retail levels but also with respect to monopoly networks. The contributors to this volume have made a substantial contribution to advancing our understanding of all these issues.


Introduction to Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies PAUL L. JOSKOW1 Department of Economics, Centre for Energy and Environmental Policy Research, Massachussetts Institute of Technology, Cambridge, USA

Introduction During the 1990s, many countries began to restructure their electric power sectors with the goal of improving sector performance.2 The restructuring programs have included privatization of state-owned enterprises, the separation (ownership or functional) of potentially competitive segments (generation and retail supply) from segments that have natural monopoly characteristics and are expected to continue to be subject to price and entry regulation (distribution and transmission), the creation of competitive wholesale and retail markets, and the application of performance-based or incentive regulatory mechanisms (PBR) to the remaining regulated segments to complement traditional cost-of-service regulation. We now have a significant amount of experience with the performance of these restructuring, regulatory reform and competition programs. Evaluating the performance record is useful in order to provide guidance to policymakers considering comprehensive electricity sector reforms where they have not yet occurred, to guide refinements in the reform programs aimed at improving performance where they have already been implemented and, importantly, to provide factual evidence to respond to questions about sector performance from policymakers who continue to be skeptical about the value of at least some of the reforms that have been implemented or are being proposed. This book contains studies of the electricity sector reform programs, their consequences and remaining policy issues for a large number of countries. The studies cover countries with diverse characteristics. Both developed (e.g. Britain, Central Europe, the Nordic countries, US, Canada, New Zealand) and a few developing countries are covered (e.g. Columbia, Argentina, Brazil). It covers countries or states/provinces that entered the reform process with primarily privately owned regulated vertically integrated utilities (e.g. US, Japan, Germany, Alberta) and countries or states/provinces whose electricity sectors were dominated by state-owned

1

I am grateful for research support from the MIT Center for Energy and Environmental Policy Research and the Cambridge-MIT Institute. Comments from Stephen Littlechild were very helpful. This paper draws heavily on Joskow (2000b, 2003, 2005a, 2006). 2 A few countries started comprehensive restructuring programs earlier (e.g. Chile and the UK).

1

2


monopolies (e.g. Britain, much of Central Europe, Argentina, the Nordic countries, Ontario).3 The nature of the reform programs has also varied quite widely, from comprehensive “textbook” restructuring, regulatory reform and competition initiatives (e.g. Britain, Argentina, Texas, New Zealand, Alberta) to those that have tried to introduce competition with minimal structural or regulatory reform (e.g. Germany, France, Japan) and examples of reform programs that fall between the two extremes (e.g. Chile, California, the states within Pennsylvania–New Jersey–Maryland regional transmission operator (PJM)). Since the reforms were implemented at different times and with different rates of change in different countries, the experience with meaningful reforms that we can draw upon varies from more than 10 years (e.g. Britain, Argentina, Chile) to much shorter time periods (e.g. Texas, Germany, Japan, Ontario, Brazil). In all of the countries covered, the initial reforms were eventually followed by a “reform of the reforms” in response to real or imagined problems that emerged over time. This diversity in experience is valuable since it provides an opportunity to better understand how different approaches to restructuring, competition and regulatory reform affect sector performance. However, experience drawn from different countries also needs to be utilized with some care. The objective levels of performance in different countries, in terms of costs, productivity, price levels and price structures, service quality, etc., varied widely from country to country when the reforms were implemented. The effects of the reforms on these and other performance indicia are likely to vary as well, reflecting where the sectors started as well as the nature of the reform initiatives themselves. The countries covered in the chapters in this book also vary widely in the basic legal and political infrastructure upon which the reforms had to proceed. This institutional infrastructure includes attributes, such as laws governing contracts and property rights, the powers and independence of the judicial system, experience with and effective authority of independent regulatory agencies, cost accounting, auditing and performance measurement systems, state/province versus federal authority and variations in the faith in and commitment of policymakers to competitive markets. It is particularly dangerous to draw inferences from the experience in developed countries for the design of reform policies in developing countries without carefully taking into account the differences in starting points and institutional infrastructures. The process of continuing “reform of the reforms” may provide additional insights into how different structural and regulatory mechanisms affect performance. However, continuing reform itself – an absence of a stable institutional environment for the sector – may also have adverse affects on sector performance, especially on investment incentives.

Why Restructuring, Regulatory Reform and Competition? In order to evaluate the performance of electricity sector reforms we must articulate clearly what the goals of the reforms are meant to be. I realize that some reform proponents believe that competition per se is a primary goal. Most policymakers have more precise performance goals in mind. And it is against these goals that the effects of the reforms should be measured. Electricity sectors almost everywhere on earth evolved with (primarily) vertically integrated geographic monopolies that were either state-owned or privately owned and subject to price and entry regulation as natural monopolies. The primary components of electricity 3

Although the electricity sectors in most countries started with private firms, state ownership typically came later, often after World War II. There are useful lessons that may be drawn from a careful consideration of the factors that led to a transition from private to state-owned firms in the mid-20th century, but that is not the focus of this book.

Electricity Sector Liberalization: Lessons Learned from Cross-Country Studies

3

supply – generation, transmission, distribution and retail supply – were integrated within individual electric utilities. These firms in turn had de facto exclusive franchises to supply electricity to residential, commercial and industrial retail consumers within a defined geographic area. The performance of these regulated monopolies varied widely across the countries. In many developing countries, the sectors were characterized by low labor productivity, poor service quality, high system losses, inadequate investment in power supply facilities, unavailability of service to large portions of the population and prices that were too low to cover costs and support new investment (Besant-Jones, 1993; World Bank, 1994; Bacon and Besant-Jones, 2001). Industrial customers sometimes had to respond to frequent system outages by building their own isolated generating facilities, increasing their costs of doing business. Sector performance in developed countries was generally much better (Joskow, 1997), but high operating costs, construction cost overruns on new facilities, costly programs driven by political pressures, wide variations in performance across firms with similar supply opportunities, and high retail prices required to cover these costs stimulated pressures for changes that would reduce costs and retail prices (Joskow, 1998, 2000a). The overriding reform goal has been to create new governance arrangements for the electricity sector that provide long-term benefits to consumers. These benefits are to be realized by relying on competitive wholesale markets for power to provide better incentives for controlling construction and operating costs of new and existing generating capacity, to encourage innovation in power supply technologies, and to shift the risks of technology choice, construction cost and operating “mistakes” to suppliers and away from consumers. Retail competition, or “customer choice,” is supposed to allow consumers to choose the retail power supplier offering the price/service quality combination that best meet their needs. Competing retail suppliers were then expected to provide an enhanced array of retail service products, risk management, demand management and new opportunities for service quality differentiation to better match individual consumer preferences. They also increase competition on the buying side in the wholesale market. It has also been widely recognized that significant portions of the total costs of electricity supply, distribution and transmission, would continue to be regulated. Accordingly, reforms to traditional regulatory arrangements governing the distribution and transmission networks have generally been viewed as an important complement to the introduction of wholesale and retail competition to supply consumer energy needs. Privatization of distribution and transmission companies combined with the application of PBR regulation imposes hard budget constraints on regulated network firms and provides better incentives for them to reduce costs and improve service quality (Beesley and Littlechild, 1989; Joskow, 2005e). In addition, the efficiency of competitive wholesale and retail markets depends on a wellfunctioning supporting transmission and distribution network infrastructure. These goals are to be achieved in a way that is consistent with environmental laws, regulations and policy goals. In many countries, the electricity sector is a major producer of conventional air pollutants (SO2, NOx, particulates) and CO2 emissions. Generating, distribution and transmission facilities can have significant impacts on land use and are not perceived as attractive neighbors by most individuals. Generating plants use large amounts of water for cooling and the release of warm water can have ecologic impacts. They also produce significant amounts of toxic waste which must be disposed of properly. Moreover, the electric power industry has been a target of environmental regulations (and related regulations governing energy efficiency and renewable energy programs) at levels that are greater than the sector’s relative impact on the environment, compared to the industrial sector, for example. I believe that this is the case because the electricity sector has been composed of regulated vertically integrated monopolies, perceived to be immune from competition from other states, provinces

4


and countries, and the sector provided an attractive target for “taxation by regulation,” allowing politicians to bury the costs of these programs in regulated electric power prices. These traditional approaches to use the electric power sector to pursue environmental policy agendas may not be compatible with the successful development of competitive markets. Alternative approaches that are more compatible with competitive markets may be necessary.

Textbook Architecture For Restructuring and Competition While a number of variations are potentially available (Hunt, 2002; Joskow, 2000a, 2005a), it is my view that the “textbook” architecture for restructuring, regulatory reform and the development of competitive markets for power involves several key components: 1. Privatization of state-owned utilities to create higher-power incentives for performance improvements and to make it more difficult for the state to use these enterprises to pursue costly political agendas. The components of these political agendas have included the use of state-owned monopolies for patronage employment, macroeconomic and redistributive policies, to favor domestic suppliers of fuel and equipment, and to funnel revenue to government budgets outside of the tax system. 2. Vertical separation of competitive segments (e.g. generation, marketing and retail supply) from regulated segments (distribution, transmission, system operations) either structurally (through divestiture) or functionally (with internal “Chinese” walls or “ring fencing” separating affiliates within the same corporation). These changes are thought to be necessary to guard against cross-subsidization of competitive businesses from regulated businesses and discriminatory policies affecting access to distribution and transmission networks upon which all competitive suppliers depend. 3. Horizontal restructuring of the generation segment, to create an adequate number of competing generators to mitigate market power and to ensure that wholesale markets yield reasonably competitive performance results. 4. Horizontal integration of transmission and network operations to encompass the geographic expanse of “natural” wholesale markets and the designation of a single independent system operator to manage the operation of the network, to schedule generation to meet demand and to maintain the physical parameters of the network (frequency, voltage, stability). These structural changes are necessary to provide an efficient platform for wholesale and retail competition to proceed consistent with the physical constraints and operating protocols that must govern the operation of any electric power network that meets standard reliability criteria. Horizontal integration eliminates inefficient institutional seams between physically synchronized networks, allows for more effective use of network capacity, expands the geographic expanse of competition, reduces distortions caused by inefficient transmission prices and supports the operation of wholesale markets with a minimum of intervention by system operators or regulators. An independent system operator (whether in the form a system operator with responsibility only for balancing supply and demand in real time consistent with the network’s topology and reliability criteria or a “Transco” that owns and operates the network’s transmission facilities as well) is needed to guard against discrimination that might arise if there is common ownership between markets participants who use the network and the entity that operates the network and supporting market mechanisms. 5. The creation of voluntary public wholesale spot energy and operating reserve market institutions to support requirements for real time balancing of supply and demand for


6.

7.

8.

9.

10.

5

electric energy, to allocate scarce network transmission capacity, to respond quickly and effectively to unplanned outages of transmission or generating facilities consistent with the need to maintain network voltage, frequency and stability parameters within narrow limits, and to facilitate economical trading opportunities among suppliers and between buyers and sellers. Physical (or near physical) spot wholesale power markets do not design themselves. They must be designed as part of the restructuring process. There are differences in opinion about what the most effective design elements are, but there should be no difference of opinion about the need for (at least) public balancing markets for energy and ancillary network support services that cover a geographic region that encompasses as large a fraction as possible of the energy trading that can take place using the network. The design of public spot energy and ancillary services markets should be compatible with the evolution of private markets for bilateral forward contracts for energy and associated derivatives, including instruments that can be relied upon to hedge basis risk associated with transmission congestion, power exchanges and other institutions to facilitate financial arrangements between buyers and sellers. The application of regulatory rules and supporting network institutions to promote efficient access to the transmission network by wholesale buyers and sellers in order to facilitate efficient competitive production and exchange, including mechanisms efficiently to allocate scarce transmission capacity among competing network users, and to provide for efficient siting and interconnection of new generating facilities. The unbundling of retail tariffs to separate prices for retail power supplies and associated customer services to be supplied competitively from the regulated “delivery” charges for using distribution and transmission networks that would continue (primarily) to be provided by regulated monopolies. This makes it possible for retail consumers eligible to choose their power suppliers competitively to purchase their power supplies from competing retail suppliers without having to overcome barriers caused by behavior of incumbents that may have the effect of increasing entry barriers for independent competitive suppliers. Where policymakers have determined that retail competition will not be available (e.g. for domestic and small commercial customers), distribution companies or alternative designated suppliers would have the responsibility to supply these customers by purchasing power in competitive wholesale markets or, if they choose, to build their own generating facilities to provide power supplies. However, in the latter case the associated charges for power would be subject to wholesale market-based regulatory benchmarks, primarily competitive procurement processes. The creation of independent regulatory agencies with good information about the costs, service quality and comparative performance of the firms supplying regulated network services, the authority to enforce regulatory requirements, and an expert staff to use this information and authority to regulate effectively the prices charged by distribution and transmission companies, and the terms and conditions of access to these networks by wholesale and retail suppliers of power, are also important but underappreciated components of successful reforms. Regulators should rely on well-designed PBR mechanisms that meet budget balance, rent extraction and efficiency criteria, given the information available to them (Joskow, 1998), and must create a stable and credible regulatory environment that will support the attraction of the capital needed to improve the performance and expand the regulated network platforms upon which competition depends. Transition mechanisms that are compatible with the development of well-functioning competitive markets need to be built into the reform program. If there is anything to be learned from the chapters in this book it is that there will be problems that need to be addressed as the reforms proceed. It makes sense to anticipate them and to build effective

6

Electricity Market Reform transition mechanisms into the reform process. These include forward contracts between generators and distributors with continuing retail customer service obligations, retail default service pricing arrangements available to retail customers before they switch to retail suppliers, and market monitoring and mitigation mechanisms. The challenge is to identify transition mechanisms that are temporary and actually support a transition rather than mechanisms that act as barriers to the development of competitive markets. It must also be recognized that it is much easier to smooth the transition when there is excess generating capacity and little network congestion than when the system has tight supplies, a lot of congestion that creates small “load pockets” and requires additional investment in generation and transmission capacity quickly to meet demand reliably.

As I read the chapters in this book it is quite clear that a few countries/states/provinces followed this textbook restructuring architecture, others embraced several of its key components, while still others took a minimalist approach to restructuring to start with and then made sequential reforms in response to (often bad) experience. As I will discuss further below, it is fairly clear that electricity sector reforms designed to create competitive wholesale and retail markets yield better performance if they follow the major elements of this textbook restructuring program than if they do not. Performance Assessments There have been few comprehensive “social cost benefit” assessments of the effects of electricity restructuring in specific countries. Newbery and Pollitt’s (1997) analysis of the welfare consequences of reforms in the UK is an exception, though it covers a period that precedes changes in the ownership concentration of generation and changes in wholesale market institutions that increased competition in the wholesale market and led to lower price–cost margins. There has been much more work on individual segments of the industry affected by the restructuring programs in particular countries (e.g. labor productivity in generation and distribution; integration of wholesale markets, investment in generation) and more of what I would call studies that examine “fragments of evidence” associated with the performance of specific segments of the sector. Most of the chapters in this volume fall in the latter two categories. Nevertheless, they convey useful information and exhibit some common themes. One of the challenges that must be confronted in doing a performance assessment is to choose a suitable counterfactual benchmark for comparison purposes. That is, we need to measure various performance metrics and compare them with what these metrics would have been if the reforms had not been made at all or if they had been made differently. The easiest approach is to examine changes in performance over time using time series data and to attribute improvements or deteriorations in performance “before and after” the reforms were implemented to the reforms themselves. This type of analysis is always based on explicit or implicit assumptions about how various exogenous variables affect performance over time and whether or not they have been controlled for adequately. So, for example, in examining the behavior of wholesale electricity prices, controlling for changes in the prices of fuels used to generate electricity over time would be important.4 Another approach is to compare the performance of electricity sectors in countries or states/provinces that have implemented reforms with those that have continued with the traditional industry structure and regulatory arrangements. This kind of approach has 4

Although the fuel price itself may be affected by the restructuring program as for coal prices in Britain before and after restructuring.


7

been utilized extensively in the US to compare performance of various regulatory and deregulatory policy initiatives in different states that have implemented particular policies in different ways (Joskow and Rose, 1989). In a recent paper (Joskow, 2006) I have attempted to make a preliminary assessment of the effects of wholesale and retail competition programs on retail prices using cross-state data over time. This kind of cross-sectional approach has not yet been used extensively or systematically to evaluate electricity sector reforms and the chapters in this book provide a vehicle for doing this kind of cross-sectional comparison using case studies for different countries. A third approach to evaluate the performance of regulatory reform and deregulation initiatives is a “structural” simulation approach (Joskow, 2005c). Here one uses historical experience to construct a structural simulation model for the behavior and performance of the industry under different institutional arrangements. The model is then used to simulate what key performance metrics would have been if historical institutional arrangements had continued per the status quo. The simulated performance can then be compared to actual performance. With a good structural model one can also simulate how policy changes that are different from those that were implemented would have affected performance. This kind of approach has been applied to electricity restructuring policies primarily to evaluate wholesale market power issues. Newbery and Pollitt (1997) produce a counterfactual for England and Wales based on various assumptions about the incumbent state-owned generation and transmission company’s investment plans and performance trends. Green and Newbery (1992) develop a simulation model based on a particular model of imperfect competition that is parameterized to match the supply and demand attributes of the electricity sector in England and Wales in 1990. The model is used to simulate wholesale prices, generator profits, entry, consumer and producer welfare under different market structures. Wolfram (1998, 1999) effectively tests the predictions of that model using ex post data on actual prices, costs and bidding behavior. A related competitive benchmarking approach has also been applied to the evaluation of wholesale pricing behavior in California during the first few months of the so-called California electricity crises in 2000–2001 (Borenstein et al., 2002; Joskow and Kahn, 2002). All three of these approaches can provide useful insights into the effects of policy reforms on various performance indicia. However, in each case it is important to adopt what Oliver Williamson (1985) refers to as a comparative governance approach to the evaluation of the performance of alternative institutional arrangements for any industry. It has two components: (a) performance assessments must recognize that observed performance should be compared with performance under a clearly defined alternative set of institutional arrangements and (b) “ideal” textbook performance that we associate, for example, with perfectly competitive markets, is never achievable in reality. Policymakers should be looking for the best that they can do in an imperfect world.

Lessons Learned 1. Electricity sector reforms have significant potential benefits but also carry the risk of significant potential costs if the reforms are implemented incompletely or incorrectly: Based on the analyses presented in the chapters in this book, as well as other studies of electricity sector reforms and their performance, I believe that it is fair to say that when electricity restructuring and competition programs are designed and implemented well, electricity sector performance can be expected to improve significantly compared to either state-owned or private regulated vertically integrated monopolies. Note that this conclusion is not inconsistent with a finding that there are some regulated vertically integrated monopolies that perform quite

8


well and that, in such cases, the kinds of comprehensive reforms reflected in the textbook model might have little positive effect on performance. And if the reforms are poorly or incompletely implemented, could even lead to a deterioration in performance. Rather, it is a statement about what expectations policymakers, faced with imperfect and asymmetric information about the performance of the regulated sector, should have in the typical cases. These expectations can and should be refined by policymakers by using whatever information they have available about the pre-reform levels of performance and how far their sector is from the performance frontier. As I indicated earlier, the pre-reform levels of performance varied widely across countries and even across firms in the same country. The scope for improvements from reform initiatives necessarily vary as well. Policymakers should take the starting point into account because, as we shall see, there is always the chance that design or implementation flaws in the reforms can lead to performance problems of various kinds. Several chapters in this volume along with other research make it clear that successful implementation of liberalization reforms is not easy and that there is a risk that costly performance problems may emerge. California is the textbook case of reforms gone bad, though it is not at all clear that the right lessons have been learned from that experience. I will discuss the California case in more detail presently. As described in the chapters examining the reforms in Ontario and Brazil, they have also experienced very significant problems. Wholesale markets with good performance attributes have been slow to emerge in some countries. The promised benefits of retail competition for residential and small industrial customers have been slow to be realized in many countries. The mobilization of adequate investment to expand generation, transmission and distribution capacity has been a problem in many of the countries that have implemented reforms. Getting the reforms right at the outset is very important. 2. The textbook model of restructuring, regulatory reform and market design is a sound guide for successful reform: The use of the phrase “deregulation” to characterize the attributes of the most successful electricity sector reform programs is misleading. This is not the trucking industry and the traditional industry structure based on vertically integrated regulated monopolies is not conducive to simple “deregulation” without supporting structural, regulatory and market design reforms (Joskow and Schmalensee, 1983)! Restructuring, regulatory reform, wholesale and retail market design, and deregulation of competitive wholesale and retail segments go together. The most successful reform programs discussed in this volume and elsewhere have followed the “textbook model” outlined earlier fairly closely: privatization of state-owned enterprises, vertical and horizontal restructuring to facilitate competition and mitigate potential self-dealing and cross-subsidization problems, PBR regulation applied to the regulated transmission and distribution segments, good wholesale market designs that facilitate efficient competition among existing generators, competitive entry of new generators, and retail competition, at least for industrial customers. In my view, the gold standard for electricity sector reform is England and Wales (as discussed in David Newbery’s chapter). The reforms followed the basic architecture of the textbook model and have led to significant performance improvements in many dimensions. This is not to say that everything worked perfectly. Clearly, the decision to create only three generating companies, two of which set the clearing price in the wholesale market in almost all hours, led to significant market power problems that persisted for several years. Not only were wholesale prices too high, but there was probably an inefficiently high level of entry of new gas-fired combined-cycle gas turbine (CCGTs) technologies. Congestion on the transmission network made some generators “must run,” creating an additional “locational” market power problem. However, a combination of entry of new generators, divestitures of existing generating plants by incumbent suppliers, and transmission investments has made


9

the wholesale market structurally more competitive over time. Price–cost margins eventually fell dramatically and there is a lively debate about whether it was the reduction in seller concentration or the introduction of the New Electricity Trading Arrangements (NETA) to replace the pool that is the cause of the reduction in market power observed in the last few years (Evans and Green, 2005). Privatization and the application of high-powered regulatory mechanisms have led to improvements in labor productivity and service quality in electric distribution systems in England and Wales as well (Domah and Pollitt, 2001). The application of incentive regulation mechanisms to the independent transmission company also led to a dramatic reduction in the costs of managing network congestion and the costs of balancing the system and maintaining network reliability. During the 1990s there was substantial entry of new generating capacity, largely replacing existing generating capacity (that eventually retired), rather than to meet a need for new capacity to meet growing peak demand. The retail competition program in England and Wales has been reasonably successful, though there continue to be debates about whether the benefits of extending retail competition to domestic (residential) customers was worth the costs (Newbery chapter on Britain, Green and McDaniel (1998) and Salies and Waddams Price (2004)). England and Wales is not the only country that has followed the textbook model. Argentina followed most features of the basic textbook model and, prior to the country’s macroeconomic collapse, currency crisis, and rejection of contractual and regulatory commitments in 2002, experienced excellent performance. Argentina experienced significant improvements in the performance of the existing fleet of generating plants, significant investment in new generating capacity, and improvements in productivity and a reduction in losses (physical and due to thefts of service) on the distribution networks (Argentina/Columbia chapter, Estache and Rodriguez-Pardina (1998); Bacon and Besant-Jones (2001); Rudnick and Zolezzi (2001); Pollitt (2004)). Unlike the case in England and Wales, Argentina made a serious effort at the outset to create a generation sector that was structurally competitive and there is little if any evidence of market power in the wholesale market there. These improvements in performance indicia were realized despite (or perhaps partially because of) the fact that Argentina did not have a real unregulated spot market for electricity. Following the model established in Chile, Argentina’s so-called spot market was structured as a securityconstrained marginal cost-based power pool in which the clearing price is determined mechanically by the marginal cost of the generator that clears the market in an efficient costbased merit order dispatch. This mechanism effectively caps prices in the spot market at very low levels (about $150/MWh) under scarcity conditions. However, the spot market revenues are supplemented by revenues from a capacity payment mechanism. As discussed by Adib and Zarnikau in their excellent chapter, Texas also took a comprehensive approach to restructuring, regulatory reform and market design that followed many of the basic attributes of the textbook model. However, rather than adopting a poolbased wholesale market as in the UK and Argentina, Texas took an approach to wholesale market design that relied as much on bilateral contracts and as little on organized public markets operated by the ISO as possible. Texas also endeavored to implement structural remedies (i.e. generation divestiture) to respond to concerns about market power. However, congestion management and associated market power issues have been an area of concern in Texas and it appear that Texas will be moving to a nodal pricing model more like those operating in the Northeast in the future. Texas also adopted an approach to retail competition that is similar to that adopted in the UK, except retail competition was opened to all classes of customers from the beginning. At least in terms of switching behavior, Texas has the most successful retail competition program

10


in the US, especially for smaller customers. However, I question the estimates of customer savings achieved by retail competition based on comparisons between the prices available from competitive retailers and the “price to beat.” I do not think that the “price to beat” is the appropriate benchmark against which to measure actual or potential customer saving relative to what prices would have been under regulation, especially with extremely high natural gas prices. Indeed, there is evidence that retail prices in Texas, especially for customers who have not switched, are higher than they would have been during the short period of time retail competition has been in effect in Texas (Joskow, 2006). If regulation had continued, customers would have been partially hedged against changes in natural gas prices as the costs of nuclear and coal-fired capacity would have been averaged in with the cost of gas-fired capacity in determining regulated retail prices. The price to beat does not reflect this hedge. Nor do competitive wholesale market prices in markets that clear with gas-fired capacity during a large fraction of the hours in the year reflect this hedge (as in New England). New Zealand, portions of Australia and the Nordic countries adopted many of the key components of the textbook model and have had reasonably successful reform programs, though retail competition opportunities vary between these countries. Australia, the Nordic countries, Ontario, Australia and Brazil have proceeded with their reforms without fully privatizing the generation segment of the sector. The continued mix of public and private generating companies raises some interesting issues for short-run market performance and longer-run investment incentives that I will return to presently. Chile is often identified as the first country to adopt the textbook electricity sector reform model. While I believe that the Chilean reforms have led to large efficiency improvements compared to what proceeded them, and that there is much to be proud of in the reforms that were made there in the beginning of 1980s, the Chilean system has involved less restructuring, less competition and more regulation than first meets the eye (Joskow, 2000b). While Chile theoretically separated generation, transmission and distribution, for many years a single holding company owned the largest generating company, the primary distribution company in the Santiago region and the primary transmission company serving the largest region of the country.5 The generation segment is not structurally competitive and there is significant potential market power that would be exercised if Chile really had a competitive wholesale market, but it does not. What is generally referred to as a spot market in Chile is not really a market in the same sense as are the spot markets for energy in PJM, Alberta, England and Wales, or Norway. Indeed, it is little different from the pre-restructuring centrally dispatched power pools that existed in the US for the last few decades. Generators are dispatched based on estimates of their marginal production costs and the marginal cost of the last supply unit called to meet demand determines the market clearing price. Network congestion and constraints are centrally managed by the system operator (the CEDEC in Chile) in conjunction with the “least-cost” dispatch of generators. Generators have incentives to keep their costs low and their availability high under this system. However, the “spot market” does not yield prices that are high enough when generating capacity is fully utilized to balance supply and demand (the effective price cap is on the order of $150/MWh) and, as a result, scarcity rents are too low to attract adequate investment in generation to the market. Nor has there been a market mechanism to pay for operating reserves and other ancillary services. Accordingly, it should not be surprising to find that the incentives to invest

5

The transmission company Transener was sold to an affiliate of Hydro Quebec by Endesa after its acquisition of the Enersis holding company which had owned the largest distribution, transmission and generating companies in the SIC region of Chile.


11

in new generating capacity have been inadequate. Moreover, the government’s decision to go out to tender for additional generating capacity is not an attractive solution to the investment incentive problems caused by wholesale and retail market imperfections. On the retail competition front, large industrial customers in Chile were theoretically free to contract directly with generators for their supplies (though they are not permitted to buy directly from the “spot market”), but as a practical matter only the very largest customers which could connect directly to the high-voltage transmission system have this opportunity. I understand that the distribution company serving Santiago had many customers that theoretically could contract directly with generators, but very few actually took advantage of this opportunity. The reasons are (a) there was no unbundled delivery tariff available to customers that separated delivery charges from generation charges and (b) generators were reluctant to steal the distribution company’s retail customers since the distribution company itself was a major contract purchaser of the generators’ wholesale power supplies. Under the Chilean model, distributors were supposed to enter into contracts with generators to meet their load obligations, but the prices in these contracts are regulated based on forecasts of nodal prices in the “wholesale market” and the associated costs are, in turn, passed through to retail consumers. The nodal prices are theoretically collared by the “free market” prices paid by large industrial customers – they must be in a ⫾10% band of negotiated contracts between generators and large customers, but, as I have just noted, the competition in the free market is more limited than first meets the eye.6 Accordingly, there is not really a competitive wholesale contract market either. Theoretically there has been free entry of new generators in Chile for many years, but until a new law was passed in 2004 there was no open access transmission tariff,7 no obligation placed on transmission companies to plan for or build transmission capacity in advance, and the major transmission company was owned by the major generator and they must, by necessity, interact closely with one another. There has been little entry of new generating companies in Chile, though existing generating companies have expanded generating capacity significantly over time.8 Nevertheless, creating adequate investment incentives for new generating capacity is a continuing issue in Chile. So, whatever the success that the Chilean reforms achieved, they are not primarily the result of vibrant unregulated competitive wholesale or retail markets for electricity or real vertical and horizontal restructuring. Privatization, incentive regulation, a simulated competitive spot market, contractual obligations placed on distribution companies and free entry by incumbent suppliers in response largely to administratively determined generation prices have all contributed to the performance improvements. California and many of the Northeastern US states appear to have adopted many of the components of the textbook model as well. Yet California is often put forward as the textbook case of “deregulation” gone bad. The California restructuring and competition program (but not the T&D regulatory framework) were heavily influenced by the earlier reforms in England and Wales. Several of the commissioners of the California Public Utilities Commission (CPUC) visited England in early 2004 and were very impressed with what they saw there. The initial reform proposals contained in the so-called “blue book” included many of the features of the reform program in England and Wales. And, although disputes about wholesale

6

Prices negotiated in the free market are confidential and I have seen no analysis that indicates whether the negotiated contract prices are a binding constraint and, if they are, how large is their effect on nodal prices. 7 Such as provided for by Order 888 in the US or the Grid Code in England and Wales. 8 The relatively recent availability of natural gas in Chile may make entry of new suppliers and competition from cogeneration easier in the future.

12


and retail market design led eventually to a reform program that departed from several aspects of the textbook model, it still retained many of its basic features. California’s utilities were effectively required to divest their fossil-fueled generating plants in a way that led to a significant reduction in horizontal concentration and the instant creation of a large independent power sector. California’s generating sector had relatively low levels of concentration after the incumbent utilities divested their gas/oil-fired generating capacity, especially when California’s substantial interconnector capacity is taken into account. Moreover, the financial and regulatory commitments placed on the incumbent utilities during a transition period destroyed any incentive they might have had to exercise market power using their remaining generating plants. An independent system operator was created to consolidate the control areas of California’s three investor-owned utilities and to operate a single control area transmission network and manage congestion. Public day-ahead spot markets for power (the PX), a separate real time balancing market for energy (operated by the ISO) and ancillary service markets were created (operated by the ISO), and there was some integration between these markets and the allocation of scarce transmission capacity yielding a set of zonal prices (though intra-zonal congestion was not priced transparently and the associated costs were included in an uplift charge). All customers were made immediately eligible for retail competition or so-called “customer choice.” While there were a number of problems that emerged in the wholesale markets in California after they began operating in 1998, wholesale market prices between April 1998 and April 2000 were roughly equal to what had been predicted (3 cents/kWh) prior to the reforms and well below the embedded cost of regulated generation service (about 6.5 cents/kWh). The meltdown began in June 2000 and continued for about a year. What was the source of the problems? Interestingly, the New England states, New York, New Jersey and Pennsylvania had implemented very similar reforms at about the same time and experienced some of the same exogenous shocks to demand and fuel prices in 2000 and 2001. Yet they did not experience the same system meltdown as did California. So, there is something to learn as well from comparing some of the more detailed aspects of the reforms in California with those in these other US states. Many explanations have been advanced to explain what happened in California. One interpretation of what transpired and why can be found in James Sweeney’s chapter in this volume. My views, written at about the time the crisis was winding down and before the Enron and other marketers’ tapes were released, can be found in Joskow (2001). The most frequent explanation that I hear is that there was a shortage of generating capacity in California and that this shortage was a result of poor investment incentives inherent in California’s wholesale market design. This is nonsense. There was little investment in generating capacity anywhere in the US during the time period when the California reforms were being designed and implemented (1994–1998). This is because there was excess capacity in most regions of the US during the early 1990s. Uncertainties about the future path of restructuring, regulatory and competitive reforms that began to be discussed seriously at this time was also a deterrent to potential investors waiting until the rules of the game were specified more clearly. Indeed at the time of the crisis there was a long queue of developers that had applied for permits to build new generating plants in California after the market opened in April 1998. It is unrealistic to expect that even under the best of circumstances any significant amount of new generating capacity could have come out of the construction pipeline in 2 years. Moreover, California is a summer peaking system. The biggest problems, in terms of high prices, operating reserve emergencies and rolling blackouts did not occur until the winter of 2000–2001. The problem was not that there was inadequate physical generating capacity in place, but rather that a large fraction of the existing generating capacity was not available


13

to generate electricity. The best that can be said from a resource adequacy perspective is that a lot of generating capacity that should have been available to meet demand had broken down and was not capable of producing electricity. The “shortage” of generating capacity may perhaps be explained by older plants breaking down and by their owners’ reluctance to supply when it became unclear about January 2001 whether or not they would be paid. However, there is also abundant evidence that some suppliers exploited opportunities to engage in strategic behavior to jack up market prices. At least in the summer of 2000, some generators were taking advantage of a tight supply situation to exercise market power (Borenstein et al., 2002; Joskow and Kahn, 2002). The tapes of the conversations of traders for Enron and other companies that subsequently were released make it clear that they saw and took advantage of opportunities to withhold supplies and increase market prices during the crisis. It is true that California’s wholesale market would have been stressed during the second half of 2000 even if there had been no market power problems. Demand was unusually high throughout the Western Interconnection, natural gas prices and NOx permit prices rose significantly. However, even after taking account of these factors it is hard to explain what happened during the second half of 2000 only as the result of the interplay of supply and demand in a competitive market. It is interesting to note that the Northeast experienced similar exogenous shocks to those experienced in California at this time. Little new generating capacity had come out of the construction pipeline yet and supplies were tight during the summers of 2000 and 2001. While wholesale prices rose, there was no meltdown similar to what took place in California. What made the situation in the Northeast different from California? It wasn’t a more competitive generation market structure since the Northeastern markets had more concentrated generating sectors than California. However, unlike the case in California, in the Northeast when utilities divested their generating plants and took on regulated retail default service obligations during a transition period they also entered into forward contract obligations to match their retail supply obligations and were largely (though not completely) hedged against large movements in wholesale prices. The utilities in California were not permitted to hedge fully their retail supply obligations, though contrary to the conventional wisdom, the two largest California utilities were partially hedged as a result of the nuclear, hydro, coal and existing power supply contracts they retained. In addition to providing a hedge, the “vesting contracts” in the Northeast also created incentives for the generator counterparties to these contracts to supply at least enough to meet their supply commitments and significantly mitigated any potential market power that they might have had.9 Moreover, most of the utilities in the Northeast were permitted to recover (with a lag) wholesale power costs that exceed the regulated default service prices that had been negotiated with their regulators. As wholesale market prices rose along with natural gas prices, distribution companies could book reasonable wholesale market purchase costs and recover them later through their delivery or retail default service prices. In California, the CPUC initially took the 9

These transition hedging arrangements have or soon will expire and the expected transition to competitive supply of retail customers has been quite slow. This means that distribution companies are increasingly going to the short-term wholesale market to arrange for wholesale power supplies to meet the loads of their customers who have not switched to competitive supplies. As this is written, forward wholesale market prices are significantly greater than the prices reflected in the expired(ing) transition contracts, largely as a result of large increases in natural gas prices, and retail customers relying on default service will soon see a large increase in their electricity rates. This is likely to increase the influence of those who argue that the pro-competition reforms were a mistake.

14


position that retail prices were fixed during the transition period and that no additional cost recovery was allowed. This decision first created a payment crisis by January 2001 and then led to the actual bankruptcy of PG&E and the de facto bankruptcy of SCE. In my view, if California had implemented similar transition arrangements to those implemented in the Northeast, in particular if the California utilities had more completely hedged their retail supply obligations with forward contracts and had the opportunity to recover from retail customers reasonable costs of the power they purchased in wholesale markets, there would have been no California electricity crisis. This is not to say that deficiencies in the design of California’s wholesale markets would not have led to inefficiencies that would have driven up wholesale power costs to some degree. Rather, there would not have been a sudden financial collapse and California would have had time to improve its wholesale market and transmission institutions as in the Northeast. Instead, California responded to the crisis with costly long-term contracts negotiated by the state, long-term procurement obligations, a freeze on retail competition and a strange mix of regulatory obligations and competitive markets that does not bode well for the future. 3. Departing significantly from the textbook model of restructuring, competitive market institutions and regulatory reform is likely to lead to performance problems: In the reform cases discussed above and others covered in the book, restructuring, regulatory reform and the creation of wholesale and retail market institutions proceeded more or less in parallel. There was a clear recognition that a viable retail market depended on the existence of a robust wholesale market and that a robust wholesale market depended on horizontal and vertical restructuring and regulatory reform at the T&D levels. This is not how the reforms proceeded in much of continental Europe (Spain and the Netherlands being the primary exceptions), in Japan, and in large portions of the US. The reform process stimulated by European Union (EU) directives was in my view especially strange, but perhaps politically quite astute. The initial focus of the EU reforms was on “market opening” for retail customers. That is, the focus was on retail competition. This focus ignores the fact that “market opening” alone will not lead to meaningful retail competition in the absence of appropriate wholesale market and network access and pricing institutions. Retail customers may be given the freedom to shop around for their power needs, but unless they can obtain delivery services on reasonable terms and conditions, and there is a well-functioning competitive wholesale market where they or their agents can shop, there will be no meaningful opportunity to take advantage of this freedom. Accordingly, it should not be surprising that “market opening” for retail competition alone doesn’t lead to much in the way of meaningful competition. I view the slow pace of development of competition in many of the countries in continental Europe as being largely attributable to their failure to restructure vertically and horizontally and to create the necessary network access, pricing and wholesale market institutions to create a robust wholesale market. The situation described in the chapter on Germany provides the starkest example of how retail competition without restructuring and the creation of competitive market and supporting regulatory institutions leads to performance problems. The German electric power system continues to be dominated by vertically integrated utilities with interests in generation, transmission and distribution. They control the operation of the transmission networks, which are operated as separate control and balancing areas rather than as a single balancing area as in other European countries. There is no independent system operator. Generation ownership is fairly concentrated. Until recently, there was no regulator to determine network costs and prices or to enforce unbundling rules necessary to support retail and wholesale competition, however, the anti cartel office followed the transformation of the industry and made a number of important decisions. Competition developed quite dynamically in the


15

wholesale market based on an agreement between large consumers and suppliers and supported by a power exchange. Without a clear regulatory framework it should not be surprising that meaningful competitive retail markets have been slow to emerge in Germany. The modest reform program laid out in the chapter on Japan suggests that Japan is on a similar path. Whether it is by design or accident, however, the EU’s focus on market opening for retail consumers has led it to look more closely at supporting reforms upstream at the wholesale and transmission levels as time has passed. EU countries are now creating independent system operators and transmission entities, sometimes relying on ownership separation and sometimes on functional separation or ring fencing. Germany has been forced to create a regulator to regulate (at least) network charges and unbundling protocols. While the EU and other panEuropean institutions have focused on transmission facilities that connect individual member countries, rather than getting involved in intra-country market design or competition issues, the member countries are increasingly realizing that efficient use of interconnector capacity requires some compatibility between intra-country wholesale market designs and coordination between them (ETSO and EUREPX documents). In my view, the EU started at the wrong end of the system, but given the EU’s limited legal authority perhaps this was the only way they could get a credible reform process going in light of the opposition of many incumbents to reform. The EU and members countries are now moving back upstream to implement a variety of structural and institutional reforms that would have, ideally, been done first rather than last. The chapter on Brazil provides another perspective of how reforms can go bad. The reforms in Brazil were accompanied by a reasonably complete reform blueprint. However, the blueprint was only partially implemented and the program was overwhelmed by a water shortage that would have led to problems under any circumstances. The problems were probably worse because of the incomplete implementation of the reforms and were blamed unfairly on the reforms themselves. 4. Public spot energy and ancillary services markets should be integrated with the allocation of scarce transmission capacity. The most efficient design of spot wholesale energy markets continues to be a subject of dispute among interest groups and independent experts (Hunt, 2002; Stoft, 2002; Joskow, 2005a). Should the market be built around a pool or rely on bilateral contracts? Should there be locational pricing of energy and operating reserves? How should scarce transmission capacity be allocated? Should transmission rights be physical or financial (Hogan, 1992; Joskow and Tirole, 2000)? While there is some room for flexibility, and some of the disputes reflect the self-serving arguments of interest groups that expect to benefit from inefficient markets, I believe that the experience to date supports the desirability of several basic wholesale market design features. These basic design features include the creation of voluntary public spot markets for energy and ancillary services (day-ahead and real time balancing) that accommodate bilateral contracts and self-scheduling of generation; locational pricing reflecting the marginal cost of congestion and losses at each location; the integration of spot wholesale markets for energy with the efficient allocation of scarce transmission capacity; auctioning of (physical or financial) financial transmission rights that are simultaneously feasible under alternative system conditions to hedge congestion, serve as a basis for incentives for good performance by system operators and transmission owners, and partially to support new transmission investment;10 an active demand side that can respond to spot market price 10

The allocation of transmission rights can, however, affect the incentives of firms to exercise market power and this should be taken into account in the design of rights allocation mechanisms and restrictions on the entities that can purchase these rights (Joskow and Tirole, 2000; Gilbert et al., 2002).

16


signals (Borenstein et al., 2002). These are the attributes of the PJM markets, as well as those in New England, New York and the Midwest ISO in the US (Joskow, 2006). California is proposing to implement a similar “nodal price” market design and Texas is considering it. 5. Market power is a significant potential problem in electricity markets, but the cure can be worse than the disease. Try to deal with potential market power structurally ex ante rather than ex post. The potential for market power to be a particularly severe potential problem in electricity markets was recognized many years ago (Joskow and Schmalensee, 1983, Chapter 12). It arises as a consequence of transmission constraints that limit the geographic expanse of competition, generation ownership concentration within constrained import areas, the non-storability of electricity and the very low elasticity of demand for electricity (Joskow, 1997; Borenstein, 2002). Generator market power was a serious problem for several years following the launch of the privatization, restructuring and competition program in the UK (Wolfram, 1999). Concerns about market power in the US were reinforced by the events in California in 2000–2001 (Borenstein et al., 2002; Joskow and Kahn, 2002) where market power and the exploitation of market design imperfections contributed to the explosion in wholesale prices beginning in June 2000. Market power issues of various kinds have been identified by chapters in this volume and elsewhere (New Zealand, Chile, Columbia, PJM, Texas, Alberta, Brazil and some areas of Central Europe). The problems can be attributed to the interactions between the attributes of electricity networks noted above, too few competing generating companies, wholesale market design flaws, vertical integration between transmission and generation that creates the incentive and opportunity for exclusionary behavior, excessive reliance on spot markets rather than forward contracts, and limited diffusion of real time prices and associated communications and control technology that facilitates the participant of demand in wholesale spot markets. Clearly, market power is an issue that must be taken seriously. No market design will work well if there are not an adequate number of competitive suppliers of generation service or the market power of dominant firms has not been mitigated in some way (i.e. with regulated forward contracts). Ex ante structural remedies are likely to be superior to ex post behavioral remedies, but the former are difficult to implement once generators have been privatized and deregulated. As a result, market power mitigation strategies have become an important component of wholesale market reforms in many countries. In the US, FERC market monitoring and market power mitigation protocols have been a central component of all of its reform initiatives. All of the ISOs in the US have market monitoring units, wholesale price caps have been implemented and special bidding and mitigation restrictions have been placed on generators located in small geographic load pockets. These market monitoring and mitigation protocols appear to have been reasonably successful in mitigating the ability of suppliers to exercise significant market power in these situations as well. Indeed, these measures may have been too successful, constraining prices from rising to competitive levels when demand is high, capacity is fully utilized and competitive market prices should reflect scarcity values that exceed the price caps in place. Thus, these efforts to mitigate market power in the short run may create adverse generation investment incentives in the long run (Joskow and Tirole, 2005b), a subject to which I shall return presently. 6. Good transmission and distribution network regulatory institutions are important but sometimes neglected components of the reform process. Most of the chapters in this book focus on the development of wholesale and retail markets. But it is important to remember that the textbook model includes the development and application of a well-designed regulatory framework to govern the distribution and transmission networks that will continue to be subject to government regulation of prices, costs, service quality, access rules, and investment programs.


17

These “residual” regulated segments of the electricity sector often represent a significant fraction of the total retail price for services paid for by consumers (prices for competitive plus regulated services). Moreover, the performance of the regulated segments can have important effects on the performance of the competitive segments since the regulated segments provide the infrastructure platform upon which the competitive segments rely (e.g. the electric transmission and distribution networks). Accordingly, the welfare consequences of electricity sector restructuring and competition reforms depend on the performance of both the competitive and the regulated segments of these industries. Regulatory reform focused on applying incentive or PBR regulatory mechanisms was a central feature of the liberalization program in the UK and the regulatory institutions and mechanisms that have evolved there also represent the gold standard of effective incentive or performance-based network regulation (Beesley and Littlechild, 1989; Joskow, 2005d). Privatization and the application of high-powered regulatory mechanisms has led to improvements in labor productivity and service quality in electric distribution systems in England and Wales, Argentina, Chile, Brazil, Peru, New Zealand and other countries (Newbery and Pollitt, 1997; Estache and Rodriguez-Pardina, 1998; Bacon and Besant-Jones, 2001; Domah and Pollitt, 2001; Rudnick and Zolezzi, 2001; Pollitt, 2004). Sectors experiencing physical distribution losses due to poor maintenance and antiquated equipment, as well as resulting from thefts of electric service, have generally experienced significant reductions in both types of losses. Penetration rates for the availability of electricity to the population have increased in those countries where service was not already universally available and queues for connections have been shortened. Distribution and transmission network outages have declined. Improved performance of regulated distribution (and sometimes transmission) systems has accompanied privatization and the application of high-powered PBR mechanisms almost everywhere it has been tried. The debates about alternative regulatory frameworks in Germany (as reviewed in Brauknecht and Brunekreeft’s chapter) assumed that regulators have a stark choice between applying cost-of-service regulation or incentive regulation, implying that these regulatory frameworks are substitutes. I do not think that this is a correct perspective regarding the choice of regulatory framework and associated mechanism design choices in practice. In fact, costof-service regulation and incentive regulation are complements rather than substitutes. Much of the cost information and financial analysis required to implement a cost-of-service regulatory regime is also required to implement an incentive regulation regulatory regime. Incentive regulation is an enhancement to traditional cost-of-service regulation not a replacement for it (Joskow, 2005d). Nor is pure price cap regulation optimal in theory (Schmalensee, 1989) or used in practice. Price caps with “ratchets” that reset prices every few years to reflect prevailing costs are frequently utilized. However, the reset process makes prices partially contingent on actual cost realizations over time. Price caps with ratchets are more like a sliding scale or profit sharing arrangement than the pure price caps that appear to theoretical models. Moreover, in order to reset prices to reflect costs from time to time it is necessary to have a good capital and operating cost accounting system, to measure key financial variables like the cost of equity and debt, to agree on asset valuation (rate base or regulatory asset base), and associated depreciation policies (Joskow, 2005d). This is the same kind of information that is required to implement a cost-of-service regulatory regime. That’s why cost-of-service regulation and incentive regulation are complements rather than substitutes. It is also now widely recognized that cost reduction efforts by network owners can potentially lead to a deterioration of service quality – increases in network outages, delays in service restoration, delays answering telephone inquiries. Accordingly, well-designed regulatory programs include PBR mechanisms that apply to various dimensions of service quality

18


(Joskow, 2005d). These mechanisms reward or penalize network companies based on their performance against pre-specified service quality benchmarks. Bertram’s chapter on New Zealand contains a comprehensive discussion of restructuring of the distribution sector and its regulation. It is one of the few papers that I have seen that examines issues associated with network asset-valuation decisions and how these decisions can affect the prices paid by consumers for distribution service (see also Bertram and Twaddle (2005) upon which this section of the chapter on New Zealand is based). When electricity sector restructuring takes place, one decision that must be made is how to value the assets of the distribution and transmission companies that will be used for regulatory purposes going forward. The typical approach has been to carry forward the existing depreciated book value of historical investments in transmission and distribution into the new regime so that the base level of distribution and transmission charges associated with the recovery of capital-related charges does not change as a consequence of the transition. Incremental investments are then accounted for more or less as they were under the old regime (as in the US and Canada) or economic/inflation accounting methods and approximations to economic depreciation are applied (as in the UK). In New Zealand, however, a decision was made to “write up” the value of distribution assets to reflect a specific measure of their (higher) replacement cost (called “optimized deprival value” (ODV)) and to use these higher valuations to set the base level of network prices. This valuation method led to higher prices and higher price–cost margins for distribution network owners. The argument for adopting this valuation approach was that this would allow prices to rise to their efficient level and provide consumers with appropriate price signals. The arguments against this revaluation were that (a) it would lead to significant price increases and unfairly burden consumers, (b) non-linear pricing could be used to restore the correct price incentives on the margin and (c) it created windfall profits for distribution network owners and undermines support for restructuring and competition. While Bertram’s discussion focuses on the effects of this asset-revaluation program on distribution service price levels in New Zealand, the research results reported in Bertram and Twaddle (2005) also demonstrate that operating costs incurred by distribution companies in New Zealand fell very significantly during the same period of time. These cost reductions appear to reflect both the consolidation of small distribution companies through mergers and the incentives for cost reduction provided by a high-power incentive scheme. Moran’s chapter on Australia identifies significant productivity improvements in distribution as well, without any apparent deterioration in network reliability. After the first few years following restructuring significant productivity improvements in both distribution and transmission were realized in the UK as well. Effective regulation of networks does not occur by accident. It requires good regulatory institutions. Regulatory institutions that are independent, are well staffed and have access to necessary information about costs, prices and service quality continue to be an important linchpin of successful electricity reform programs. Inadequate attention has been paid to creating good regulatory institutions in many countries. Germany and New Zealand’s initial decisions to proceed with a liberalization initiative without any sector regulator at all, relying instead on negotiated prices and the constraints of competition law, were clearly a mistake. I do not think that the issues are as complicated as may be suggested by Knieps’ chapter. The presence of an electricity network regulator with the proper goals and tools is a necessary component of a successful electricity sector reform program. 7. Creating a well-functioning transmission investment framework is important but continues to be a significant challenge in many countries. As wholesale markets have developed, congestion on the transmission network has not only increased but is increasingly recognized as a


19

significant constraint on the development of efficient competitive wholesale markets for power. In several of the countries studied in this volume, investment in transmission capacity, especially interregional transmission capacity, has not kept pace with the expansion in demand, generating capacity, or the volume of wholesale trade. In Europe and the US there has been almost no investment in interregional transmission capacity since the early 1990s. Inadequate transmission investment is identified as a problem in Brazil and in Chile as well. Texas (ERCOT) appears to have responded to intra-regional transmission congestion with new investment, but ERCOT is still effectively disconnected from the rest of North America. In addition to the effects of transmission congestion on wholesale power prices and the associated social costs of congestion, a congested transmission network makes it more challenging to achieve efficient wholesale market performance. Transmission congestion and related reliability constraints create load pockets, reducing effective competition among generators and leading policymakers to impose imperfect market power mitigation rules that create other distortions. Congestion makes it more challenging for system operators to maintain reliability using standard market mechanisms, leading them to pay specific generators significant sums to stay in the market rather than retire and to rely more on out of market (OOM) calls that depress market prices received by other suppliers. In New England, the amount of generating capacity operating subject to special reliability contracts with the ISO has increased from about 500 MW in 2002 to over 7000 MW projected for 2005 (ISO New England, 2005, p. 80), amounting to over 20% of peak demand.11 These responses to transmission congestion undermine the performance of competitive markets for energy, exacerbate the net revenue problem discussed above and lead to additional costly administrative actions to respond to market imperfections resulting from transmission congestion. In the UK and Argentina, the restructuring process included a comprehensive set of institutions and regulatory mechanisms to govern transmission operating cost and reliability, the allocation of scarce transmission capacity and approvals of transmission investment programs, as an integral aspect of the reform process. In many other countries, the regulatory framework governing transmission operation and investment was not given too much attention and allowed to evolve along with the markets. Stimulating performance improvements in the operation of transmission networks and, especially, attracting adequate investment to reduce congestion and to increase the geographic expanse of competition to reduce market power and the associated need to regulate wholesale markets to mitigate it, has been a challenge. The transmission systems that have exhibited the best performance are organized with a single independent transmission company that spans a large geographic area, integrates system dispatch, congestion management, network maintenance and investment under PBR regulation (e.g. NGC in England and Wales). Fragmented transmission ownership, separation of system operations from transmission maintenance and investment and poorly designed incentive regulation mechanisms reduce performance. Relying primarily on market-based “merchant transmission” investment, that is where new transmission investments must be fully supported by congestion rents (the difference in locational prices times the capacity of a new link) is likely to lead to inefficient investment in transmission capacity (Joskow and Tirole, 2005a). The frameworks for supporting transmission investment in many countries continue to have deficiencies. Progress is being made, however. PJM and other ISOs in the Northeastern US have adopted spot market mechanisms that integrate energy and ancillary services markets

11

FERC has ordered the ISO to replace these agreements with a locational capacity market mechanism built around an administratively determined “demand curve” for generating capacity. However, implementation has now been delayed until at least October 2006.

20


with the allocation of scarce transmission capacity. They have also refined their transmission planning and investment programs significantly to capture investment needs driven by both reliability and economic considerations, though their interdependence has been slow to be recognized. These processes accommodate merchant transmission investments but do not rely on them. Chile has also introduced transmission pricing reforms recently. The Nordic markets take a different approach to integrating day-ahead energy markets with the allocation of scarce transmission capacity, but a transmission investment framework appears to be much less well developed there. Mechanisms being developed through cooperative activities of European transmission and power exchange operators and regulators for integrating energy markets with the allocation of scarce inter-country transmission capacity are moving forward in Central Europe. And recent EU rules governing investment in interconnector capacity that expands transmission capacity between countries are very constructive. There is still a great deal of work to be done on the creation of effective institutional arrangements governing the organization of transmission operations, operating costs, congestion management, reliability and investment to expand capacity. 8. “Resource adequacy” is a concern of policymakers in almost every country. The jury is still out on whether and how competitive power markets can stimulate appropriate levels of investment in new generating capacity in the right places at the right times. Many policymakers are increasingly expressing concerns about “supply security” and “resource adequacy.” It is not always very clear precisely what these phrases refer to (see the excellent conceptual discussion of different dimensions of supply security in the chapter on the Nordic market by Von der Fehr, Amundsen and Bergman). One dimension of supply security relates to the operating reliability of the network as measured by involuntary losses of power – non-price rationing or controlled rolling blackouts – given the existing stock of capital on the network. Customers may experience blackouts due to failures on the distribution system, the transmission system, or due in inadequate generating capacity and price sensitive/interruptible demand to balance supply and demand in real time consistent with maintaining the physical integrity of the network. Failure to keep the system in balance can lead to cascading uncontrolled blackouts and network collapses affecting large regions (as occurred in the US and Italy in 2003). There is also a longer-run concept of “supply security” that reflects the adequacy of investments in distribution, transmission and generating capacity. Over time, investment in additional capacity should be made as long as the incremental value of the investments exceeds the incremental cost of the investment. If too little investment is made, costs and prices, including the costs associated with non-price rationing of demand and network collapses as discussed above, will be too high. Thus, long-run concepts of supply security or resource adequacy are related to short-run concepts of supply security or network reliability. I have already discussed network investment issues and now will turn to issues associated with investment in new generating capacity. Creating appropriate investment incentives for new generating capacity is perceived to be a growing problem in many countries. At first blush, this concern may be surprising since the early experience with reforms during the 1990s suggested that competitive wholesale markets could and would mobilize adequate (or more than adequate) investment in new generating capacity. Substantial amounts of capital were mobilized during the late 1990s to support construction of new efficient generating capacity in many countries that have implemented reforms. In the US, over 200,000 MW of new generating capacity went into service between 1999 and 2004, most of it merchant capacity, an increase of nearly 30% in total US generating capacity (Joskow, 2006). About 40% of the stock of generating plants in service in England and Wales was replaced with modern efficient CCGT technology between 1990


21

and 2002 as old coal-burning generators have been closed and expensive dirty coal plants have been displaced by cheaper and cleaner CCGT capacity. Many other countries implementing reforms during the 1990s, including Argentina, Chile and Australia, also attracted significant investment in new generating capacity (Jamasb, 2002) after the reforms were initiated. So, why are policymakers so concerned now? First, we should recognize that liberalization has evolved in much of Europe during a period when there was significant excess generating capacity, Spain and Italy being the major exceptions. Even in the UK, the quantity of generating capacity in service today is not much greater than it was in 1990, with most of the investment in generating capacity during the 1990s being stimulated by opportunities to replace the inefficient stock of old generators that the state-owned Central Electricity Generating Board (CEGB) kept in service to maximize consumption of expensive British coal, long-term contracts entered into by retail suppliers early in the UK’s liberalization program, and the high prices available in the wholesale market, influenced by the exercise of market power as already discussed. These investments were not the result of a significant need for new generating capacity to meet rapidly growing peak demand, as is the case in several of the countries discussed in this volume. Second, the environment for financing new generating investments has changed dramatically in the last few years as a result of financial problems faced by merchant trading and generating companies in Europe, the US and Latin America, as well as macroeconomic and political instability in Latin America and Asia (De Araujo, 2001; Jamasb, 2002; Joskow, 2005a). After peaking at 55,000 MW of new capacity entering service in the US 2002, only about 15 000 MW of new generating capacity entered service in 2005, most of which was built either for municipal utilities that have not been subject to restructuring and competition reforms or wind projects that benefit from special subsidies and contractual arrangements. Concerns about future incentives for investment in additional generating capacity are noted in several of the studies in this book (Chile, Brazil, Columbia, New Zealand, California, Germany, PJM). In some cases, state-owned entities have stepped in to contract for additional generating capacity (e.g. Chile, Brazil, New Zealand, Ontario, California) to mitigate resource adequacy concerns. The actions by state-owned entities to support investment in new generating capacity may have salutary short-run effects, but these actions are likely to discourage private investment in the longer run. Programs designed to stimulate investments in renewable generation (mostly wind) with special tax subsidies, contractual benefits, or mandatory purchase obligations, further complicate the investment picture for “ordinary” generating plants. Potential private investors in new generating capacity are looking for stable market rules and longer-term contractual commitments before they will commit capital for new generating facilities. Continuous market redesign, regulatory actions that limit prices, system operators’ “reliability” actions that depress market prices, and other market and regulatory imperfections are being pointed to as deterrents to private investors in unregulated generating plants. Financing investments in peaking capacity, which rely heavily on wholesale market prices creating “rents” to support fixed investment costs in a relatively small number of hours, is especially problematic. Analyses done of regional markets in the US make it fairly clear that “energy-only” markets do not produce adequate revenues to attract investment in generating capacity consistent with the reliability standards that are still applicable to them and have now become mandatory (Joskow, 2005a, 2006). The chapters on the UK, Norway and Australia are more optimistic about generation investment incentives. I am unconvinced, though I hope that they are right. The UK is in a rather unique position since the investment wave of the 1990s has left it with a relatively large stock of older mothballed-generating units that can enter and leave the market on relatively

22


short notice as forward (e.g. 1 year) contract prices change, providing a fairly elastic shortrun supply of “reserve” generating capacity. A stock of mothballed reserve capacity of this magnitude does not exist in most countries and investment in new capacity is forecast to be needed to meet growing demand with conventional levels of reliability. Von der Fehr, Amundsen and Bergman’s paper on Nordic markets is also optimistic, but I am not convinced by the evidence presented to support this optimism. The fact that spot market and forward prices have not justified new investment in the last several years begs the question of whether or not these prices are providing the right price signals. This is an issue that has been analyzed for all of the organized US markets and the answer is the same – wholesale energy market prices do not provide adequate price signals (Joskow, 2006). There has been little if any net increase in generating capacity in the Nordic market area for many years and reserve margins are declining. There are a few generating projects in process, including a nuclear plant in Finland developed by a public/private consortium. Maybe the stateowned generating companies will step in and fill any gaps, but whether this will be a market response or a political response is yet to be determined. A number of US states are considering imposing resource adequacy or forward contracting obligations that would be placed on the entities that provide retail service to overcome imperfections in wholesale spot markets to restore incentives for investments in generating capacity and demand-response capabilities consistent with traditional reliability levels (California Public Utilities Commission, 2005; Joskow, 2006). The organized markets in the US, Chile, Argentina and Columbia have such obligations or administratively determined capacity payments. These policies are and will continue to attract considerable attention and debate as they should. 9. Environmental policy initiatives can and should be designed to be compatible with competitive wholesale and retail electricity markets. The electric power sector has been a focus of environmental policy for decades. As noted earlier, this has been the case both because of the sectors’ significant impacts on the environment and the political attractiveness of using a regulated industry as a facilitator of environmental policies through “taxation by regulation.” The liberalization of electricity sectors does not appear to have reduced policymakers’ zeal for finding ways to exploit opportunities to exploit taxation by regulation opportunities to pursue environmental agendas. Going forward, environmental policy initiatives affecting the electric power sector need to be compatible with competitive market mechanisms or they will undermine the efficient evolution of competitive markets. Cap and trade programs, as are described in the chapters focused on European countries, for controlling CO2 emissions and as applied in the US for controlling SO2 and NOx emissions (Ellerman et al., 2003) are likely to be compatible with competitive electricity markets if they have the right design features. In particular, as long as emissions permits are fully tradeable, allocations of permits are not updated to reflect changes in emissions, and allowances or credits can be sold when generating plants are retired rather than being recaptured from the plant’s owners by the government, cap and trade programs provide an approach to controlling emissions that is compatible with the development of efficient competitive markets for power. The reliance on tax subsidies, special procurement obligations and contract regimes, subsidized transmission service, and exemptions from balancing and settlement obligations in wholesale markets, all designed to promote renewable energy technologies is of more concern. If we can internalize the relevant externalities with a cap and trade program (or taxes on emissions), the only argument for additional renewable energy subsidies are significant market distortions caused by non-convexities associated with learning by doing economies. However, learning by doing economies are ubiquitous and affect many valuable goods and


23

services. Taking this argument to the extreme would imply that the government should subsidize all new products. Moreover, the benefits of subsidies once relevant externalities are internalized should be balanced against the costs of the market distortions that are likely to be caused by the special regime policies governing renewable energy. To the extent that obligations to purchase or pay for above market costs of renewable energy technologies are imposed on retail electricity suppliers, it is important that the obligations apply symmetrically to all electricity suppliers in a symmetrical and non-discriminatory fashion. Otherwise these policies will create an industry devoted to competing for customers by bypassing the costs of renewable energy obligations. 10. Retail market design and the terms and conditions of default service provided by incumbents have important implications for the success of retail competition programs. Consumers can benefit in at least four ways from the introduction of retail competition. First, even if they do not switch to a competitive retailer or competitive electricity service provider (ESP), they may benefit from reductions in regulated “default service”prices that have typically accompanied the restructuring process as an outcome of the bargaining over stranded cost recovery and the terms and conditions under which the incumbents can move their regulated generating plants into unregulated affiliates. Second, consumers can benefit by receiving lower prices than the default service price from an ESP that has competed successfully for their business. Third, ESPs may offer consumers a variety of value-added services, including price risk management, demand management and energy efficiency services. Finally, competing ESPs may be able to provide “retailing” services more efficiently that can the incumbent. However, here we must recognize that retail service costs are a small fraction of a typical customer’s bill, amounting to 0.3–0.4 cents/kWh or about $3–$7 per month for a typical residential customer (depending on assumptions about fixed versus variable components of retail service costs – Joskow, 2000a). Since the incumbent monopolies did not have to incur marketing and advertising costs to attract customers, these are additional costs that are not now reflected in regulated retail prices but would have to be incurred by ESPs. The design of retail competition programs vary widely from country to country and even within countries where reforms have been driven by states and provinces. All countries that have adopted market liberalization reforms allow large customers to buy power competitively at the outset of their restructuring programs. In some countries, retail competition remains available only to such large customers. Residential and small commercial and industrial customers then continue to buy power from their local distribution companies which in turn procure their power in competitive markets and pass along the associated costs in the prices charged to these groups of retail consumers. Other countries have gradually expanded retail competition opportunities to customer classes that consume smaller amounts of power, with the long-run goal of opening up the retail market to all customers. In this case, the distribution company (or a retail affiliate) buys power in the wholesale market and passes along the associated costs to the remaining retail customers during a transition period. Finally, retail competition is sometimes (e.g. in the states in the US that have adopted retail competition programs) made available to all customers at the outset of the reform program. However, since customers, especially smaller customers, do not switch instantly to competitive suppliers, some type of “default service” must be provided to them, typically by their local distribution company or a retail affiliate. Thus, in all cases, there is some period of time during which a significant fraction of retail consumers continue to be served under some type of regulated default service tariff. There are three primary sets of issues associated with the introduction of retail competition that are discussed in the chapters in this volume. First, there are issues associated with the

24


details of the unbundling and network pricing model applied. Second, there are issues associated with the terms and conditions of retail default power supply services. Third, questions continue to be raised about whether the benefits of making retail competition available to smaller customers is worth the costs relative to the best available wholesale competition alternative. These are all important issues. Unbundling prices and costs for network services from those for competitive electricity supply services, regulating the prices charged for network services and making network services available to all users (customers and retail suppliers) of the network in a nondiscriminatory fashion, are basic and essential steps for retail electricity competition to be successful. Otherwise, the incumbent vertically integrated utility will have an incentive to charge high prices for network services, where it will make a large profit, and low belowmarket prices for power supply services, where it may lose money. This strategy will make it impossible for competitive retail suppliers to compete with the incumbent. This is precisely one of the barriers to retail competition that emerged in Germany in the absence of an electricity sector regulator to establish network access and pricing protocols. To the extent that the incumbent distribution company is also an active unregulated competitor in the retail market, as opposed to a passive supplier of regulated default service that carries no profit margin, competitive retail activities should be fully separated or “ring fenced” from the firm’s regulated activities. The distribution company’s retail affiliate would then look to recover all of its costs, including customer service costs, from revenues received in the market. Functional or ownership separation of retail supply mitigates concerns about cross-subsidization and discriminatory access to information about customers and to the network. This is the approach taken in the UK, Texas and eventually in New Zealand. However, when a distribution company has an obligation to provide “passive” default service, metering, billing and customer services at regulated prices, regulators have an obligation to provide for full recovery of efficiently incurred costs. The treatment of customer service and billing costs is especially challenging since there are significant scale economies in providing these services, making the costs avoided by the incumbent as customers migrate to competitive retailers lower than the average costs of providing these services. As discussed in the chapter by Tschamler on retail competition in the US, the terms and conditions of retail default service can have significant effects on the ability of competitive retailers to attract customers. In the US and some other countries (e.g. Spain), default service prices or tariffs have been used to support a number of objectives other than promoting a robust retail market. These include commitments that retail customers will receive an immediate and sustained price reduction of some magnitude, stranded cost recovery considerations, income redistribution goals and consumer protection goals. As a result, default service prices have sometimes been set at levels below the wholesale cost of power, or wholesale prices have risen over time, closing or reversing the gap between default prices and wholesale market prices. Under these circumstances it is impossible for a competitive retailer profitably to offer services that can attract customers away from default service. Regulatory rules that allow customers to come and go between regulated default service prices and competitive market prices, whichever happens to be lower at a point in time, further undermines the ability of competitive retailers to build a stable customer base. If as a matter of policy regulators want to protect customers from high market prices by giving them access to regulated tariffs fixed at prices below market then retail competition will never be successful. Such policies may also signal a lack of faith and commitment by policymakers in retail competition. The experience in Pennsylvania (a state that is part of the PJM wholesale market) provides a good example of the effects of mixing regulated default pricing with retail competition. Different default service prices were set for each utility in Pennsylvania, reflecting


25

Percentage of residential load

historical regulated costs of generation service and stranded cost recovery settlements. The prices were fixed in 2000 for a term of up to 10 years, with some adjustments for fuel and other input price changes. Figure 1 provides time series data on the fraction of residential customers which switched to a competitive retailer for each utility in Pennsylvania. Figure 2 provides the same data for industrial customers. There is both wide variation in the initial fraction of customers who switched to competitive retail suppliers and significant evidence of their switching back and forth between regulated default service and regulated services. The inter-utility variations must be attributable to differences in regulated default service prices since there is no inherent reason why customers in Pittsburgh should be more likely to shop for alternatives than are customers in Philadelphia. By July 2005 nearly all residential customers had returned to regulated default service and a large fraction of the industrial customers who initially opted for default service had also returned to default service. This is attributable to rising nominal wholesale prices in PJM which have reduced or eliminated the “headroom” between the regulated default service price and the wholesale market price for

Pennsylvania direct access load: residential (%)

40

April 2000 July 2000 January 2001 April 2001 October 2001 January 2002 April 2002 July 2002 October 2002 January 2003 January 2004 April 2004 July 2004 July 2005

35 30 25 20 15 10 5 0 Allegheny power

Duquesne light

GPU energy

PECO energy

Penn power

PPL

Percentage of industrial load

Fig. 1. Pennsylvania direct access load: residential (%). Source: Pennsylvania office of consumer Advocate.

Pennsylvania direct access load: industrial (%)

80

April 2000 July 2000 January 2001 April 2001 October 2001 January 2002 April 2002 July 2002 October 2002 January 2003 January 2004 April 2004 July 2004 July 2005

70 60 50 40 30 20 10 0

Allegheny power

Duquesne light

GPU energy

PECO energy

Penn power

PPL

Fig. 2. Pennsylvania direct access load: industrial (%). Source: Pennsylvania office of consumer Advocate.

26


power. Accordingly, in the US, the biggest problem faced by competitive retailers is “competition” from default service, a service for which the incumbents typically make no profit either. The general pattern of retail switching behavior in most countries is that large industrial customers are more likely to switch and to do so more quickly than smaller industrial and commercial customers. Residential customers switch more slowly and are more likely to remain with the incumbent, especially when the incumbent must offer a regulated default price that is at or below the wholesale market price of power. However, for residential and small commercial customers, even if the regulated default service price is equal to the comparable competitive wholesale market value of the power supplied, retail suppliers need a significant additional margin both to induce sticky retail customers to switch suppliers and to cover their retail supply costs. This margin has turned out to be much larger than anticipated when retail competition was first introduced. In particular, the retail supply costs for the mass market (residential and small commercial) are much higher than many retailers had anticipated. Billing, customer service, bad debt, advertising and promotion costs add up quickly. Accordingly, the default service price may have to be much higher than the comparable wholesale market price to induce more customer switching. Moreover, the evidence from England and Wales and Texas suggests that price reductions of 5–10% of the total bill compared to the default/incumbent service price are necessary to get significant customer switching for mass market (residential and commercial) customers. If the generation component of the retail price is 50% of the total bill, then price reductions of 10–20% on the generation component are necessary to get significant switching. To this must be added about another 5–10% for retail service costs. So, a margin of 15–30% between the default service price and the comparable wholesale market value of power may be necessary to induce significant switching by residential and small commercial customers. A margin of this magnitude may be incompatible with reducing retail prices from their prevailing or “but for” regulated levels. This naturally leads to the final issue. Is retail competition worth the trouble compared to a regime where the distribution company procures power competitively and resells it at cost to residential and small commercial customers? Unfortunately, there is little if any good empirical analysis available to evaluate this question rigorously, though there is no shortage of strong ideological views. Looking at switching rates alone is not very informative as an index of the welfare consequences of retail competition. The presumption has been that retail competition is a good thing to offer larger customers, where transactions costs are low, opportunities to offer risk management and demand-management products are greater, and customers are expected to be able to shop intelligently. There are also benefits for the development of competitive wholesale market resulting from having more buyers active on the demand side, reducing monopsony problems that might emerge if distributors were the only buyers. Moreover, if the alternative is competitive procurement by the distribution company, regulators must become involved in determining procurement rules, including the attributes of the contracts that will be put out for bids. Industrial customers and their agents should be in a better position to express their risk preferences than are regulators (see Littlechild (2003) for these and other arguments in support of retail competition). And indeed, where default prices have been allowed to float to reflect spot wholesale market prices (including capacity prices), large customers appear to migrate fairly quickly to the market and to sign contracts that hedge price volatility from 1 to 3 years into the future. It is far from obvious to me, however, that residential and small commercial customers have or will benefit much, if at all, from retail competition compared to a regime where their local distribution company purchased power for their needs by putting together a portfolio of short-term forward contracts (from days to several years) acquired in wholesale markets


27

(Joskow, (2000a, b) and Littlechild (2003) for a different view)). Indeed, New Jersey has used the so-called basic generation service (BGS) auction process quite effectively to buy power competitively for residential and small commercial customers. There is little evidence that residential and small commercial customers are getting any significant value-added services from retail suppliers aside from some billing options in the UK and, at least in Norway, choices between contracts of different durations. Retail competition with load profiling leads to some inefficiencies (Joskow and Tirole, 2005c). There is evidence that there are significant costs associated with implementing a retail competition program for residential consumers (Green and McDaniel, 1998) and that they may make poor shopping decisions (Salies and Waddams Price, 2004). I remain unconvinced that retail competition for small customers is worth the bother. If policymakers are committed to fostering retail competition for residential and small commercial customers, despite the possibility that retail prices will rise in the short run due to increased transactions costs, switching costs and market power, the framework adopted by the UK, Texas and the Nordic countries is likely to be the most successful in stimulating retail shopping and the development of a viable retail supply sector. 11. Vertical integration between retail supply and generation is likely to be an efficient response to imperfections in wholesale markets. It may also create market power problems. Thus, policymakers must confront a tradeoff: In several countries with active retail competition programs there appears to be a growing movement to an industry structure where competitive retail suppliers acquire generating capacity to meet a significant fraction of their retail commitments. This trend is likely to reflect an efficient response to relatively high transaction costs associated with real wholesale power markets in practice (Coase, 1937; Williamson, 1975; Carlton, 1979). There is no inherent competition problem with vertical integration of this type as long as there are a sufficient number of vertically integrated suppliers that continue to compete in the market. However, if there is significant market power in the upstream or downstream markets, vertical integration could lead to a further reduction in competition by increasing the operating or entry costs of rival retail suppliers (Ordover et al., 1990; Riordan, 1998). The discussion of retail competition in the chapter on New Zealand suggests that the intensity of competition declined significantly as retail suppliers became vertically integrated while the chapter on Australia suggests that vertical integration did not lead to market power problems there. Thus, there may be a tradeoff between increases in efficiency and increases in market power. The welfare properties of retail competition with different horizontal and vertical market structures has received too little serious theoretical and empirical analysis and more work on these issues would be desirable. 12. Expanding demand response in spot wholesale energy markets needs more attention. In markets for most goods and services, when demand grows and supply capacity constraints are reached, prices rise to ration demand to match the capacity available to provide supplies to the market. In electricity markets, however, as generating capacity constraints are reached, relatively little demand can be rationing by short-term price movements and, instead, must by rationed with rolling blackouts. This reflects both the limited use of real time pricing and the system operator’s need to adjust demand very quickly at specific locations. The possibility of broader uncontrolled cascading blackouts and regional network collapses further exacerbates this problem and necessarily leads to regulatory requirements specifying operating reserves, operating reserve deficiency criteria and associated administrative actions by system operators to balance the system to meet voltage, stability and frequency requirements in an effort to avoid cascading blackouts (Joskow and Tirole, 2005b). In addition, retail competition has

28


more attractive welfare properties if the real time consumption of retail consumers can be measured instead of relying on load profiling (Joskow and Tirole, 2005c). The challenges faced by network operators to maintain system reliability and avoid nonprice rationing of demand would be reduced if additional demand-side response instruments were at their disposal. These instruments include the ability to rely on demand response by more customers who can see and respond to rapid changes in market prices and expanded use of price-contingent priority rationing contracts (Chao and Wilson, 1987). As a general matter, too little demand-side response has been developed to date most countries. (Texas has quite a bit of demand response but this is likely to be associated with the relatively large number of oil refineries, petrochemical and other energy intensive industries that can manage their load and have self-generation facilities that can be used when market price rise sufficiently.) The demand-response instruments that are available are poorly integrated with spot markets and are likely to have the effect of depressing prices inefficiently. Moreover, the prices that are paid for demand response or the prices that can be avoided by responding to price signals are too low compared to the cost of carrying generating capacity reserves to meet planning reserve margins in some cases. Improving demand response should be given higher priority in wholesale market design (Borenstein, Jaske and Rosenfeld, 2002). There is no reason why these efforts must be limited to large customers. Controlled air conditioning and water heater devices have been available to residential and small commercial customers in several countries for many years. Italy’s decision to install real time meters for every electricity consumer premises is an important step forward in this area. By making real time metering universal, Italy will gain the benefits of an infrastructure that can support additional efficient demand response and avoid the impediments to retail competition identified in the chapter on the Nordic markets. 13. Electricity sector reform appears to be a continuing process of improvement, but a process of continuing reforms of the reforms has both potential benefits and potential costs. It is quite clear from the chapters in this book that none of the reform programs got it all right out of the box. Initial reform programs are followed by additional reforms, some major and some minor, to respond to performance problems that emerge in practice or lessons learned about best practices from other countries. On the one hand, reforms that are needed to fix major performance problems certainly should be considered carefully. On the other hand, a process of ongoing reforms that have significant and uncertain future financial impacts on market participants is not likely to create a framework that is conducive to investment in long-lived assets whose value is subject to policy reform risks. Policy reforms may also be used opportunistically to respond to political pressures that arise under market conditions when investors properly expected that they would achieve high returns, effectively truncating the upper end of the return distribution and leading investors to require higher expected returns from other states of nature than would otherwise be the case. So, for example, it would be unfortunate if policymakers were to require renegotiation of restructuring arrangements because natural gas prices have risen dramatically, benefiting deregulated coal, nuclear and hydroelectric facilities. Similarly, if policymakers belatedly conclude that there are market power problems, the tools available to them are likely to be much more limited after assets have been privatized, deregulated, traded, constructed, etc. than before, if they are to limit themselves to policies that do not amount to expropriation of private property rights that they previously conveyed. Policy reform risk is extremely difficult to hedge, except perhaps through the regulatory process itself and is a potentially significant deterrent to investment. Accordingly, it makes a lot of sense to try to get the reform program as correct as possible the first time around. It is also important for policymakers to recognize that the search for


29

perfection can be the enemy of the good. Policymakers need to make sure that the benefits of any additional reforms exceed their short- and long-run costs, in particular those related to investment incentives. And if there are to be reforms of the reforms it is desirable to package them together so that there can be one reform of the reforms rather than a continuing stream of them. Finally, if policymakers are serious about competitive markets for power they will have to rethink the long tradition of relying on taxation by regulation of the electric power industry to implement policies in ways that hide the associated costs from taxpayers. 14. A strong political commitment to reform is important. Implementing a good electricity sector liberalization program is a technical, institutional and political challenge. Almost everywhere, some unanticipated (at least by the policymakers) problems emerged that required major or minor refinements to the original reform program. In some cases (e.g. UK, New Zealand, Alberta, Australia, Texas) the reforms were consistent with the continuing development of competitive markets and in other cases they were not (e.g. California, Ontario, Brazil). It appears that reforms that have strong pro-competition political support are more likely to respond to problems by identifying market or institutional imperfections and trying to fix them in ways that are consistent with the continued successful evolution of competitive wholesale and retail markets. They are also likely to be willing to live with some imperfections, recognizing that no market is perfect and that the cures can be worse than the disease. Where the commitment to competitive electricity markets is weak, when problems emerge policymakers are more likely to seek what appear to be quick fixes that undermine continued evolution of competitive markets or just cut and run from the competitive market agenda. If the commitment to competition is not strong in the first place, of course, the reforms are likely to be timid and have little effect on the status quo anyway, Japan and many US states being the prime examples. If policymakers do not have a strong commitment to competition and are unable to follow the basic textbook reform model outlined at the beginning of this chapter, sector performance may be better if they focus their attention on improving the traditional regulated monopoly model rather than dabbling with timid and flawed approaches to competition.

Conclusion Structural, regulatory and market reforms have been applied to electricity sectors in many countries around the world. Significant performance improvements have been observed in some of these countries as a result of these reforms, especially in countries where the performance of state-owned monopolies was especially poor. Privatization and PBR mechanism applied to regulated distribution companies has generally yielded significant cost reductions without reducing service quality. Wholesale markets have also stimulated improved performance from existing generators and helped to mobilize significant investments in new generating capacity in several countries. However, efforts to create well-functioning competitive wholesale and retail markets have revealed many significant challenges and the restructuring and competition reforms remain a work in progress in most countries. The California electricity crisis, electricity crises in Brazil, Chile, Ontario, and elsewhere, scandals involving energy trading companies like Enron, the failure of poorly designed reforms in countries such as Brazil, macroeconomic problems undermining investments in generally well-designed systems as in Argentina, and ongoing political interference undermining private sector investments as in India and Pakistan, have certainly made policymakers more cautious (but not necessarily more thoughtful) about

30


electricity sector reforms. The challenges associated with successful reforms have sometimes been underestimated. However, these problems and challenges do not imply that restructuring, regulatory reform, and promoting the development of competitive wholesale and retail markets for power, are ill-advised. The problems that have emerged are now much better understood and solutions to many of them are at hand. The primary question is whether governments properly can choose between competing solutions and have the political will to resist interest group pressures and pursue reforms that will lead to more efficient markets and better performance of the network platforms upon which competition depends.

References Bacon, R.W. and Besant-Jones, J.E. (2001). Global electric power reform, privatization and liberalization of the electric power industry in developing countries. Annual Reviews of Energy and the Environment, 26, 331–359. Beesley, M. and Littlechild, S. (1989). The regulation of privatized monopolies in the United Kingdom. Rand Journal of Economics, 20(3), 454–472. Bertram, G. and Twaddle, D. (2005). Price–cost margins and profit rates in New Zealand electricity distribution networks: the cost of light handed regulation. Journal of Regulatory Economics, 27(3), 281–307. Besant-Jones, J.E. (ed.) (1993). Reforming the Policies for Electric Power in Developing Countries. Washington, DC: World Bank. Borenstein, S., Bushnell, J. and Wolak, F. (2002). Measuring market inefficiencies in California’s restructured wholesale electricity market. American Economic Review, 92(5), 1376–1405. Borenstein, S., Jaske, M. and Rosenfeld, A. (2002). “Dynamic Pricing, Advanced Metering and Demand Response in Electricity Markets,” Center for the Study of Energy Markets Working Paper, University of California at Berkeley, October. California Public Utilities Commission (2005). “Capacity Markets White Paper,” August 25. Carlton, D. (1979). Vertical integration in competitive markets under uncertainty. Journal of Industrial Economics, 27, 189–209. Chao, H. and Wilson, R. (1987). Priority service: pricing, investment and market organization. American Economic Review, (77), 899–916. Coase, R. (1937). The nature of the firm. Economica, 4, 386–405. De Araujo, J.L.R.H. (2001). Investment in the Brazilian ESI – What Went Wrong? What Should Be Done? Institute of Economics, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil. Domah, P.D. and Pollitt, M.G. (2001). The restructuring and privatisation of the regional electricity companies in England and Wales: a social cost benefit analysis. Fiscal Studies, 22(1), 107–146. Ellerman, A.D., Joskow, P.L. and Harrison, D. (2003). Emissions Trading in the United States. Pew Center on Global Climate Change, Washington, DC. Estache, A. and Rodriguez-Pardina, M. (1998). “Light and Lightening at the End of the Public Tunnel: The Reform of the Electricity Sector in the Southern Cone,” World Bank Working Paper, May. Gilbert, R., Neuhoff, K. and Newbery, D. (2002). “Allocating Transmission to Mitigate Market Power in Electricity Networks, Cambridge-MIT Electricity Project Working Paper, October. Green, R. and McDaniel, T. (1998). Competition in electricity supply: will “1998” be worth it? Fiscal Studies, Vol. 19, No. 3, pp. 273–293. Green R. and Newbery, D. (1992). Competition in the British Electricity Spot Market. Journal of Political Economy, 100(5), 929–953. Hogan, W. (1992). Contract networks for electric power transmission. Journal of Regulatory Economics, 4, 211–242. Hunt, S. (2002). Making Competition Work in Electricity. New York, Wiley. ISO New England (2005). “2004 Annual Markets Report.” http://www.iso-ne.com/markets/ mkt_anlys_rpts/annl_mkt_rpts/2004/2004_annual_markets_ report.doc Jamasb, T. (2002). Reform and Regulation of the electricity sectors in developing countries, June 2002, Department of Applied Economics, University of Cambridge.


31

Joskow, P.L. (1989). Regulatory failure, regulatory reform and structural change in the electric power industry. Brookings Papers on Economic Activity: Microeconomics, 125–199. Joskow, P.L. (1997). Restructuring, competition and regulatory reform in the US electricity sector. Journal of Economic Perspectives, 11(3), 119–138. Joskow, P.L. (1998). Electricity sectors in transition. Energy Journal, 19, 25–62. Joskow, P.L. (1998a). Electricity sectors in transition. Energy Journal, 19, 25–62. Joskow, P.L. (2000a). Deregulation and regulatory reform in the US electric power sector. In Sam Peltzman and Clifford Winston (eds.), Deregulation of Network Industries: What’s Next? Brookings Institution Press, Washington, DC. Joskow, P.J. (2000b). Regulating the electricity sector in Latin America, comments 2000b. Economia, 1, 199–215. Joskow, P.L. (2001). California’s electricity crisis. Oxford Review of Economic Policy, 17(3), 365–388. Joskow, P.L. (2003). Electricity sector restructuring and competition: lessons learned. Cuadernosde Economia (Latin American Journal of Economics), 40, 548–558. Joskow, P.L. (2005a). The difficult transition to competitive electricity markets in the United States. In J. Griffin and S. Puller (eds.), Electricity Deregulation: Where To From Here? University of Chicago Press, Chicago. Joskow, P.L. (2005b). Transmission policy in the United States. Utilities Policy, 13, 95–115. Joskow, P.L. (2005c). Regulation and deregulation after 25 years: lessons for research. Review of Industrial Organization, 26, 169–193. Joskow, P.L. (2005d). “Incentive Regulation in Theory and Practice: Electric Transmission and Distribution,” NBER Regulation Project, August. Joskow, P.L. (2006). Markets for power in the US: an interim assessment. The Energy Journal, 27(1), 1–36. Joskow, P.L. and Kahn, E. (2002). A quantitative analysis of pricing behavior in California’s wholesale electricity market during summer 2000. The Energy Journal, 23(4), 1–35. Joskow, P.L. and Rose, N.L. (1989). The effects of economic regulation. In R. Schmalensee and R. Willig (eds.), Handbook of Industrial Organization. North-Holland, Amsterdam. Joskow, P.L. and Schmalensee, R. (1983). Markets for Power: An Analysis of Electric Utility Deregulation. MIT Press, Cambridge. Joskow, P.L. and Tirole, J. (2000). Transmission rights and market power on electric power networks. Rand Journal of Economics, 31(3), 450–487. Joskow, P.L. and Tirole, J. (2005a). Merchant transmission investment. Journal of Industrial Economics, 53(2), 233–264. Joskow, P.L. and Tirole, J. (2005b). Reliability and competitive electricity markets, September (revised) http://econ-www.mit.edu/faculty/download_pdf.php?id⫽917. Joskow, P.L. and Tirole, J. (2005c). Retail electricity competition, Rand Journal of Economics (forthcoming) http://econ-www.mit.edu/faculty/download_pdf.php?id⫽918. Littlechild, S.C. (2003). Wholesale spot market passthrough. Journal of Regulatory Economics, 23(1), 61–91. Newbery, D. and Pollitt, M. (1997). The restructuring and privatization of Britain’s CEGB – was it worth it? Journal of Industrial Economics, 45(3), 269–303. Ordover, J., Salop, S. and Saloner, G. (1990). Equilibrium vertical foreclosure. American Economic Review, 80, 127–142. Pollitt, M. (2004). “Electricity reforms in Argentina: lessons for developing countries. CMI Working Paper 52, Cambridge Working Papers in Economics. http://www.econ.cam.ac.uk/electricity/publications/ wp/ep52.pdf Riordan, M. (1998). Anticompetitive vertical integration by a dominant firm. American Economic Review, 88, 1232–1248. Rudnick, H. (1996). Pioneering electricity reform in South America. IEEE Spectrum, August, 38–45. Rudnick, H. (1998). Market restructuring in South America. IEEE Power Engineering Review, June, 3–6. Rudnick, H. and Zolezzi, J. (2001). Electric sector deregulation and restructuring in Latin America: lessons to be learnt and possible ways forward. IEEE Proceedings Generation, Transmission and Distribution 148, 180–184.

32


Salies, E. and Waddams Price, C. (2004). Charges, costs, and market power: the deregulated UK electricity retail market. The Energy Journal, 25(3), 19–37. Schmalensee, R. (1989). Good regulatory regimes. Rand Journal of Economics, 20(3), 417–436. Stoft, S. (2002). Power System Economics. Wiley Interscience, New York. Williamson, O. (1975). Markets and Hierarchies: Analysis and Antitrust Implications. Free Press, New York. Williamson, O. (1985). The Economic Institutions of Capitalism. Free Press, New York. Wolfram, C. (1999). Measuring duopoly power in the British electricity spot market. American Economic Review, 89(4), 805–826. World Bank (1994). Infrastructure for Development: World Development Report 1994.

PART I What’s Wrong with the Status Quo?


Chapter 1 Why Restructure Electricity Markets? FEREIDOON P. SIOSHANSI1 AND WOLFGANG PFAFFENBERGER2 1

Menlo Energy Economics, Walnut Creek, CA, USA; 2International University Bremen, Bremen, Germany

Summary For nearly two decades, governments in a number of countries have restructured and/or liberalized their electricity supply industry. The motivations to change the regulatory regime and the expectations for the outcomes vary despite many commonalities. The experience of different countries, however, has been decidedly mixed, highly to mildly successful in some places, disappointing or disastrous in others. Understanding why some restructured markets are functioning successfully while others have not is the central theme of this book, which offers a variety of perspectives from different parts of the world. This chapter provides a review of the shortcomings of the traditional regulated monopoly structure, examines the motivations for restructuring and the expectations of policy-makers in introducing market reforms.

1.1. Why Change the Status Quo? As the following chapters of this book point out, before the restructuring developments of 1980s and 1990s, the electricity supply industry (ESI) was generally organized and operated under one or a combination of the following two structures: ● ●

as state-owned enterprises (SOEs), or as privately owned, regulated monopolies.

In the former case, typically there was no independent regulator but a separate part of the government (e.g. the treasury) might exercise regulatory oversight or provide financial or fiduciary function. This model has some shortcomings. The taxpayers, for example, usually bear most investment risks and there may be poor accountability since the government agency is not directly accountable to consumers or shareholders. Moreover, under this scenario there tends to be some circularity, since different arms of government would be engaged in forecasting, planning, building, investing and operating as well as managing the network, setting and collecting retail tariffs. SOEs may not be overly sensitive to customer needs and may 35

36


lack sufficient incentives to improve customer service or engage in technology innovation. In the case of many rapidly growing economies, central government may not have sufficient resources to adequately invest in infrastructure, resulting in chronic power shortages and poor service reliability. In the case of privately owned, regulated monopolies, which are prevalent in more advanced economies, the private sector owns and operates some or significant components of the ESI under the supervision of an independent regulator. Regulated monopolies could potentially offer some advantages relative to the centrally planned SOEs especially when there is a competent, resourceful, independent and vigilant regulator. At least in theory, this model could capture the significant economies of scale of large vertically integrated monopolies while controlling their abusive tendencies. Regulated monopoly regime could provide a degree of price stability and encourages long-term planning. In some countries, this model has resulted in the formation of a number of large vertically integrated companies, which – according to some studies – offer significant cost savings.1 America, Germany and Japan, which are covered in the following chapters, offer examples of this regulatory paradigm, with a small number of large vertically integrated utilities dominating the business. Despite these purported advantages, the regulated monopoly model has a number of shortcomings, including the following: ●

●

●

●

●

1

Over-investment in rate base: Under traditional rate of return regulations in the USA, firms are rewarded for how much they have invested in infrastructure. Since the rate of return is typically fixed, regulated monopolies, all else being equal, prefer to have more assets in the so-called “rate base.” This leads to a perverse tendency to over-invest, if permitted by the regulator, who typically has to approve capital investments in advance.2 Security of supply: In the regulated monopoly model, the trade-off between quality and price was usually biased toward security of supply as customers had no choice. As will be described in several chapters of this volume, in the transformation process to a marketoriented model, this may cause some problems as customers, who are used to taking security of supply for granted, may not be conscious about the cost of quality or reliability.3 Risks borne by ratepayers: Regulated monopolies typically pass all investment risk to captive customers, who must bear it during the long life of capital assets. No customer choice: In exchange for “the obligation to serve”, regulated utilities get an exclusive service area, which means that customers have no choice in selecting a service provider even if a lower-cost option is available.4 While this may not be much of an issue for most small consumers, it may be a source of frustration for large, sophisticated consumers. Price disparities: In many countries where neighboring areas are served by different utilities and/or under different regulatory regimes, significant price disparities have evolved

For a comprehensive discussion of the benefits of vertical integration see Michaels (2005) and his references. 2 For example, see the seminal paper of Averch and Johnson (1962). 3 Kollmann, Andrea; Tichler, Robert; Schneider, Friedrich: Netztarife in Österreich: Bestandsaufnahme und Interantionaler Vergleich, Linz (2005) show that there is a strong correlation between network fees and quality of service. 4 Competition also forces unbundling of costs of service components and the removal of cross-subsidies among different customer classes – a desirable development from efficiency point of view. It also encourages suppliers to pay closer attention to customer demand and offer services tailored to their changing needs.

Why Restructure Electricity Markets?

●

37

over time, causing significant pressure for change. Yet rigid regulations typically do not allow customers to switch suppliers. Price subsidies: In many countries, the regulators, usually with tacit approval of politicians, find it convenient to allow, or encourage, cross-subsidization of rates between different customer classes. The same applies to various cost components of bundled service, making it difficult, for example, to decipher the cost of energy versus transmission services, versus distribution. As will be pointed out, competition encourages and/or requires costs to be unbundled and be divulged, making it difficult to cross-subsidize certain customer classes or provide cross-subsidies between various service components.

On the other hand, the regulated model has a number of other attributes with potentially significant consequences for efficiency and price transparency: ●

●

●

5

Supply adequacy: Both SOEs and regulated monopolies are subject to a highly skewed reward and penalty system, which predisposes them to over-build. The reason is obvious. There will be a public inquiry and an outcry if there is a shortage of capacity, while hardly anyone notices if there is a little extra capacity sitting idle. One is tempted to conclude that because of this phenomenon, regulators rarely had to intervene in the past to ensure adequate reserve capacity – the ESI would always err on the side of having a little too much. By contrast, the issue of who is responsible for resource adequacy has become a hot topic in a number of countries since the introduction of competition.5 Manipulation by politicians: SOEs and regulated monopolies offer irresistible opportunities for regulators and politicians to micro-manage and/or “meddle” in the business.6 Examples include efforts to manipulate prices, offer subsidies to certain constituents, or engage in other political or social experiments – which would be hard to justify in a competitive environment with publicly listed companies. For example, in many Indian provinces, farmers pay next to nothing for electricity – as a matter of political expediency. In Britain before liberalization, the coal unions enjoyed strong political influence, which included the use of uneconomic British coal by government-run Central Electricity Generating Board (CEGB).7 In many US jurisdictions, commercial and industrial customers routinely subsidize residential customers. In California, for example, residential customers in hot interior climate zones are subsidized by those living in the more temperate coastal zones. Similar examples abound in other countries. It must be noted that governments continue to meddle in the utility business even after the firms are privatized and operate competitively – by imposing, for example, renewable portfolio standards, energy efficiency mandates and other measures. Nuclear energy: Capital-intensive nuclear industry owes special thanks to the relative long-term stability of the regulated environment. Without long-term contracts and/or

In the USA, for example, the Federal Energy Regulatory Commission (FERC), devoted a major section to discussion of supply adequacy in its standard market design (SMD) proposal in 2002. The issue of supply adequacy is covered in a number of chapters in this book. 6 An example of political meddling, not unique by any measure, occurred in late 2003 in Italy, where the independent regulator, the Authority for Electricity and Gas, proposed to reduce ENEL’s average rates by 2% over a 4-year period just as the government was trying to partially privatize the company. The regulatory decision had a negative effect on ENEL’s public offering, then 61% owned by the government. The government subsequently put intense pressure on the regulator to “rethink” its earlier decision, making a mockery of its independence. 7 See chapter on Britain in this volume for further details.

38

●

Electricity Market Reform government assurance, private investors would shun large, lumpy investments with significant upfront investments, a hallmark of nuclear power plants.8 Public disclosure and coordination: Another potential advantage of regulation, not fully appreciated until recently, is that it may encourage public disclosure as a part of the regulatory process while enforcing some coordination in forecasting and planning.9

Under a “deregulated” paradigm, where vertically integrated utilities are typically broken apart, there may be little coordinated planning. The problem, some believe, is particularly acute due to the bifurcation of generation and transmission planning – which now takes place within different, often competing, companies.10 Moreover, in a competitive environment, there may be strong disincentives in communication and coordination.11 In the North American context, recent “boom and bust” cycles in capacity investment may have been exacerbated by the introduction of competition in the wholesale market. In the recent past, there have been significant capacity gluts in certain regions and subsequent drops in wholesale prices. Wholesale competition has also been blamed for increased transmission congestion problems since transmission and generation planning have become virtually disjointed. In some cases, the result has been significant local capacity build-ups while shortages afflict other areas. In some regions, long-term resource planning has become chaotic with concerns about who is responsible for maintaining adequate reserve margins, as a number of chapters in this book point out.12 Further research appears warranted to identify and quantify the existence and prevalence of these problems and whether they impose significant risks. As the chapter on Europe claims, the introduction of competition on the basis of the 1996 European Commission directive may have resulted in dampened interest in generation investment. This, in turn, has raised the significance of “security of supply,” which is further described in several chapters, notably in case of Germany. In the European context, current projections show potential problems on the horizon regarding supply adequacy (Fig. 1.1). It will be a test for the market to see if additional generation capacity will develop in the new framework.

8 The chapter on Japan describes the difficulties facing Japanese regulators in defining a sustainable future for nuclear power, which has significant energy security implications for Japan. 9 One does not need a regulator for this. For example, the California Energy Commission maintains public records on all plants under construction in California regardless of their ownership including IPPs and co-generation units that are not regulated. 10 There are those who point out that coordination is not all that complicated, and future and forward contracts can substitute for lack of ownership of the long supply chain. This is true in theory but not always in practice, for example, in cases where generators cannot secure long-term contracts for their output, or secure transmission rights to transmit their output to distant markets. In the case of wind power, for example, developers typically select wind parks with little concern about availability of transmission capacity or distance to load centers. 11 In fact, it can be argued that competing generators have every incentive to confuse their potential rivals by spreading false rumors about their future investment plans, bluffing or making misleading announcements designed to confuse and distract. The last thing a private generator wants to do is to spell out its plans to its rivals. 12 Under regulation, non-competing utilities had strong incentives to coordinate their plans and make sure that the “neighborhood” had sufficient capacity to maintain reliable service and to keep the regulators at bay.

39

Why Restructure Electricity Markets? GW 70,0 65,0

62,0

62,4

60,5

(60,5)

(60,6)

60,0

(56,3) 55,0

55,0 50,0

50,2

51,6

53,0

(53,4)

53,8

(54,4)

January 2008

January 2009

52,0

45,0 January 40,0 35,0 30,0 January 2004

January 2005

January 2006

Remaining capacity

January 2007

January 2010

5% of GC ⫹ margin against peak load

Fig. 1.1. Projections of installed capacity and peak demand for Europe, January 2004–2010, in GW. Source: UCTE.

1.2. Drivers of change in regulatory paradigm Starting in 1980s, some economists and policy-makers started to question the wisdom and the necessity of centrally planned SOEs and regulated monopolies. A number of factors contributed to the new thinking13 including one or more of the following – not necessarily in any order and not applicable in all cases: ●

●

●

13

Gas turbines: The advent of highly efficient natural gas-fired turbines made it possible to build smallish units in record time with little risk. This broke significant barriers to entry in generation and made large, capital-intensive plants less attractive. As documented in Figure 1.2 for the USA, the dominance of gas-fired technology has been overwhelming in the recent past, and the phenomenon is not unique to America. Sympathetic regulators in a number of countries have removed the barriers for newcomers to enter the generation business and compete with the incumbent regulated utilities. Independent power producers (IPPs) have made significant inroads in many markets, in many cases, using highly efficient gas turbines. Ideology and politics are believed to have played a role in some cases.14 Privately listed companies, under commercial competitive pressures, can be expected to trim bloated staff levels, shed uneconomic contracts, and be forced to divulge performance information.

Among 62 countries studied, Pollitt (1997) found liberalization in 51, privatization in 30 and vertical separation in 27. The interest in market reform, by all indications, has intensified since 1997. 14 Newbery (1999) says, “one rather cynical view of electricity privatization (in Britain) is that it was part of a campaign against the overmighty public sector unions which was designed to undermine the power base of the opposition labor party.” The number of British coal miners which stood around 160,000 in 1984 dropped to 10,000 in 1994, and has continued to fall.

40

Electricity Market Reform 200

186

180 160 140 GW

120 100 80 60 34

40

11

20 0

Natural gas

Coal

Wind

6 Hydro

15 Other

Fig. 1.2. The dominance of gas-fired turbines is evident in the US context. New capacity built or proposed in the US between 2003 and 2007 in GW. Source: Platt’s RDI.

●

●

●

● ●

15

Public debt may have been a motivating factor in a few cases, such as in Victoria, Australia, where the sale of state-owned assets brought relief to the heavily indebted government of the time. Regulatory complexity played a role in California, where state regulators reached the humbling conclusion that, despite their best efforts, California electricity prices were 50% above national average. In this case, the regulators conceded that they could no longer do an adequate job of regulating the utilities, leading to the decision to “restructure” the industry to allow market discipline to regulate the industry instead.15 As further described in Chapter 10, the restructuring did not work very well, resulting in even higher prices for California consumers. Inadequate investment in infrastructure is among the primary reasons for many rapidly growing economies, which privatize and/or liberalize their ESI to attract foreign investment.16 Poor accountability by SOEs is among the motivating factors in some cases.17 Decentralized decision-making is a motivator in cases where central government can no longer cope with the growing complexities of forecasting, financing, constructing, operating and maintaining the network.18

In mid-1990s, the IOUs in California where encumbered by heavy debt having completed a number of costly nuclear power plants while the new IPPs were building highly efficient natural gas-fired gas turbines at a time when natural gas prices were low. This led to protracted discussion about “stranded costs,” which, in retrospect proved to be illusory. Refer to Chapter 10 for further details. 16 Financial lending institutions including The World Bank, the Inter-American Development Bank (IADB) and The Asian Development Bank (ADB) were unanimous in recommending privatization as a way to attract foreign investment in infrastructure, sometimes without adequate regard to the local underlying institutional constraints leading to significant problems in implementation. 17 In some countries like France and Korea, the dominant position of the monopoly utility company has simply been overwhelming. 18 In Iran, a frustrated official related a story that the head of a provincial generating plant had called to inform him that the plant would be shut down in a few days time because the fuel in the storage tanks


41

Box 1.1 Terminology Terminology used to describe different approaches to change the regulatory paradigm: Restructuring is a broad term, referring to attempts to reorganize the roles of the market players, the regulator and/or redefine the rules of the game, but not necessarily “deregulate” the market. Liberalization is not synonymous with restructuring. It refers to attempts to introduce competition in some or all segments of the market, and remove barriers to trade and exchange. The European Union, for example, refers to their efforts under this umbrella term. Privatization generally refers to selling government-owned assets to the private sector, as was done in most countries that have embarked on market reform. Corporatization generally refers to attempts to make SOEs look, act and behave as if they were for-profit, private entities. In this case, a SOE is made into a corporation with the government treasury as the single shareholder. For example, former SOEs in New South Wales, Australia, have been corporatized. They vigorously compete with one another, while all belong to the same, single shareholder, namely the Government Treasury of NSW. Deregulation refers to removing or reducing sector-specific regulation and subjecting the ESI to the monitoring by the anti-cartel authority. However, no electricity market has been (or, in fact, can be) fully deregulated. There is agreement now that the monopolistic bottlenecks of the market in transmission and distribution need specific regulation in addition to general anti-cartel policies.

Depending on the real or perceived problems with the status quo, and what was viewed as the solution, a different approach was pursued in different countries.19 The following chapters of this book offer a number of examples from different countries on what prompted a change in the regulatory paradigm, how the new market structure was designed and implemented, and what have been the results thus far. Box 1.1 provides definitions of a few key words. 1.3. Alternative views on competition and regulation In most cases, the shift in regulatory paradigm includes a desire to encourage competitive market forces to substitute for command and control regulations, or bureaucratic and often inefficient management of SOEs. Since some segments of the ESI are natural monopolies and cannot be made competitive, the focus of market reform has been predominantly on generation and supply functions, which can be competitively provided.20 The question is how to make the whole system work efficiently, while parts of it remain regulated. 18

(continued) was running short, pending shipments ordered by the central authorities. The manager could not imagine taking the matter into his own hands and securing additional fuel. Examples like this abound in countless other centrally planned economies. 19 Jamasb and Pollitt (2005) provide a comprehensive discussion of the European market reform initiatives. 20 Refer to Chapter 2 as well as Jamasb and Pollitt (2005) for a discussion of these issues.

42


This book’s Foreword, Introduction and Chapter 2 offer useful insights gained after two decades of studying markets. The following chapters provide a variety of models and market structures. At a fundamental level, there may be two basic views on competition and the role of regulation in the ESI: ●

●

Create level playing field where the regulator plays the role of the referee: Those who subscribe to this view see the prime purpose of restructuring as devising a set of rules and conditions that encourage free and fair competition among the players – and believes that the main role of the regulator is to be an impartial referee. This implies regulating only the monopolistic bottlenecks of the industry and deliberately refraining from regulating generation or supply. There are good reasons for such a strategy as Knieps points out in Chapter 2. A number of authors in this book favor a light-handed regulatory philosophy, leaving as much to markets as markets can reasonably handle. Littlechild, for one, believes in “competition where possible, regulation where not.”21 Write a script and make the puppets dance: Those who subscribe to this view prefer a more prescriptive market and a more hands-on role for the regulator. In his case, the regulator prepares a prescriptive plot for the play, making sure that the players act according to the script. As noted by Knieps in Chapter 2, “It is well known from the positive theory of regulation that regulators have strong incentives to over-regulate, mix regulatory instruments in an unsuitable way, favor the application of detailed regulation and call for a heavy-handed supervision of firms.” Authors in some chapters appear to favor the strong hand of regulator over the invisible hand of the market.

The central ingredients of the former are well-known, namely: ● ● ● ●

● ●

●

Remove barriers to entry in generation.22 Privatize or corporatize all players so they are competing on the same footing.23 Unbundle vertically integrated enterprises to remove cross-subsidies and self-dealing.24 Ensure that the transmission network is open and accessible to all under transparent and non-discriminatory prices.25 Create wholesale markets that are open and transparent. Ensure that the grid is managed by an independent operator who maintains reliability, manages transmission congestion, operates various markets to facilitate trade, liquidity and risk management. Foster competition in the supply business.

These conditions will most likely encourage a number of firms to enter various segments of the business, engage in short- and long-term transactions with one another and with customers. With sufficient degrees of freedom, the participants will find suitable arrangements to transact, and the regulator’s role is primarily that of a vigilant referee, making sure that 21

See Littlechild (2005). In the US context, the Public Utility Regulatory Act (PURPA) of 1978 accomplished this goal to a great extent. 23 In the Australian context, Victoria has done this while the neighboring New South Wales has maintained government control of both generation and distribution, creating an uneven playing field especially in the retail sector. 24 As pointed out in Chapters 9 and 18, lack of rigorous physical unbundling is considered as a major obstacle the creation of a vibrant competitive market. 25 Chapters 9 and 14 address the problems associated with this thorny issue in the European and US context. 22


43

the rules of the game are obeyed, infringements are caught and offenders are punished. Otherwise, market participants are free to roam as long as they obey by the rules and play within the field. Most restructured markets have adopted a variation of this theme – with varying levels of autonomy and authority for the referee. This assumes that the players know best, and the market is simply the sum of its components. Free market advocates and those favoring laissez faire prefer a limited role for the regulator, allowing market participants maximum flexibility.26 The alternative view of the market assumes a stronger, perhaps intrusive role for the regulator, most likely requiring frequent intervention in the market. In this case, the roles of the players are carefully orchestrated and their moves are monitored and controlled. There is, of course, no assurance that the scriptwriter, no matter how clever, can get all the parts right. Nevertheless, examples of this line of thinking persists and may be found in some countries, notably those who have had a disappointing experience with market liberalization and reform. There may also be considerable public pressure on governments to replace the invisible hand of the market by the intrusive hands of government. The degree of government interference is part of the political culture of a country that has evolved over a long time and often is amazingly stable. Tradition, to some extent, explains the deviation between the textbook approach to restructuring and the implementation of reforms in different countries.

1.4. In search of insights The following chapters of this book expand on many of the topics identified here while exploring others. Many important answers are provided in the context of examining specific experiences of particular markets around the world. While many insights are offered, a number of questions remain essentially unanswered, either because we have not examined the issues or have not found the answers. Additionally, three important caveats must be mentioned at the outset: ●

●

●

First, the co-editors of this volume made a conscious decision not to influence the personal perspectives of individual authors or to restrict their views in any way. Consequently, this volume offers a wide range of philosophies on markets, which the reader should find refreshing. Second, while we have made an effort to cover most interesting restructured markets, not all are included. Third, we have purposely limited the scope of the discussions to market design, structure and performance issues – leaving out other important topics such as renewable energy technologies, global climate change, environmental and energy security issues, to name a few.

In this book’s Foreword and Introduction, Stephen Littlechild and Paul Joskow share their own considerable experiences and perspectives, as well as an overview of many market design issues covered in this book. In Chapter 2, Guenter Knieps presents the analytical foundation of network-specific market power and provides a survey of the localization of monopolistic bottlenecks in different networks. Modern economic theory has developed the concept of disaggregated regulation

26

For example, generators would not be obliged to bid their capacity in the market, nor would they be required to explain their bidding strategies or prices.

44


of monopolistic bottlenecks against end-to-end regulation. He shows that there are solid reasons that regulation should be restricted to the bottleneck part of the network making sure that markets for complementary services like generation and supply will be able to breathe. As far as market power in these fields is concerned, it should be subjected to general competition policy rather than by sector-specific regulation. In Chapter 3, Ricardo Raineri describes the pioneering restructuring and privatization experience of the Chilean ESI, which began in 1980s. Chile has had its share of success and failures over the years and offers an early example of the difficulties of dealing with three distinct segments of the value chain, namely generation, transmission and distribution, along with an independent system operator (ISO). In the Chilean context, power generation was left to competitive investors, transmission was turned into an open–access regime, and distribution was left as a regulated natural monopoly. The system was tested during a severe drought in the late 1990s and subsequent crises, including chronic shortages of natural gas from neighboring Argentina. After 25 years of fine-tuning, the electricity market in Chile is still far from perfect, and because of the diverse and extreme conditions that it has experienced and what has been done, it offers useful insights for policy-makers and regulators around the world. Among the key lessons that can be learned from the Chilean experience are the importance of flexible prices, demand-side management tools, ancillary services, diversified sources of energy, binding international bilateral agreements and the constraints of modeling assumptions. In Chapter 4, David Newbery describes the experience of Britain, another pioneering market from which much has been learned, and continues to be learned. The British market may be characterized as the exemplar of electricity market reform, demonstrating the importance of ownership unbundling and workable competition in generation and supply. The original privatization created a de facto duopoly that supported increasing price–cost margins and induced excessive entry. This problem was addressed by trading horizontal for vertical integration in subsequent mergers. Competition arrived just before the Pool was replaced by the New Electricity Trading Arrangements (NETA). NETA, which intended to address many of the claimed shortcomings of the Pool, however, has had ambiguous market impacts. Increased competition caused prices to fall, inducing generators to withdraw capacity from the market and resulting in concerns about security of supply. Subsequently, price–cost margins increased and capacity was returned to the market. NETA was extended to Scotland in 2005 as the British Electricity Trading and Transmission Arrangements (BETTA) and the entire British transmission system is now under a single system operator. In Chapter 5, Eirik Amundsen, Nils-Henrik von der Fehr and Lars Bergman provide an overview of the evolution and subsequent expansion of the Nordic market, considered among the most successful competitive markets in the world. The Nordic market was put to a severe test during the drought of 2002–2003 where reservoirs in the hydro-dominated system fell to unprecedented low levels. But despite this natural shock, the market held together, without mandatory rationing, blackouts, price manipulation or major financial ruin of any of the players. This, in contrast to other markets – notably California – is an important hallmark of the Nordic market. In Chapter 6, Alan Moran describes how the Australian electricity market drew from the experiences of the UK and has forged a national electricity market (NEM) out of the five interconnected states. The NEM has proven effective in providing adequate incentives to encourage appropriate levels of new investment and the increased competition has driven prices down. Part of the system, which prior to 1994 was exclusively government owned, is now privatized, while some states have maintained government ownership. Despite shortcomings


45

of government ownership, increased efficiency is observable throughout the industry; the available data does however indicate that the private businesses have lower costs and better reliability. Future risks include government intervention, which might discourage optimal investment decisions, may increase costs and could threaten reliability. A major policy question has concerned transmission and how to ensure that it is built according to an incentive structure that is comparable to that facing generation. Some maintain that there is inadequate transmission between major state systems while others argue that there is over-build centered on some state governments’ wishes to export electricity without their generators facing the full costs. Australia’s long, stringy transmission network throws into greater relief issues seen elsewhere namely that new capacity and increased transmission can be substitutes and if one is favored over the other this can lead to poor investment choices and eventually cause supply shortfalls. In Chapter 7, Geoff Bertram traces New Zealand’s failure over two decades of reform to establish a viable industry self-governance framework, and the parallel failure to achieve restraint on monopoly profits by means of light-handed regulation. Starting from a classic monopoly of generation, transmission, distribution and retailing, New Zealand corporatized the ESI, separated lines businesses from generation and retail, removed retail franchises, and broke up the monopoly generator into five companies. These measures were insufficient to achieve competitive outcomes in the absence of hands-on regulation. Generators integrated vertically by takeover of retailers, and the resulting retail oligopoly erected an effective barrier to entry by withholding affiliated generators’ capacity from the very thin market for hedge contacts. Grid pricing and contract provisions foreclosed demand-side innovation and distributed generation. Distribution lines businesses ramped up markups from 30% to 70% without any regulatory restraint and were allowed to revalue their assets to underwrite the new high margins. Faced with failure of the original design, the Government in 2003 established a new industry regulator and invested in a new state-owned thermal generation to plug the country’s yawning gap in reserve generating capacity. In Chapter 8, Gert Brunekreeft and Dierk Bauknecht provide an overview of the German electricity market, still confronting major obstacles. The authors examine the problems and prospects for, and consequences of, new investment in the German market. The potential ramifications of a variety of energy policies are examined in the context of three important energy policy goals: preservation of the environment, provision of supply adequacy and fostering competition. While many countries are concerned about the adequacy of generation capacity, Germany follows a hands-off approach. To promote competition, Germany has now translated the latest European directive into national law, creating a specific regulator for the ESI. Environmental policy has a strong focus on promoting renewables, outside of the electricity market. The new European emissions trading scheme is shown to have a strong influence on future development. In Chapter 9, Reinhard Haas, Jean-Michel Glachant, Nenad Keseric and Yannick Perez cover the evolving situation in the key continental European markets, including Austria, Belgium, Czech Republic, France, Germany, Hungary, Luxemburg, The Netherlands, Poland, Portugal, Slovenia, Slovakia, Spain and Switzerland. Due to its geographic size, multiplicity of national regulatory agencies and trans-boundary transmission issues, attempts by the European Union to create a fully integrated market is progressing slowly. The authors identify a number of necessary conditions, not yet fully achieved, for a wellfunctioning competitive market including separation of grid from generation and supply, sufficient transmission and generation capacity, significant number of generators and a fully liberalized market. Failure to meet these conditions will result in distorted markets with suboptimal results.

46


In Chapter 10, James Sweeney provides an overview of California’s electricity restructuring in mid-1990s. California’s goal was to deregulate both the wholesale and the retail electricity market, after a transition period during which utilities would be able to recover stranded costs. Three regulatory rules were in place during the transition period: (a) regulators forced investor-owned utilities (IOUs) to divest at least 50% of their generating assets; (b) regulators precluded the utilities from entering forward contracts to acquire electricity; and (c) regulators imposed retail price caps on the utilities. In the midst of the transition period, the Western electricity crisis, driven primarily by a sharp reduction in available hydropower and limitations on natural gas supplies, drove wholesale electricity prices throughout the western USA to triple-digit or quadruple-digit levels. The three regulatory rules together proved disastrous, driving one utility into bankruptcy and another near bankruptcy, and left the state with billions of dollars of losses. Flaws in the California market design encouraged market gaming or exercise of market power by generators, traders, utilities and the state itself. With the crisis over, California has now established resource adequacy rules for utility electricity acquisition, encourages long-term contracts, and has eliminated retail price caps. An ISO-led market redesign project is underway to restructure transmission congestion management. Retail deregulation has been temporarily put on hold. And the future regulatory direction remains uncertain, with some advocates trying to force a return to a vertically integrated monopoly structure and others striving to improve regulatory rules within a partially regulated, partially deregulated system. In Chapter 11, Parviz Adib and Jay Zarnikau describe the much more successful experience of the Texas market, widely acknowledged as the most “robust” in North America to date. The Electric Reliability Council of Texas (ERCOT) has benefited from a confluence of positive factors, including a phased approach where the restructuring of the wholesale market preceded the retail choice, ample generating capacity at the outset of retail competition, and an intra-state market where a single state-level regulatory authority wielded near-exclusive jurisdiction over the implementation of the state’s restructuring plan. A number of challenges, however, remain as Texas seeks to implement more efficient means of managing transmission congestion and ensuring resources adequacy. In Chapter 12, Michael Trebilcock and Roy Hrab provide an overview of the Canadian market, pointing out that, so far, only limited restructuring has occurred. The chapter focuses on restructuring initiatives in Canada’s most populous province, Ontario, including initial setbacks and ramifications of re-regulation, and the lessons learned. The restructuring experience of the province of Alberta is also briefly examined. Both provinces altered their restructuring plans after experiencing unexpected price increases. In particular, the Ontario experience illustrates the importance of political commitment and how restructuring policies can be reversed quickly when a government fears a political backlash. In Chapter 13, Joe Bowring provides an overview of the PJM, considered by many as a role model for an efficient, centrally managed electricity market. The PJM interconnection manages the largest centrally dispatched control area in North America and operates the largest competitive wholesale electricity market in the world. PJM operates a bid based, security constrained, economically dispatched, locationally priced market with open–access transmission and financial transmission rights. PJM has achieved its success to date based on a number of factors including considerable experience as a power pool, well-defined market rules, a workable governance structure and an independent market monitoring function. The PJM model offers useful lessons for other aspiring ISOs and transmission system operators (TSOs). In Chapter 14, Richard O’Neill, Udi Helman, Benjamin Hobbs and Ross Baldick provide a background on important decisions emanating from the Federal Energy Regulatory Commission


47

(FERC) and some of the frustrations the agency has confronted as it attempts to implement its vision of an efficient national market with standardized rules and protocols. While Chapters 10, 11 and 13 examine particular markets with an ISO, this chapter provides a cross-cutting perspective on all the US ISO markets and FERCs broader public policy agenda. The authors provide perspective on regulatory reforms in the USA that began the movement toward organized, competitive markets. These reforms were accompanied by intense debate over the appropriate design of such markets and whether and how to mitigate market power. The chapter examines what motivated these debates and how different ISO regions experimented with alternative approaches. It explains why certain design features eventually were considered best practices and what issues remain for design consideration. It examines what measures of market performance have been implemented and how they should be interpreted. Finally, the chapter provides a summary view of the state of market design and market performance across each US ISO market. In Chapter 15, Taff Tschamler reviews the decidedly mixed experience in the US retail market. Despite a few state-level success stories, a national competitive US retail market has failed to materialize and does not even appear to be in the cards. This chapter provides an overview of the USA experience with competitive retail power market design with focus on default generation service. It sets forth the author’s assessment of effective default service policy, assuming an objective of vigorous competition and customer participation in retail electric markets. It also provides a summary of the various default service approaches or models adopted by states that allow customer choice. This chapter also provides historical data on the level of market activity by jurisdiction. In Chapter 16, Joao Lizardo R. Hermes de Araujo describes rather unique features of Brazil’s large hydro-based power system. A striking feature of market liberalization, which followed the debt crisis of 1980s, was that the divestment process and market reform followed two parallel and nearly independent paths. This, plus the mismanagement of reform and transition – particularly in view of the difficulties found in privatizing large generators – led to the 2001 power crisis. The new arrangement instituted by the Lula administration purports to ensure adequate expansion investment with an expanded role for central planning and coordination including a mechanism for regulated expansion auctions and contracts. There also is a role for contracting for the short and middle terms, which may grow. In principle, the new arrangements could solve the conundrum of thermal power investment in a large hydrodominated system. There are some positive signs that investment in transmission appears to be working. However, two major issues remain to be addressed: streamlining of environmental licensing of hydro plants, and how to ensure an adequate supply of gas to thermal plants and build the gas network in a market-oriented context. In Chapter 17, Isaac Dyner, Santiago Arango and Erik Larsen describe the market reform experience in Argentina and Colombia, not the first in Latin America, but presently considered among the most successful. This verdict, however, might change in the near future. Both are developing nations with a large hydroelectricity generation base with many similarities, yet followed different approaches to market reform more than 10 years ago. The elapsed time provides an opportunity to compare and contrast how well each market has fared and yield insights into the consequences of the initial and subsequent decisions on the market design. Together, the four Latin American countries covered in this volume offer a rich opportunity to examine the effects of different approaches to market liberalization. In Chapter 18, Mika Goto and Masayuki Yajima describe the early experience of introducing limited competition in the Japanese electricity market.

48


Japan started the market liberalization process in 1995 with the objective of improving the efficiency of vertically integrated monopoly power companies. While some other countries have taken bold – sometimes hasty – steps, Japanese policy-makers are taking a methodical and judicious approach. This, the authors argue, provides an opportunity to incorporate lessons from experiences of other countries, avoiding some of the pitfalls. This chapter reviews the ongoing debates about the vertical separation of utility companies, introduction of retail competition in the residential sector, dilemmas on how to preserve the country’s strategically important nuclear industry and insure adequate investment in infrastructure.

References Averch, H. and Johnson, L.L. (1962). Behavior of the firm under regulatory constraint. American Economic Review, 52, 1053–1069. Jamasb, T. and Pollitt, M. (2005). Electricity market reform in the European Union: review of progress towards liberalization and integration. The Energy Journal Special Issue on European Electricity Liberalization, 11–41. Kollmann, A. et al. (2005). Netztafiein Österreich: Bestandsaufnahme und Internationaler Vergleich, Linz. Littlechild, S. (2005). Beyond Regulation, IEA/LBS Beesley Lectures on Regulation Series XV. 4 October 2005, forthcoming in volume edited by C. Robinson, published by Edward Elgar. Michaels, R. (2005). Rethinking vertical integration in electricity. Department of Economics, California State Fullerton, Fullerton, CA, May. Newbery, D. (1999). Privatization, Restructuring, and Regulation of Network Utilities. MIT Press, Cambridge, MA. p. 148. Pollitt, M.G. (1997). The restructuring and privatization of the electricity supply industry in Northern Ireland – Will it be worth it? Mimeo, Cambridge University, February.

Chapter 2 Sector-Specific Market Power Regulation versus General Competition Law: Criteria for Judging Competitive versus Regulated Markets GÜNTER KNIEPS Institute for Transport Networks and Regional Policy, Albert-Ludwigs-Universität Freiburg, Freiburg, Germany

2.1. Introduction Since the comprehensive abolishment of legal barriers to entry into (almost) all network sectors, network economics has experienced a paradigm shift. Before the opening of the markets the controversial question was if and to what degree competition in networks could function at all; meanwhile, the central controversy of network economics has shifted to the distribution of tasks between sector-specific regulation and general competition law. The application of sector-specific regulatory intervention constitutes a massive intervention in the market process which requires a particularly well-founded justification. That the general competition law should be applied to the opened network sectors, too, is beyond dispute. Sector-specific regulatory interventions with competition policy objectives, on the other hand, are only justified if there is network-specific market power.1 Insofar as vague legal terms originating from general competition law – such as, for instance, market dominance – are being used to determine the need for sector-specific intervention they have to be corroborated by a localization of market power that is substantiated by network economics. Otherwise it is to be expected that market power is simply postulated, but not actually localized. A suitable economic reference model for establishing the regulatory activity necessary for disciplining market power in network sectors must be able to take into account the essential characteristics of networks without automatically equating them with market power. In the debate on the potentials of competition in opened network sectors the theory of contestable markets, developed by American economists in the 1970s (e.g. Baumol et al., 1982), 1

Technical regulatory functions (network security, allocation of frequencies, number administration, etc.) and the pursuit of universal service objectives by means of entry-compatible instruments (e.g. universal service fund) also constitute long-term sector-specific regulatory tasks, but are not examined in detail here.

49

50


is of central importance. This theory formulates the conditions that have to apply for potential competition to function as perfect substitute for active competition – under the assumption that there is a natural monopoly situation where one supplier can serve the market at lower cost than several suppliers. Critics of the theory of contestable markets refer in particular to the non-robustness of this theory (cf. e.g. Schwartz and Reynolds, 1983; Schwartz, 1986). Even the smallest alterations in the assumptions of the theory of contestable markets would destroy the disciplinary effects of potential competition and thus, it is argued, this concept is largely irrelevant for regulatory policy. As a consequence of this non-robustness of the theory of contestable markets the necessity of global sector-specific regulation may be postulated by the proponents of regulation. In the following it is shown that while the theory of contestable markets formulates sufficient conditions for the absence of network-specific market power, it does not provide a comprehensive definition of the preconditions for the existence of network-specific market power and the resultant need for sector-specific regulation. Whereas the theory of contestable markets examines only the role of potential competition with identical cost functions for both active providers and potential competitors (cf. Panzar and Willig, 1977; Baumol, 1982), effective network competition is not limited to potential competition. Newcomers can successfully distinguish themselves from the incumbent and find their own niche in the market, in particular by means of technological and product differentiation. An essential contribution of the theory of contestable markets to the study of potential competition consists in basing its analysis on Stigler’s concept of a barrier to entry (Stigler, 1968). For Stigler, economies of scale do not constitute barriers to entry, as they do not cause long-term asymmetries between incumbent and potential competitors. At this point, the theory of monopolistic bottlenecks, which is centrally concerned with localizing stable network-specific market power, becomes relevant. Based on this theory, a disaggregated regulatory policy can enable effective competition on complementary network parts by means of suitable access and price regulation. In this context, the essential facilities doctrine, which originates from US antitrust law, is taken up and generalized as a rule for the class of monopolistic bottlenecks. A monopolistic bottleneck in need of regulation only exists when both potential and active competition are lacking. In addition to potential competition, it is therefore also necessary to take into account the various potentials of active competition by means of product differentiation, price differentiation, technological differentiation, etc. As the theory of contestable markets disregards active competition, it is incapable of providing a comprehensive description of the potentials of competition after network opening. The theory of monopolistic bottlenecks was not specifically developed for a particular network sector, but is rather an economically sound instrument for localizing and disciplining remaining network-specific market power in all network sectors (e.g. railways, air traffic, telecommunications, etc.). It is shown that network-specific market power can be localized in most network industries, although strongly varying in extent among different network industries. In particular, monopolistic bottlenecks are relevant in electricity grids (transmission and distribution). Gas distribution pipelines are monopolistic bottlenecks too, although active network competition in supra-regional transmission pipelines may occur. Insofar as there are monopolistic bottleneck areas in network sectors, they require sectorspecific regulation in order to discipline the remaining market power. In particular, symmetric access to monopolistic bottleneck areas for all active and potential suppliers of network services must be guaranteed, so that competition can become fully effective on all complementary markets. It is shown that the application of regulatory instruments should be limited to the monopolistic bottleneck basis. Price-cap regulation should be applied, leaving the design of flexible pricing structures for network access to individual firms. Although

Market Power Regulation versus General Competition Law

51

perfect regulatory instruments do not exist, neither end-to-end regulation, nor access holiday regulation can be recommended. This chapter is organized as follows: Section 2.2 is devoted to an analytical foundation of the concept of network-specific market power. After an outline of the theory of contestable markets there follows a short description of the theory of monopolistic bottlenecks and its relation to the essential facilities concept. Section 2.3 contrasts the theory of imperfectly contestable markets with the theory of monopolistic bottlenecks. It is shown that the controversy about the relevance of the theory of contestable markets, in particular the criticism of non-robustness, does not constitute a fundamental argument against the importance of localizing network-specific market power. On the contrary, it is shown that monopolistic bottlenecks are robust structural characteristics that exist independently of alternative behavior assumptions. In Section 2.4 a short survey of the localization of monopolistic bottlenecks in different network industries is provided, leaving a more explicit analysis of localizing monopolistic bottlenecks within the energy sector to Section 2.5. In Section 2.6 the concept of a disaggregated regulation of monopolistic bottlenecks is provided, avoiding end-to-end regulation as well as access holidays.

2.2. The Search for Sector-Specific Market Power 2.2.1. The theory of contestable markets The concept of contestable markets takes its starting point from the question of the disciplinary effects of potential competition in natural monopoly areas.2 The threat of potential market entry was recognized as early as 1859 when Chadwick pointed out the difference between “competition for the field” and “competition within the field of service”. In 1968 Demsetz proved that the existence of a natural monopoly does not per se necessitate regulation. The almost axiomatic connection between the production structure of a natural monopoly and its need for regulation with respect to market entry, market exit, and pricing in a world without uncertainty does therefore not stand up to economic analysis. Auctioning off the right to supply a market that is a natural monopoly can, if necessary, replace competition in the market. According to Demsetz (1968, p. 58) the following two conditions are crucial for the functioning of such an auctioning process: competition on the input markets (many potential bidders), and no prior arrangements between competing bidders. In order to define the effects of the threat of potential competition more precisely, in the second half of the 1970s the theory of contestable markets was developed (Baumol, 1977, 1982; Panzar and Willig, 1977; Baumol et al., 1982). A market is termed contestable if entry is completely free and exit is completely free of charge (cf. Baumol, 1982, p. 3). The term “free entry” is here characterized by the absence of barriers to entry as defined by Stigler (cf. Bailey and Panzar, 1981, p. 128). However, this does not mean that market entry is free of charge or easy, but rather that new entrants have no cost disadvantages compared to the active supplier. As long as the inputs are available to both active and potential market participants under identical conditions, according to Stigler they do not produce barriers to entry.3 Thus economies 2

A natural monopoly exists if the cost function in the relevant area of demand is subadditive (cf. Baumol, 1977, p. 810). When examining the cost side of networks, economies of scale and scope in providing services, are of particular importance, which can result in a single supplier being able to serve the market at a lower cost than several suppliers. 3 Stigler defines a barrier to entry as “… a cost of producing (at some or every rate of output) which must be borne by a firm which seeks to enter an industry but is not borne by firms already in the industry” (Stigler, 1968, p. 67).

52


of scale do not cause a barrier to entry, as long as the entrant has access to the same cost function. Stigler’s concept further implies that traditional competition parameters such as product differentiation and the resultant creation of reputation and goodwill or the requisite capital do not constitute barriers to entry either, as they, too, affect active and potential competitors alike. In other words, in these contexts the cost function is only dependent on those factors to which all (potential and active) market players have symmetrical access.4 The theory of contestable markets proposes a model framework where, in the case of natural monopolies, potential competition is the perfect substitute for absent actual competition provided by active market players. The following assumptions are being made: ●

●

●

Free market entry: There is a large number of potential competitors with unlimited access to the most cost-efficient technology of a natural monopoly. Absence of irreversible costs:5 The investment necessary for market entry can be fully recovered in case of market exit. Exit is possible without incurring costs or loss of time. Bertrand–Nash behavior: The potential competitors calculate their chances on the market by taking the incumbent’s current prices as given and undercutting them. Perfect information on the part of all players is assumed, that is, there are no searching costs, so that even small changes in prices result in an immediate switch of the total demand to the cheapest supplier. Customers do not feel bound to a specific firm; they can switch with little effort when a new supplier appears.

The essential characteristic of a contestable market is its susceptibility to hit-and-run entry. Even a small, temporary profit on the part of the active supplier creates incentives for the entry of a potential competitor. Prerequisite for this, however, is a sufficient flexibility in prices, so that a potential entrant can undercut the active supplier (cf. Bailey, 1981, p. 178). The question may arise if and to what extent the basic assumptions of the theory of contestable markets are relevant in the real world at all. It is indeed an essential characteristic of the functions of competition on the open markets for network services that business strategies such as product differentiation, price differentiation, creation of goodwill, creation of an efficient distribution network, etc. are also used strategically. In addition, information problems (search costs, asymmetric information, etc.) can be relevant.6 It is true that the 4

The barriers to entry that traditional industrial economics – based on Bain (1956) – distinguishes (economies of scale, product differentiation, high requirements of capital), on the other hand, do not allow a reliable proof of stable market power (cf. e.g. Schmalensee, 1989). For instance, von Weizsäcker (1980a, b) shows that reputation and goodwill constitute efficient mechanisms for reducing uncertainty, and thus may lead to an increase in social welfare. 5 For the incumbents, irreversible costs no longer affect decision-making. Potential entrants on the other hand have to decide whether or not to incur these irreversible costs in the market they wish to enter. The incumbents, therefore, have lower decision-relevant costs than the potential entrants. Irreversible costs in combination with a natural monopoly constitute a credible threat that may discourage a second network operator from entering the market. Although even the irreversible costs have to achieve riskequivalent rates of return, they would be irrevocably lost after market entry. Therefore, the threat that the incumbent could temporarily reduce its prices to the variable cost level is indeed credible. 6 The Bertrand–Nash assumption of the theory of contestable markets does not aim to deny the more or less severe information problems of real markets. No sector-specific stable market power can be derived from information problems alone, because markets are resourceful where the (endogenous) development of institutions for overcoming information problems is concerned, for instance by the creation of goodwill. On the other hand, in natural monopolies with irreversible costs stable market power exists, even if all players are perfectly informed.


53

simplified models employed by the theory of contestable markets do not provide a comprehensive description of the manifold potentials of competition. However, this fact should not lead to the opposite conclusion that competition will therefore not work on principle. Neither does this mean that the application of general competition law on these markets should be rejected. However, as on all other competitive markets, the burden of proof whether market power exists and whether it is being abused rests with the competition authorities. In contrast to a general sector-specific regulation, such interventions into the competitive process should only be made case-by-case and ex post.7

2.2.2. The theory of monopolistic bottlenecks 2.2.2.1. Localizing network-specific market power The theory of monopolistic bottlenecks is based on a strict application of Stigler’s concept of barriers to entry in order to identify network-specific market power. It constitutes a central component of the disaggregated regulatory approach (cf. Knieps, 1997a, pp. 327ff.; 1997b, pp. 362ff.): the localization of network-specific market power in order to determine the minimal regulatory basis. Its objective is to derive, based on the principles of network economics, a regulatory basis consistent for all network sectors which justifies sector-specific regulatory inventions, irrespective of historical or institutional coincidence. All other network areas are subject to general competition law. The special focus of regulatory activity should be on the design of a symmetrical regulation of the access to monopolistic bottlenecks, combined with a regulation of access charges. The issue of the monopolistic bottlenecks, especially the concomitant problem of network access, is frequently discussed in network economics (Baumol and Willig, 1999, p. 44; Laffont and Tirole, 2000, p. 98; Kuhlmann and Vogelsang, 2005, p. 34). The conditions for a monopolistic bottleneck facility are fulfilled: 1. If the facility is necessary for reaching consumers, that is, if no second or third such facility exists, that is, if there is no active substitute available. This is the case if there is, due to economies of scale and economies of scope, a natural monopoly situation, so that one supplier can provide this facility at a lesser cost than several suppliers. 2. If at the same time the facility cannot be duplicated in an economically feasible way, that is, if no potential substitute is available. This is the case if the costs of the facility are irreversible. The entire value chain has to be examined in a disaggregated manner that is it has to be differentiated into those network areas that do have bottleneck characteristics and those that do not. Non-bottleneck areas are characterized by effective competition. The latter is by no means confined to potential competition. Both active and potential competition with and without technological differentiation as well as product differentiation and innovation (of both products and processes) constitute potential parameters of effective competition. Service networks due to the absence of irreversible costs unquestionably have non-bottleneck character; they may or may not possess the characteristics of a natural monopoly. When establishing 7

Competition authorities have to weigh up two possible sources of error. A false positive error occurs, when the competition authority intervenes in the competitive process, even though competition is functioning and there is no need for any active competition policy. A false negative error occurs when the competition authority fails to act, even though competition policy measures are indeed called for. Within the context of analyzing predatory pricing this differentiation has already been pointed out by Joskow and Kleworick (1979).

54


proof that a facility is a monopolistic bottleneck, it is crucial to concentrate exclusively on those network areas where there is a lack of active as well as potential competition and, consequently, no economically feasible alternative network access on the downstream markets. For instance, if a service network provider can choose between alternative network infrastructure providers, there is no monopolistic bottleneck, even if the infrastructures in question are not identical but, in accordance with the theory of monopolistic competition, characterized by product/technology differentiation.8 In network areas without irreversible costs no stable market power can be localized. Nevertheless fixed costs and economies of scale play an important role in these areas. Weitzman (1983) argues that the natural monopoly problem can only occur when irreversible costs are involved. There are certainly cases conceivable where market entry and exit at no charge is linked to small or non-existing economies of scale; in network sectors, however, economies of scale are usually of crucial importance, even in the absence of irreversible cost.9 Network areas where economies of scale are relevant but sunk costs do not occur are areas where competition is basically effective. The central concern of the theory of monopolistic bottlenecks is the differentiation, in the context of the disaggregated regulatory approach, between those network areas where competition is effective and those in need of being regulated, which are characterized as monopolistic bottlenecks. 2.2.2.2. Monopolistic bottlenecks and the concept of the essential facility When applying competition rules in order to discipline network-specific market power, the concept of the essential facility is of crucial importance. A facility or infrastructure is termed essential if it simultaneously: ● ● ●

is indispensable for reaching consumers and/or for enabling competitors to do business; is not otherwise available on the market; objectively cannot be duplicated by reasonable economic means.

This concept suggests the connection to the essential facilities doctrine, derived from US antitrust law, which is meanwhile being increasingly applied in European competition law also (cf. e.g. Lipsky and Sidak, 1999). The doctrine states that a facility is only to be regarded as essential if the following conditions are fulfilled: entry to the complementary market is not effectively possible without access to this facility; it is not possible for a supplier on a complementary market to duplicate this facility at a reasonable expense,10 and there are also no substitutes (Areeda and Hovenkamp, 1988).11 In the context of the disaggregated regulatory approach the essential facilities doctrine is no longer applied case by case – as is common in US antitrust law – but to an entire class of cases, namely, monopolistic bottleneck facilities. The design of non-discriminatory conditions of 8 However, the absence of stable network-specific market power when there are two or more infrastructure suppliers present does not mean that practical competition policy (e.g. merger control) is redundant. 9 For this cf. also Baumol (1996), pp. 57f. 10 Thus it is not feasible to offer, for instance, a ferry service without access to ports. 11 Occasionally an additional criterion for applying the essential facilities doctrine is formulated, namely, that the use of the facility is essential for competition on the complementary market, because it reduces prices or increases supply on this market. This criterion, however, merely describes the effects of access.


55

access to essential facilities must be specified in the context of the disaggregated regulatory approach. It is important in this context to view the application of the essential facilities doctrine in a dynamic context. Therefore, an objective for the formulation of access conditions must be to not obstruct infrastructure competition, but rather create incentives for both research and development activities, and innovation and investment on the facility level. 2.2.3. The theory of contestable networks as precursor for the theory of monopolistic bottlenecks The theory of contestable markets was developed at a time when even in the USA the process of opening the network sectors for competition was still in its early stages.12 The question that concerned economic policy was whether competition can function if economies of scale are involved (e.g. Willig, 1980): The theory of contestable markets was focussed on those markets that, due to significant economies of scale, promised, not so much active competition, but rather effective potential competition: for example, air transport or bus transport.13 The competition policy point of reference was the concept of the “invisible hand” of perfect competition, which is formulated in an extreme version by the general equilibrium theory. As the general equilibrium theory disregards the existence of economies of scale (e.g. Debreu, 1959), the theory of contestable markets was to permit them (and the concomitant non-convexities of production technology) in its model analysis, as they are of central importance for network sectors. Even though the theory of contestable markets is conceived as a partial analysis, all the other assumptions of the general equilibrium theory – complete information (no searching costs, no asymmetric information), no externalities, price as a perfect strategy parameter, no product or price differentiation, no dynamics – were transferred to the modeled world. Whereas market entry is not explicitly dealt with in the general equilibrium theory (all firms are seen as “atoms”, their active number determined endogenously in equilibrium), for the theory of contestable markets the role of potential competition is the focal point. The theory of contestable markets states that if there is a network structure with its concomitant economies of scale, competition in the form of potential market entry may quite possibly be effective. It is precisely by formulating the maximum role of potential competition for a scenario with a natural monopoly in combination with an absence of active competition that the theory of contestable markets laid an important cornerstone for the economically well-founded localization of network-specific market power after comprehensive market opening. The theory of monopolistic bottlenecks and the theory of contestable markets have their common origin in Stigler’s concept of a barrier to entry. The focus is therefore on the longrun cost asymmetries between active supplier and potential entrant. Although industrial economy’s quest for the “right” concept of a barrier to entry is still ongoing (cf. Carlton, 2004; McAfee et al., 2004; Schmalensee, 2004), Stigler’s concept has proven to be undoubtedly suitable for localizing stable network-specific market power. Neither adjustment costs nor the evolutionary dimension of market processes detract from the pivotal role of Stigler’s concept of a barrier to entry for localizing network-specific market power. The theory of contestable markets was rightly criticized for, due to its static character, disregarding essential functions 12

Thus interstate aviation was opened for competition in 1978, and interstate telecommunications – voice telephony in 1982, while intrastate network sectors remained almost completely closed. 13 In the USA, market opening for railway transport was not even considered for a long time; for the opening of this sector Europe was the pioneer.

56


of competition (cf. Carlton, 2004, pp. 468ff.); as regards competitive network sectors, however, the only conclusion to be drawn from this is that the manifold types of active competition are also important. However, it is by no means possible to derive the non-existence of stable network-specific market power by referring to network dynamics. Therefore, a periodic review of the phasing out potentials of monopolistic bottleneck regulation due to newly developed network alternatives would seem to be necessary (cf. Knieps, 1997a). 2.3. Imperfectly Contestable Markets versus Monopolistic Bottlenecks 2.3.1. The focus of the empirical studies on the theory of contestable markets The empirical studies on the role of potential competition examined the performance effects (e.g. price drops) from an end-to-end perspective, that is, on the end consumer markets.14 The effects of potential competition and price and product differentiation (e.g. hub formation, frequent flyer programs) were included, as well as the role of airports with scarce starting and landing capacities. Although the special importance of access to airports was particularly stressed (e.g. Bailey and Panzar, 1981, pp. 132f.), network access problems were analyzed on a level with other market imperfections.15 Two contrary estimates for the changes in performance to be expected on contestable markets can be distinguished. According to Bailey and Baumol (1984, pp. 130ff.), following market opening, the importance of nonoptimal network structures and of cost structures diverging strongly between businesses should decrease and contestability should improve.16 According to Bailey and Williams (1988, p. 191) market opening leads to a considerable increase in competitive strategies, via the price-cost range of the industries. According to Graham et al. (1983, p. 137) “schedule and service reliability” causes asymmetry in favor of the incumbent. This is designated as sunk costs, even though it does not constitute a barrier to entry as defined by Stigler. A disaggregated regulatory approach differentiating systematically between network areas with bottleneck character and those without bottleneck character, and thus dealing with the design of non-discriminatory access to airport capacities separately was not developed at that time. Consequently, there was no analysis of the question to what extent adequate network access conditions create the basis for active and potential competition on service markets in the first place. Thus no sufficient differentiation was possible between, on the one hand, the effects of market imperfections (e.g. information problems) – disregarded by contestability theory, because they do not only occur in network sectors but also on other markets – and, on the other hand, network-specific structural characteristics that need to be regulated, because they systematically impede network access and constitute network-specific market power. 2.3.2. One-sided focus on the role of potential competition At the center of the empirical studies on the theory of contestable markets was the question whether market opening did indeed show the effectiveness of potential competition. Those markets on which performance was influenced not only by potential suppliers but also by 14

Among the large body of literature on this issue (e.g. Bailey and Panzar, 1981; Caves et al., 1984; Morrison and Winston, 1987; Bailey and Williams, 1988). 15 Measuring the effects of market opening empirically is inherently very difficult (cf. Bailey and Panzar, 1981, p. 144). Thus questions arose such as: Is the industry not in balance? What is the role played by fluctuating fuel prices or the cost of intermodal alternatives; or, more generally, how would the market have developed if it had not been opened? 16 Cf. for this also Caves et al. (1984, p. 484).


57

the number of active suppliers were termed “imperfectly contestable” – as opposed to “perfectly contestable markets”, on which even a single potential entrant can fulfil a comprehensive disciplinary function (Morrison and Winston, 1987, pp. 58f.). But the disciplinary role of potential competition is also derived within the concept of imperfectly contestable markets (cf. e.g. Morrison and Winston, 1987, p. 61). Deviations from perfect contestability are attributed to, among other things, the importance of switching costs. The theory of contestable markets, on the other hand, is based on Stigler’s concept of barriers to entry and thus on long-run cost asymmetries (cf. Bailey and Panzar, 1981, p. 128) and abstracts from these types of market imperfections. It is conspicuous that in those network sectors where network infrastructures did not cause access problems (road haulage, bus transport, international shipping line services) the contestability hypothesis was hardly controversial (cf. Baumol and Willig, 1986, pp. 24ff.; Davies, 1986). Even though aviation had at first been considered the prime example of a contestable market, the criticism of the theory of contestable markets focussed increasingly on this sector. Even its supporters relativized the hypothesis of perfect contestability in the direction of imperfectly contestable markets.17 In their study on the American air transport market Bailey and Panzar (1981, p. 145) differentiate strictly between the roles of active and potential competition. The authors show that active competition in the air transport sector between long distance and regional providers can discipline the market for regional transport, even if potential competition on the regional level is absent. The implicit conclusion, however, is that on opened markets not only potential but also active competition can fulfil a disciplinary function – in conjunction with a concomitant differentiation of the cost function and product differentiation (small versus large aeroplanes, different times of departure, etc.). The full effect of a comprehensive market opening cannot be registered in its entirety by the theory of contestable markets. The role of access to infrastructures is also shown to be a source for the lack of perfect contestability (Morrison and Winston, 1987, pp. 61f.). Bailey and Panzar (1981, pp. 132f.) also emphasize the relevance of sunk costs for airports. But there is no strict differentiation being made between markets with stable market power (monopolistic bottlenecks) that are in need of regulation and markets with effective competition where sector-specific regulation is redundant. 2.3.3. Imperfect contestability versus monopolistic bottlenecks The studies on imperfect contestability transcend the abstraction level of the theory of contestable markets. Thus searching costs and information problems are taken into account, which may occur on all markets. The role of cost heterogeneity and product differentiation and the concomitant active competition are regarded as an indication for the absence of perfect contestability, although it cannot be concluded from this that there is no effective competition. It has been argued that, in spite of the problems of asymmetric access to infrastructures and the resultant market imperfections, competition functions in the sense of imperfect contestability (Bailey and Panzar, 1981, p. 134; Morrison and Winston, 1987, pp. 55f.). But this view does not obliterate the necessity to determine the remaining need for regulation in opened network sectors and to discipline it by means of suitable regulatory instruments. As has already been shown in Section 2.2.2, the existence of monopolistic bottlenecks and the localization of network-specific market power needing to be regulated presupposes the 17

It should also be noted that the contestability hypothesis was not tested for the American railway sector. AMTRAK still holds the monopoly for railway passenger transport services.

58


lack of both active and potential competition. If, following market opening, both active and potential competition emerge, the objective of creating effective competition is fulfilled. The differentiation between, on the one hand, markets on which only potential competition can be observed and, on the other hand, markets where there is both potential and active competition is irrelevant for the localization of monopolistic bottlenecks. 2.3.4. On the independence of network-specific market power from alternative behavioral assumptions The theory of contestable markets is a static theory (cf. Baumol et al., 1983, p. 495). It describes, by means of the Bertrand–Nash behavioral assumption, the maximum efficiency of potential competition in natural monopoly areas without irreversible costs. Here even minor price reduction potentials lead directly to market entry, so that residual demand is perfectly elastic (cf. Knieps and Vogelsang, 1982, p. 236). It has to be kept in mind that what is concerned here is a static equilibrium concept. Behavioral assumptions about dynamic qualities are irrelevant for determining the Bertrand–Nash equilibrium. In this sense, the non-robustness criticism by Schwartz and Reynolds (1983, p. 488) is misleading, because it is based on assumptions about dynamic qualities of the relation between entry lag and price adjustment lag,18 which are unnecessary for the concept of perfectly contestable markets and do not constitute a logical consequence of the absence of irreversible costs either (cf. Baumol et al., 1983, p. 496). Beside the Bertrand–Nash behavior assumption, however, other behavioral assumptions (e.g. the Cournot–Nash or von Stackelberg behavioral assumptions) are conceivable which could impede the disciplinary effects of potential competition considerably, even if there are no irreversible costs involved (e.g. Knieps and Vogelsang, 1982, pp. 236ff.). Basically, what is concerned here are quantity pre-commitments, whereas the theory of contestable markets emphasizes the formulation of pricing behaviors without quantity pre-commitments (Baumol and Willig, 1986, p. 14). Although such quantity pre-commitments cannot be completely excluded in markets without irreversible costs – just like in all other markets – no networkspecific stable market power that would justify sector-specific regulation can be derived from this fact. In the absence of irreversible costs, network-specific market power that would be robust under alternative behavioral assumptions cannot be proven, even in the case of a natural monopoly (cf. Knieps and Vogelsang, 1982). Market power on the basis of the Cournot–Nash behavior assumption becomes immediately unstable under the Bertrand– Nash behavior assumption. Interventions by competition authorities would thus have to be based on behavior hypotheses that are difficult to verify empirically. Even according to the Bertrand–Nash behavioral assumption, with the demand side perfectly willing to switch and perfectly informed, its exploitation by the supplier of a monopolistic bottleneck facility cannot be prevented. This is all the more true if the customers of a monopolistic bottleneck facility have tied themselves to a specific infrastructure provider via pre-commitments (e.g. in the context of the Cournot–Nash behavioral assumption), even if market entry cannot be excluded for all scenarios (cf. Knieps and Vogelsang, 1982, pp. 239f.). As a consequence, in the absence of irreversible costs network-specific market power does not occur even if the behavioral assumptions are altered. The reverse is also true, that is, the 18

While the entry lag mentioned above refers to the time period between the entry of a newcomer and the time when the entrant is able to sell his product, the price adjustment lag indicates the period between market entry and the incumbent’s price reaction.


59

structural characteristics of a natural monopoly in combination with irreversible costs create market power, which remains stable even under alternative behavioral assumptions. 2.3.5. The “␧ sunk cost” argument versus the localization of network-specific market power The effectiveness of potential competition – and thus the central statement of the theory of contestable markets – was challenged with the argument of non-robustness, that is, that even the existence of minimal irreversible costs could lead to a non-contestable monopoly scenario (“␧ sunk cost” argument) (cf. Schwartz and Reynolds, 1983; Schwartz, 1986, pp. 41ff.). Baumol et al. (1983) do not completely exclude this possibility, but on their part point out conditions under which, with very low irreversible costs, the performance of the markets approximates that of a perfectly contestable market. The most elementary formulation of this argument of the non-robustness of the theory of contestable markets in case of minimal irreversible costs can be found in Stiglitz (1987, p. 889). Stiglitz assumes that minimal irreversible costs can occur even on service markets and with almost all investments; he is, however, using a broader interpretation of the term irreversible costs than Stigler in his concept of barriers to entry.19 The non-robustness argument that even minimal irreversible costs result in market power is here presented in the context of a subgame perfect equilibrium. The crucial precondition for this is the assumption that all potential entrants are identical to the incumbent, so that it is only the time asymmetry of incurring the minimal irreversible costs that constitutes the potential for a plausible threat on the part of the incumbent. In the framework of this model, newcomers have no possibility to distinguish themselves from the incumbent by means of technology, product or price differentiation. It is immediately evident that the manifold forms of active and potential competition on service markets cannot be comprehensively dealt with by this approach. As regards the problem of localizing network-specific market power, however, the controversy about the role of minimal irreversible costs is misleading. Active network competition typically goes together with product and technology differentiation, so that the existence of minimal irreversible costs does not lead to a plausible threat regarding market entry. Instead, it is the presence or absence of a monopolistic bottleneck that constitutes the crucial factor in deciding whether network-specific market power exists. Only a natural monopoly in combination with irreversible costs and the resultant lack of active and potential competition lead to a scenario in which the non-discriminatory access of all players on downstream markets must be guaranteed by a suitable form of sector-specific regulation.

2.4. Localizing Monopolistic Bottlenecks in Different Network Sectors The theory of monopolistic bottlenecks was not specifically developed for a particular network sector, but is rather an economically sound instrument for localizing and disciplining remaining network-specific market power in all network sectors (e.g. railways, air traffic, telecommunications, etc.). The character and extent of monopolistic bottleneck areas vary considerably between the individual network sectors. Individual proof as to which network area does indeed meet the criteria for a monopolistic bottleneck is necessary, and it is also important to avoid the danger of an erroneous identification of a monopolistic bottleneck. 19 “

While much of investment is not sunk, however, there is a sunk cost element in almost all investments. An airline must advertise to obtain customers; it must solve complicated routing problems” (Stiglitz, 1987, p. 889).

60

Electricity Market Reform Table 2.1. Airports as monopolistic bottleneck facilities.

Air traffic Air traffic control Airports

Natural monopoly

Irreversible costs

X/— X X

— — X

Table 2.2. Railway infrastructure as a monopolistic bottleneck facility.

Railway traffic Railway traffic control Railway infrastructure

Natural monopoly

Irreversible costs

X/— X X

— — X

Table 2.3. Local telecommunications networks as monopolistic bottleneck facilities.

Terminal equipment Telecommunications services (including voice telephone services) Satellite/mobile networks Long-distance cable-based networks Local cable-based networks

Natural monopoly

Irreversible costs

— —

— —

X — X/—

— X X

If, for instance, due to technological progress, the conditions for a monopolistic bottleneck no longer exist, the corresponding sector-specific regulation must also stop (cf. Knieps, 1997a). In an environment of competing network infrastructure providers and a variety of networks and technologies this means that existing regulation should be reduced rather than extended. The combination of natural monopoly and irreversible costs can occur in different network sectors: For example, airport infrastructures – in contrast to aeroplanes – are associated with irreversible costs. Once made, investments in terminals and runways cannot be transferred to another location, the way an aeroplane can. Thus to the extent airports are natural monopolies, they constitute monopolistic bottlenecks. Railway infrastructure, unlike rail transport services and railway traffic control, represents a bottleneck facility, because the track operator holds a natural monopoly and the building of rail tracks involves irreversible costs.20 Air traffic and railway traffic may or may not possess the characteristics of a natural monopoly (depending on the relevant market). In the telecommunications sector, bottleneck facilities can (if at all) only be found in the local loop, while in long-distance networks there is both active and potential competition (Knieps, 2004). Although legal entry barriers do still exist in some parts of letter mail conveyance, postal services as a whole do not constitute a monopolistic bottleneck. The different components of postal services may or may not possess the characteristics of natural monopolies (Knieps, 2002a). Tables 2.1–2.4 illustrate the application of the theory of monopolistic bottlenecks to the network sectors mentioned above.21 20

For a more explicite analysis of the role of monopolistic bottlenecks in the area of transport infrastructures (see Knieps, 2006). 21 For a more detailed discussion, cf. Knieps (1997a); Knieps and Brunekreeft (eds.) (2003).


61

Table 2.4. No monopolistic bottlenecks in any component of letter mail conveyance. Components

Natural monopoly

Irreversible costs

Collecting Sorting of outgoing mail Transport Sorting of incoming mail Delivery

X/— X/— — X/— X

— — — — —

2.5. Localizing Monopolistic Bottlenecks Within the Energy Sector 2.5.1. The different value chains in the electricity and gas sectors Within the electricity sector four vertically successive stages can be distinguished: 1. 2. 3. 4.

power generation, transmission grid (high voltage), distribution grid (low voltage), retail.

Within the gas sector the following stages can be distinguished: 1. 2. 3. 4.

extraction (production), supra-regional transmission pipelines, regional and local distribution, retail.

It is important to differentiate between those parts of the value chain, which have the characteristics of monopolistic bottlenecks and the other, competitive parts. The monopolistic bottleneck parts are characterized by stable network-specific market power and should therefore be regulated ex ante. Although the competitive parts are typically not characterized as atomistic markets, the ex post application of competition law is sufficient. 2.5.2. Competitive generation/production and retail It is obvious that electricity generation, gas production as well as the retail level do not possess the characteristics of monopolistic bottlenecks. Several active suppliers can be identified, so that the characteristics of a natural monopoly are not fulfilled. Both traditional industrial economics22 and modern competition theory based on methods of game theory deal quite extensively with problems of market power. However, in many game theory model approaches the market power to be localized is already assumed to exist, and it is on the basis of this premise that the numerous competition-distorting effects of such (assumed) market power are analyzed.23 It can be shown that stable market power in markets with several suppliers cannot be unequivocally proven to exist either by means of empirical/econometric approaches, or by means of game theory approaches, and that, consequently, it is not possible to derive an economically sound basis for ex ante regulatory intervention in oligopoly markets (see appendix). 22 23

For an overview, see Schmalensee (1989), Chapter 16. A good overview of this model world is given in Tirole (1989).

62


2.5.3. Monopolistic bottlenecks in the transmission and distribution networks 2.5.3.1. The specific characteristics of natural gas transmission and electricity transmission compared For a differentiated analysis of competition potentials to be possible, the specific features of the commodity and the related particularities of transmission have to be considered. As part of such an analysis, it is also very helpful to examine the fundamental differences between natural gas and electricity. Natural gas is a primary energy available as a natural resource. This means that it can only be produced where the gas reservoirs are. Most of these reservoirs are situated abroad. The gas therefore has to be transported over long distances to the centers of consumption. For this reason there is extensive cross-border gas transportation from remote sources. Moreover, there are different gas qualities, which is why customers cannot use just any gas. Natural gas transportation occurs on specific routes between entry and exit points, with the gas in the pipeline flowing predominantly in one direction, even in the long term. An increase in a pipeline’s transportation volume entails an increased pressure drop. This pressure drop describes the difference between the gas pressure at the entry point and the gas pressure at the exit point, or between two precisely defined points of a pipeline (e.g. compressor stations24). Another characteristic feature of gas is the fact that it can be stored. International gas transportation not only relies on pipelines but can also be effected through liquefied natural gas (LNG) tankers. This involves cooling the gas down to its liquid state for transportation, which is then followed by regasification at the point of destination. As a secondary energy, electricity can, in principle, be produced anywhere (subject to regulations and approvals). It is predominantly generated in the country in which it is consumed. Both the location of the power stations and the design of the transmission network are selected to meet the needs of the domestic market. Above all, unlike the gas sector, the electricity industry does not rely on foreign production sites. Power transmission routes are fundamentally different from motorways, railways, and gas pipelines. Whilst the mere existence of other railway routes does not have any direct effect on the congestion problem or the problem of internalizing the externality costs on a given route section,25 the situation is totally different in power transmission. Here, the scope of the externality costs cannot be restricted to the direct transmission path between an entry point and an exit point. It rather depends to a large extent on the simultaneous generation (feeding into the system) and withdrawal at the various entry and exit points and on the fixed overall system parameters (voltage restrictions, etc.) in the network. This can be illustrated by a simple example (Hogan, 1992, p. 217): in an interconnected power grid, it is not possible to transport power directly from a particular entry point to a particular exit point. The current will rather choose a path through the grid, which offers the lowest resistance. At least part of the power will therefore not choose the shortest connection (contract path). So which route (or detours) the current will actually take does not just depend on the transmission capacities and resistances of the different lines but more so on the feed and withdrawal

24

In order to move the gas over long distances, transmission pipeline systems have compressor stations to keep boosting the gas pressure. 25 Route-based congestion externalities have long been investigated in traffic economies. Participants usually ignore the damage (e.g. longer travel times) caused by another vehicle on a certain route. These are physical externalities which cannot be internalized through market prices. One possible step would be to charge a congestion fee to the amount of the externality costs incurred by all other participants as a result of one extra journey. This ensures that each vehicle bears the total costs of its own journey.


63

rates at all of the entry and exit points. In an extreme case, the capacity of a transmission line between a particular entry point and an exit point will exclusively be used for power transmission from other entry points (cf. Hogan, 1992, p. 213).26 Transmission and distribution of electricity break down into different voltage levels. Power transmission is restricted to extra high-voltage networks covering what are usually large geographic areas. Its main purpose is to interconnect different power plants to allow utilizing the systems’ economies of scale. In addition, the extra high-voltage networks connect power generators with the local utilities’ distribution grids. 2.5.3.2. Power transmission networks as monopolistic bottlenecks Even after the full opening of the electricity market, power transmission networks continue as horizontally linked natural monopolies. At the transmission level, the degree of bundling (interlinking) is very high. The grids operated by the different power utilities are restricted to particular geographic regions and do not overlap. Integrated utilities operate what is in some cases a very complex transmission network within their own service areas. Inside these service areas there are no competing transmission lines owned by other network operators. Owing to the system advantages of an integrated power transmission network, which are based on the interlinking of all points of generation and consumption within the service area, it is most cost-efficient for a particular geographic area to be supplied by only one utility. So there is no incentive, either from a general economic or from a private investor’s point of view, to build individual parallel lines or networks within a given service area. It is therefore highly unlikely that newcomers will build alternative lines or parallel (partial) networks, even after the full opening of the electricity networks. Moreover, the development of transmission networks involves irreversible costs tied geographically to a particular location. These would be lost if a player were to withdraw from the market, because it would be impossible to recoup the capital if a network were to shut down. The transmission network allows the transport of electricity power at high-voltage over long distances and is considered a natural monopoly with highly sunk investment (Brunekreeft, 2003, p. 232). Accordingly, the power transmission network within a utility’s service area has the features of a monopolistic bottleneck. 2.5.3.3. Active network competition in supra-regional transmission gas pipelines The national supra-regional high-pressure natural gas pipeline networks are imbedded in the international pipeline transmission system. The pipeline networks are backbones linking the international network with the regional and/or local service areas. They also feed into downstream regional grids or supply local distribution system operators and industrial users. Depending on the location of the transmission pipeline, this can either be through direct branch lines or through more or less interconnected regional networks. If it can be shown that regional network operators and/or local distribution companies can choose between at least two different operators of supra-regional gas transmission networks, then it is no longer absolutely necessary to have access to the pipelines of a particular supra-regional gas transmission company, which in turn means there is no bottleneck situation at the gas transmission level.

26

The “loop flow” phenomenon, which is based on Kirchhoff’s laws, is thus essentially the same as the economic problem of system network externality.

64

Electricity Market Reform Table 2.5. Electricity networks (transmission and distribution networks) as monopolistic bottleneck facilities.

Generation Transmission networks Regional/local distribution networks Retail

Natural monopoly

Irreversible costs

— X X

X X X

—

—

It is obvious that competing pipeline backbones can only evolve if freedom to build pipelines is guaranteed by law. If instead, free entry is not allowed due to a restrictive licences policy, which guarantees nationwide monopolies, competing pipeline backbones cannot occur, although the characteristics of monopolistic bottlenecks may not be fulfilled. The competition potential created by project companies and pipeline ownership in undivided shares relies on identical transportation routes. Access to alternative supra-regional transmission networks operated by different companies does not necessarily mean that they provide identical transmission routes. Far more important is the question whether there is access to competing pipeline backbones. A transmission company’s backbone system comprises pipelines with utilization rights under joint project company ownership, pipelines co-owned in undivided shares and pipelines owned by only one company. For these pipelines, the concept of detour traffic is of lesser importance as long as existing alternatives for pipeline access do not generate prohibitively high costs. For operators of regional and local distribution networks, the relevant issue is whether there are several alternatives of access into supra-regional transmission pipeline systems. In contrast to electricity transmission grids, supra-regional gas pipelines do not necessarily fulfil the characteristics of natural monopolies and therefore do not necessarily possess the characteristics of a monopolistic bottleneck. For example, the national supra-regional transmission networks in Germany do not possess the characteristics of monopolistic bottlenecks (Knieps, 2002b). 2.5.3.4. Monopolistic bottlenecks in the regional/local distribution networks Electricity distribution networks deliver at low voltage to the end user. Local distribution networks of gas are highly integrated grids serving gas supply purposes in local service areas. Like other local grids, the routes usually follow the network of roads. In towns and cities, the local distribution mains of gas or electricity are usually installed directly under or alongside the streets. In terms of its geometry, the local grid therefore assumes the same structure as the actual street system. The distribution networks of gas or electricity are therefore considered as regional/local natural monopolies with sunk costs. The results of this section can be summarized in Tables 2.5 and 2.6. 2.6. Disaggregated Regulation of Access to Monopolistic Bottlenecks Insofar as there are monopolistic bottleneck areas in network sectors, they require specific residual regulation in order to discipline the remaining market power. In particular, symmetric access to monopolistic bottleneck areas for all active and potential suppliers of network services must be guaranteed, so that competition can become fully effective on all complementary markets.


65

Table 2.6. Gas distribution as monopolistic bottlenecks.

Extraction (production) Supra-regional transmission pipelines Regional and local distribution Retail

Natural monopoly

Irreversible costs

— X/—

X X

X —

X —

In contrast to networks under competition, the market power involved in monopolistic bottlenecks fundamentally disturbs the bargaining processes on access conditions. One extreme alternative could be (vertical) foreclosure of competitors on a complementary service market. Such a tying can be used as a method of price discrimination, enabling a monopolist to earn higher profits (Posner, 1976; Chapter 8). Another way of abusing market power in the bargaining process on access conditions is to provide insufficient network access quality or excessive interconnection charges. An example of inferior access conditions is lower quality access to railroad tracks. Monopolistic access charges are another danger when market power due to monopolistic bottlenecks is involved. 2.6.1. Application of regulatory instruments to the monopolistic bottleneck basis The effect of a total refusal of access to monopolistic bottleneck facilities can also be achieved by providing access only at prohibitively high tariffs. This shows that an effective application of the essential facilities doctrine must be combined with a suitable regulation of access conditions to bottlenecks with regard to price, technical quality, and timeframe. However, the fundamental principle of such a regulatory policy should be to strictly limit regulatory measures to those network areas where market power potential does indeed exist. A regulation of access tariffs to monopolistic bottlenecks must therefore not lead to a regulation of tariffs in network areas without market power potential. There are two further issues that have to be taken into account. On the one hand, the existence of competition on the service level should not lead to the conclusion that there is no market power potential on the upstream network level, as long as the latter fulfils the criteria of a monopolistic bottleneck (cf. Brunekreeft, 2003, pp. 89f.). On the other hand, there is the question of the minimum regulatory depth necessary to guarantee non-discriminatory access to essential facilities, without, however, disproportionately interfering with the property rights of the regulated firm.27 2.6.2. Price-level regulation of access charges The identification of monopolistic bottlenecks is always based on an intramodal perspective, a decisive factor being the need for complementary service providers to have nondiscriminatory access to such facilities. However, the existence of monopolistic bottleneck 27

Basically one has to differentiate between, on the one hand, the question whether, due to a monopolistic bottleneck, network-specific market power exists, and, on the other hand, the question what kind of regulatory intervention is suitable. Thus the so-called Hausman–Sidak test argues that a regulatory obligation to unbundle the local loop is not justified, if, even without unbundling, the incumbent is not able to exercise market power with regard to providing telecommunications services to end users (cf. Hausman and Sidak, 1999, pp. 425f.; Hausman, 2002, p. 138).

66


facilities does not necessarily guarantee that there will be long-term economic profits. Firstly, there is the possibility of the “necessary case”, where even unregulated infrastructure providers are unable to meet their costs. Secondly, competition between modes can severely limit an infrastructure provider’s profit potential. The reference point for regulatory rules concerning access charges should be the coverage of the full costs of the monopolistic bottleneck (in order to guarantee the viability of the facility). In particular, when alternatives to bypass essential facilities are absent, the costcovering constraint may not be sufficient to forestall excessive profits. Therefore, the instrument of price-cap regulation should be introduced (cf. e.g. Beesley and Littlechild, 1989). Its major purpose is to regulate the level of prices, taking into account the inflation rate (consumer price index) minus a percentage for expected productivity increase. It seems important to restrict such price-cap regulation to the bottleneck components of networks, where market power due to monopolistic bottlenecks is really creating a regulatory problem. In other subparts of networks price setting should be left to the competitive markets. Regulation of infrastructure access charges should be limited exclusively to price capping. The basic principle underlying price-capping regulation is that price levels should be regulated in areas where there is network-specific market power. The benefits of price capping in terms of efficiency improvements and future investment activities can only unfold if price capping is applied in its “unadulterated” form and not combined with input-based profit regulation. Individual pricing agreements amount to over-regulation that is harmful to competition. 2.6.3. Flexible pricing structures for network access The question remains whether regulators should also be allowed to prescribe pricing rules focussing on tariff structures within monopolistic bottlenecks. There are serious arguments for regulators to refrain from detailed tariff regulation. In the first place, firms should have the flexibility to design (Pareto superior) optional tariff schemes (e.g. Willig, 1978). Pricing rules prescribed by the regulator could induce inefficient bypass activities. For example, a first pricing rule could be access tariffs according to the long-run average costs of the essential facility. Since in such a case a differentiation among different user groups according to different price elasticities is not possible, incentives for inefficient bypass of the bottleneck facility may be created for certain user groups. A second pricing rule would be access pricing according to the Ramsey pricing principle. Mark-ups on the marginal costs of access to the monopolistic bottlenecks are chosen according to the elasticity of demand for network access in order to maximize social welfare given the cost-covering constraint. However, Ramsey prices could become unsustainable, even if applied strictly to monopolistic bottlenecks. The technological trend toward the unbundling of monopolistic bottleneck components increases the possibilities for inefficient bypass. Secondly, the danger arises that regulators extend the regulatory basis to include the contestable subparts of networks. From the point of view of increasing static (short run) efficiency such behavior could even be justified by welfare theory. It is well known that efficiency distortions caused by applying Ramsey pricing can be reduced by extending the regulatory basis (e.g. Laffont and Tirole, 1994). Nevertheless, such an endeavor would in fact mean a return to fully regulated networks, including priceand entry-regulation of the contestable subparts. As such, this would not be a suitable response to deregulation (e.g. Damus, 1984). Regulatory authorities should not force firms to apply specific pricing rules, such as Ramsey prices or two-part tariffs, as this would hamper their quest for innovative pricing systems. It is always possible that better rules will be found in future. The design of pricing rules should be part of the decision-making process of the firms.


67

2.6.4. End-to-end regulation versus disaggregated regulation Regulatory instruments can be differentiated according to whether they are limited to the bottleneck areas (disaggregated regulation) or applied globally (end-to-end), including the competitive segments (e.g. Laffont and Tirole, 2000; Chapter 4). Since the application of regulatory rules is not costless and may also be abused strategically to disturb market forces, the advantage of the disaggregated regulatory approach is the strict limitation of the regulatory basis to bottleneck services. Its disadvantage, however, is that incentives may be created to discriminate against firms in vertically related competitive segments (e.g. Mandy, 2000). This should be kept in mind when designing adequate rules for disaggregated bottleneck regulation. From an economic policy point of view, the use of ex ante sector-specific regulation involves massive interference with the market process and must therefore be supported by a well-founded justification. Even if the nature of the network means that the bottleneck areas are complementary to the other parts of the network, there is no reason whatsoever for end-to-end regulation and a general use of regulatory tools. Both the findings of network economics and the experience with different network sectors show that tailor-made bottleneck regulation is the only way. Generally, a distinction has to be made between the existence of network-specific market power due to monopolistic bottlenecks and the question as to whether this market power is transferred to complementary parts of the market. Even if a transfer of market power from a bottleneck to other partial markets were incentive compatible, this would certainly not mean that the bottleneck and the other partial markets belong to the same market. The basic idea of the disaggregated regulatory approach employed in network economics is the very fact that it is possible to distinguish between those parts of the network that constitute bottlenecks and those parts that are characterized by active and potential competition. The all-important task then is to ensure adequate regulation of bottlenecks in order to enable equal opportunities for competition on the other markets. It is well known from the positive theory of regulation that regulators have strong incentives to over-regulate, mix regulatory instruments in an unsuitable way, favor the application of detailed regulation and call for a heavy-handed supervision of firms (e.g. Stigler, 1971; Knieps, 1998). This is the very reason why an a priori “framing” decision to limit the regulatory basis to some extent is of particular importance. This leads to the disaggregated regulatory approach, which not only identifies networkspecific market power properly as monopolistic bottlenecks, but also designs a combination of regulatory instruments limited to the bottleneck (Knieps, 1997a, p. 331). Price-cap regulation limited to monopolistic bottleneck services (wholesale services) combined with accounting separation and technical regulation is sufficient to deal with the problem of non-discriminatory access. Although access regulation cannot be perfect, it moves regulatory attention into the right direction. The aim of future regulatory policy should not be the global regulation of markets. Instead, only a disaggregated regulation of monopolistic bottlenecks is justified. The aim is then to localize the market power in monopolistic bottleneck areas and discipline this market power by regulatory intervention. Asymmetry of market power due to monopolistic bottleneck facilities, however, does not by itself require asymmetric regulation. Instead, the symmetry principle requires that all firms have access to capacities of monopolistic bottlenecks on terms identical to those of the incumbent (non-discriminatory access). The symmetry principle demands that only bottleneck facilities are regulated, irrespective of whether the owner is the incumbent or a newcomer (Shankerman, 1996, p. 5).

68


2.6.5. A critical appraisal of access holidays 2.6.5.1. The avoidance of over-regulation In recent years the focus of regulatory attention has increasingly shifted toward the incentives for investment. From an economic point of view the relation between the pricing of access to monopolistic bottlenecks and its linkage with investment incentives has to be analyzed (e.g. Newbery, 2000; Valletti, 2003). In the European Union (EU) review on telecommunications the access directive of March 2002 indicates the necessity that: “National regulatory authorities shall take into account the investment made by the operator and allow him a reasonable rate of return on adequate capital employed, taking into account the risks involved”.28 The new EU directives on the regulation of electricity and gas networks also focus on transmission and distribution tariffs allowing the necessary investments to ensure the viability of networks.29 Access holidays mean a significant period during which an investor is free from access regulation. The idea is that such holidays will increase investment incentives by allowing profits unhindered by regulatory intervention (Gans and King, 2003, p. 164). Access holidays can only be a relevant concept, if regulatory problems of network-specific market power still exist. With respect to market power two questions have to be considered. Firstly, does a new investment create network-specific market power? If not – as for example in the area of LNG terminal projects – sector-specific regulation is superfluous. From this perspective, Art. 22 of the Directive 2003/55/EC on the exemptions from access regulation seems convincing, because it requires that the exemption is not detrimental to competition. The efficient operation of exemption from regulated third party access, therefore, depends on the regulator’s ability to distinguish between essential facilities and projects subject to competition (Hernández and Gandolfi, 2005, p. 6). Secondly, do new investments phase out the bottleneck nature of the existing network infrastructure? An important example comes from the telecommunications sector. Since the comprehensive opening of the telecommunications market, the pressure of innovation has increased in local networks, too. This has led to considerable variety in technological platforms, for example, optical fibre, wireless networks, CATV networks, satellite technology, and an increase in the variety of network access products. Due to these rapid developments the local loop facilities in bigger cities and agglomerations are increasingly loosing their character of monopolistic bottlenecks. Although it is not possible at this point to predict exactly how long it will take for the monopolistic bottlenecks in the local loop to disappear completely, there cannot be any doubt that the regulation of monopolistic bottlenecks has to be viewed in a dynamic context, so that the potential for phasing out sector-specific regulation in telecommunications can be fully exhausted (e.g. Knieps, 2004). 2.6.5.2. Are access holidays justified? The basic argument in favor to access holidays is the conjectured failure of regulatory contracts. Due to the sequential nature of investment decisions (ex ante) and regulation of 28

Directive 2002/19/EC of the European Parliament and of the Council on access to, and interconnection of, electronic communications networks and associated facilities (Access Directive), OJ L108/7, 24.4. 2002. 29 Directive 2003/54/EC of the European Parliament and of the Council concerning common rules for the internal market in electricity and repealing Directive 96/92/EC, OJ L 176/37, 15.7.2003, Art. 23/2 (a). Directive 2003/55/EC of the European Parliament and of the Council concerning common rules for the internal market in natural gas and repealing Directive 98/30/EC, OJ L 176/57, 15.7.2003, Art. 25/2 (a).


69

access tariffs (ex post), a regulation-induced hold-up problem would arise. The truncation problem would result to reward only ex post-successful projects, whereas the ex ante risks of project failure would not be compensated. The question arises whether access holidays are the adequate answer to this problem. Due to the past dependency of network infrastructures the hypothetical scenario of greenfield approaches of new infrastructure networks overstates the ex ante risks, although the incentive problem for the gradual renewal of bottleneck infrastructures should not be ignored. From an investor’s point of view all relevant ex ante risks should be compensated, including the option value of waiting to invest. The design of credible regulatory contracts focussing on the financial viability of the networks is required (e.g. Knieps, 2005). An important cornerstone of the EU directives is the financial viability of the networks.30 Access tariffs in the electricity and gas networks should allow the necessary investments in the networks.31 In telecommunications national regulatory authorities are obliged to take into account the investment made by the operator and allow him a reasonable rate of return on adequate capital employed, taking into account the risk involved (Access Directive 2002/19/EC, Art. 13(1)). For railroads, full recovery of the infrastructure costs can be accomplished via access charges only, without state funding. Where a contractual agreement for state funding exists, the terms of the contract and the structure of the payment have to be agreed upon in advance to cover the whole of the contract period (Directive 2001/ 14/EC, Art. 6/1, Art. 6/4, Art. 8/1). It can be shown that the problem of regulatory opportunism is not caused by the nature of ex ante irreversible investments per se, but is based on the more general problem that regulatory agencies cannot be committed to welfare-maximizing behavior. Therefore, the regulatory agencies have to be constrained by statutes to allow the compensation of ex ante risks of irreversible investments. To conclude, the instrument of access holidays becomes superfluous. 2.7. Conclusions In order to analyze the role of sector-specific regulation compared to general competition policy in the liberalized energy markets a disaggregated approach is developed. The theory of monopolistic bottlenecks constitutes the theoretical reference point for this analysis in order to identify stable network-specific market power. A survey is devoted to of the localization of monopolistic bottlenecks in different network industries. Finally, the concept of disaggregated price-cap regulation of monopolistic bottlenecks is provided avoiding end-to-end regulation as well as access holidays. Appendix 2.A1. Empirical/Econometric Approaches The traditional approaches to proving with sufficient certainty the existence of stable market power by means of empirical/econometric methods for a certain industry have, to 30

For railway infrastructures see Directive 2001/14/EC, Art. 8; for electricity grids see Directive 2003/54/EC, Art. 19; for natural gas pipelines see Directive 2003/55/EC, Art. 18; for telecommunications networks see Access Directive 2002/19/EC, Art. 10; for slot allocation at Community airports see Council-Regulation (EEC) No. 95/93. 31 Directive 2003/54/EC, Art. 22/2a; Directive 2003/55/EC, Art. 25/2a.

70


date, not developed far enough to serve as a basis for justifying regulatory intervention. Also to be considered in this context is the usual point of reference in a market economy that in markets the 1st order error (“false positive”), that is, the fact that the authorities interfere with the competition process by way of regulatory action even though competition is working and there is no need whatsoever for state action, is particularly grave (Knieps, 1997a, p. 51). The following quotation characterizes the concept of market power established in antitrust literature: “The term ‘Market Power’ refers to the ability of a firm (or a group of firms, acting jointly) to raise price above the competitive level without losing so many sales so rapidly that the price increase is unprofitable and must be rescinded” (Landes and Posner, 1981, p. 937). Stable market power is characterized by the fact that the long-term characteristics of the market under review (particularly the production and demand conditions) allow stable profits without these being competed away (e.g. by arbitrage activities). Quite apart from this are short-term profits which can occur as a result of short-term characteristics of the market under review and are then rapidly competed away by other suppliers. These kind of unstable or fleeting profits are, however, very difficult to differentiate from pioneer profits. The much used but still problematic method of “proving” market power in antitrust cases consists in first defining the relevant market in which the market share of the company accused is to be determined; second, calculating this market share; and third, deciding whether this market share is sufficiently large to allow conclusions to be drawn regarding the necessary extent of market power. However, this method is unsuited as a basis for deriving stable market power and for establishing a regulatory policy in network industries. Market shares are not a reliable criterion for market power.32 This can already be explained by way of the known phenomenon of reversed causality. Instead of being able to impose higher prices because of a high market share, low prices lead to a high market share. In the case of free market access, competitors will enter the market at lower prices with the result that a seller with high prices is undercut and loses market shares. Obviously, it would be incorrect to assume market power without hesitation merely because of a high market share (cf. Landes and Posner, 1981, p. 977). In network industries where economies of scale are frequent, the resulting decrease in average costs augments this effect. Nevertheless, there is the risk that the authorities may use the market share criterion to conclude automatically the existence of a dominant market position. Since the size of the market share cannot be used conclusively to infer the existence of stable market power, there is the question of how far econometric models are capable of assessing whether or to what extent stable market power actually exists in an industry. The methods best known and most often used are the structural models and the “reduced form” methods,33 all of which are in themselves consistent and mathematically correct.34 However, stable market power cannot reliably be localized with any of these model approaches. Moreover, applying 32

The fundamental criticism of the structure–conduct–performance approach of traditional industrial economics is summarized extensively (e.g. in Schmalensee, 1989). 33 A critical assessment of these approaches along with extensive literature on the subject is given in Hyde and Perloff (1995). 34 In a structural model, the econometrician assesses all equations of that model. Ideally one would assess a complete model with separate behavioral equations for each company of the industry. If only aggregated industry level data are available, a demand equation, an aggregated cost equation and a balance condition are estimated. A market power assessment based on these methods is extremely sensitive to small changes in the specification of the structural model equations on which they are based. Structural


71

these different methods to the same industries typically leads to contradictory results, with at least one or even several tests rejecting the competition hypothesis for most of the industries.35 It is obvious that regulatory intervention based on these results of industry-economic research would lead to interventionism, which would be damaging to the competition process. 2.A2. Game Theory Approaches Another methodological approach aimed at unveiling stable market power in industries leads to game theory. Over the last two decades, game theory models have made significant inroads into industrial economics and competition theory-related research. Yet the extensive literature on game theory models (e.g. Tirole, 1989) fails to provide a basis for reliably localizing stable market power in industries. The result is not surprising if one looks at the strengths and weaknesses of the game theory analyzes.36 On the one hand, there are typically a multitude of alternatives for developing a game theory model, all of which appear equally appropriate a priori. On the other hand, there is typically a multitude of solutions within a certain model approach. However, the advantage of having a multitude of possibilities for modeling interaction between alternative suppliers based on game theory proves problematic when it comes to searching for robust solutions. If market power can be localized for a very specific model specification, but then disappears if there is a minor change to any of the model parameters, then this is no reliable basis for ex ante regulatory intervention. The search for robust characteristics of game theory models, which can also be examined empirically, has only just begun (cf. Sutton, 1990). Given the present state of research, it is unsuited as a theoretical basis for localizing stable market power and establishing a regulatory framework.

References Areeda, P. and Hovenkamp, H. (1988). “Essential facility” doctrine? Applications, Antitrust Law, 202.3 (Suppl. 1988), 675–701. Bailey, E.E. (1981). Contestability and the design of regulatory and antitrust policy. American Economic Review, 71(2), 178–183.

(continued) models require more extensive data and more explicit assumptions than the “reduced form” methods. A known “reduced form” method was developed by Panzar and Rosse (1987), whose market power test only requires the assessment of a single equation. The Panzar–Rosse method is easier to apply than the structural model approach. The problem with this method, however, is that the correct “reduced form” revenue function is extremely complicated, not linear, and therefore difficult to assess. In the end, a reliable distinction between collusion and competition is not possible. Another known “reduced form” method was developed by Hall (1988). Hall’s method uses results of comparative statistics to test market power, the zero hypothesis being the competition. The central weakness of the Hall approach is that it can only be used to measure the market power of an industry if one can reliably assume constant returns to scale. The reliability of the results is extremely sensitive to deviations in the constant returns to scale. Both methods differ both with regard to the scope of the data requirements and in the underlying assumptions. 35 “For poultry, butter, cheese, we cannot reject competition based on any of the tests shown in Table IV. For all other industries, one or more tests reject competition or are implausible (Hall estimate of ␮ for red meat and flour). For these other industries, there is little consistency in results across the methods, though, in most cases, the results are consistent with some form or another of oligopoly or monopolistic competition” (Hyde and Perloff, 1995, p. 481). 36 For an illustrative overview, see Fisher (1989).

72


Bailey, E.E. and Baumol, W.J. (1984). Deregulation and the theory of contestable markets. Yale Journal on Regulation, 1, 111–137. Bailey, E.E. and Panzar, J.C. (1981). The contestability of airline markets during the transition to deregulation. Law and Contemporary Problems, 44, 125–145. Bailey, E.E. and Williams, J.R. (1988). Sources of economic rent in the deregulated airline industry. Journal of Law and Economics, 31, 173–202. Bain, J.S. (1956). Barriers to New Competition. Harvard University Press, Cambridge, MA. Baumol, W.J. (1977). On the proper cost test for natural monopolies in a multiproduct industry. American Economic Review, 67, 809–822. Baumol, W.J. (1982). Contestable markets: an uprising in the theory of industry structure. American Economic Review, 72, 1–15. Baumol, W.J. (1996). Predation and the logic of the average variable cost test. Journal of Law and Economics, 39, 49–72. Baumol, W.J. and Willig, R.D. (1986). Contestability: developments since the book. Oxford Economic Papers, Special Supplement, November, 9–36. Baumol, W.J. and Willig R.D. (1999). Competitive rail regulation rules, should price ceilings constrain final products or inputs? Journal of Transportation Economics and Policy, 33(1), 43–54. Baumol, W.J., Panzar, J.C. and Willig, R.D. (1982). Contestable Markets and the Theory of Industry Structure. Harcourt Brace Jovanovich, San Diego. Baumol, W.J., Panzar, J.C. and Willig R.D. (1983). Contestable markets: an uprising in the theory of industry structure: reply. American Economic Review, 73(3), 491–496. Beesley, M.E. and Littlechild, S.C. (1989). The regulation of privatized monopolies in the United Kingdom. Rand Journal of Economics, 20, 454–472. Brunekreeft, G. (2003). Regulation and Competition Policy in the Electricity Market – Economic Analysis and German Experience, Nomos, Baden-Baden. Carlton, D.W. (2004). Why barriers to entry are barriers to understanding. American Economic Review, 94, 466–470. Caves, D.W., Christensen, L.R. and Tretheway, M.W. (1984). Economies of density versus economies of scale: why trunk and local airline costs differ. Rand Journal of Economics, 15(4), 471–489. Chadwick, E. (1859). Results of different principles of legislation and administration in Europe; of competition for the field, as compared with competition within the field, of service. Journal of the Royal Statistical Society, 22, 381–420. Damus, S. (1984). Ramsey pricing by U.S. railroads – can it exist? Journal of Transport Economics and Policy, 18, 51–61. Davies, J.E. (1986). Competition, contestability and the liner shipping industry. Journal of Transport Economics and Policy, September, 299–312. Debreu, G. (1959). Theory of Value: An Axiomatic Analysis of Economic Equilibrium. Yale University Press, New Haven and London. Demsetz, H. (1968). Why regulate utilities? Journal of Law and Economics, 11, 55–65. Fisher, F.M. (1989). Games economists play: a non-cooperative view. Rand Journal of Economics, 20(1), 113–124. Gans, J. and King, S. (2003). Access holidays for network infrastructure investment. Agenda, 10(2), 163–178. Graham, D.R., Kaplan, D.P. and Sibley, D.S. (1983). Efficiency and competition in the airline industry. Bell Journal of Economics, 14, 118–138. Hall, R.E. (1988). The relationship between price and marginal cost in U.S. industry. Journal of Political Economy, 96, 921–947. Hausman, J. (2002). Internet-related services: the results of asymmetric regulation. In: R.W. Crandall and J.H. Alleman (Eds.), Broadband – Should We Regulate High-Speed Internet Access? Brookings Institution Press, Washington DC, pp. 129–156. Hausman, J. and Sidak, J.G. (1999). A consumer-welfare approach to the mandatory unbundling of telecommunications networks. Yale Law Journal, 109, 417–505. Hernández, F. and Gandolfi, M. (2005). EU exemptions to TPA for new gas infrastructures. Energy Regulation Insights, July(24), NERA, London.


73

Hogan, W.W. (1992). Contract networks for electric power transmission. Journal of Regulatory Economics, 4, 211–242. Hyde, C.E. and Perloff, J.M. (1995). Can market power be estimated? Review of Industrial Organization, 10, 465–485. Joskow, P.L. and Klevorick, A.K. (1979). A framework for analyzing predatory pricing policy. Yale Law Journal, 89, 213–270. Knieps, G. (1997a). Phasing out sector-specific regulation in competitive telecommunications. Kyklos, 50(3), 325–339. Knieps, G. (1997b). The concept of open network provision in large technical Systems. EURAS Yearbook of Standardization, 1, 357–369. Knieps, G. (1998). Costing and pricing of interconnection services in a liberalized european telecommunications market. In: Telecommunications Reform in Germany: Lessons and Priorities. American Institute for Contemporary German Studies, Washington DC, pp. 51–73. Knieps, G. (2002a). Does the system of letter conveyance constitute a bottleneck resource? Discussion Paper No. 101, Institute of Transport Networks and Regional Policy, University of Freiburg, Germany. Knieps, G. (2002b). Wettbewerb auf den Ferntransportnetzen der deutschen Gaswirtschaft – Eine netzökonomische Analyse. Zeitschrift für Energiewirtschaft (ZfE), 26(3), 171–180. Knieps, G. (2004a). Information and communication technologies in Germany: is there a remaining role for sector-specific regulations? In: A. Moerke and C. Storz (eds.). Institutions and Learning in New Industries, RoutledgeCurzon, 2006 (forthcoming). Knieps, G. (2005). Telecommunications markets in the stranglehold of EU regulation: on the need for a disaggregated regulatory contract. Journal of Network Industries, 6(2), 75–93. Knieps, G. (2006). Delimiting Regulatory Needs. In: OECD/ECMT Round Table 129, Transport Services: The limits of the (De) regulation, OECD Publication service. Paris, pp. 7–31. Knieps, G. and Brunekreeft, G. (eds.) (2003). Zwischen Regulierung und Wettbewerb: Netzsektoren in Deutschland, 2. Edition. Physica-Verlag, Heidelberg. Knieps, G. and Vogelsang, I. (1982). The sustainability concept under alternative behavioral assumptions. Bell Journal of Economics, 13(1), 234–241. Kuhlmann, A. and Vogelsang, I. (2005). The German electricity sector – finally on the move? CESifo, DICE Report, 3(2), 30–39. Laffont, J.-J. and Tirole, J. (1994). Access pricing and competition. European Economic Review, 38, 1673–1710. Laffont, J.-J. and Tirole, J. (2000). Competition in Telecommunications. The MIT Press Cambridge, Massachusetts, London. Landes, W.M. and Posner, R.A. (1981). Market power in antitrust cases. Harvard Law Review, 94, March, 937–997. Lipsky, A.B. and Sidak, J.G. (1999). Essential facilities. Stanford Law Review, 51, 1187–1249. Mandy, D.M. (2000). Killing the goose that may have laid to the golden egg: only the data know whether sabotage pays. Journal of Regulatory Economics, 17(2), 157–172. McAfee, R.P., Mialon, H.M. and Williams, M.A. (2004). What is a barrier to entry? American Economic Review, 94, 461–465. Morrison, S.A. and Winston, C. (1987). Empirical implications and tests of the contestability hypothesis. Journal of Law and Economics, 30, 53–66. Newbery, D.M. (2000). Privatization, Restructuring, and Regulation of Network Utilities. Cambridge (MA), London. Panzar, J.C. and Rosse, J.N. (1987). Testing for “monopoly” equilibrium. The Journal of Industrial Economics, 35, 443–456. Panzar, J.C. and Willig, R.D. (1977). Free entry and the sustainability of natural monopoly. Bell Journal of Economics, 8, 1–22. Posner, R.A. (1976). Antitrust Law: An Economic Perspective. University of Chicago Press, Chicago. Schmalensee, R. (1989). Inter-industry studies of structure and performance. In: R. Schmalensee and R. Willig (eds.), Handbook of Industrial Organization. North-Holland, Amsterdam, pp. 951–1009. Schmalensee, R. (2004). Sunk costs and antitrust barriers to entry. American Economic Review, 94, 471–475.

74


Schwartz, M. (1986). The nature and scope of contestable theory. Oxford Economic Papers, Special Supplement, November, 37–57. Schwartz, M. and Reynolds, R.J. (1983). Contestable markets: an uprising in the theory of industry structure: comment. American Economic Review, 73(3), 488–490. Shankerman, M. (1996). Symmetric regulation for competitive telecommunications. Information Economics and Policy, 8, 3–23. Stigler, G.J. (1968). Barriers to entry, economies of scale, and firm size. In: G.J. Stigler. The Organization of Industry. Irwin, Homewood, IL, pp. 67–70. Stigler, G.J. (1971). The theory of economic regulation. Bell Journal of Economics, 2, 3–21. Stiglitz, J.E. (1987). Technological change, sunk costs and competition. Brookings Papers on Economic Activity, 3, 883–947. Sutton, J. (1990). Explaining everything, explaining nothing? Game theoretical models in industrial economics. European Economic Review, 34, 505–512. Tirole, J. (1989). The Theory of Industrial Organization, 2nd printing. MIT Press, Cambridge. Valletti, T.M. (2003). The theory of access pricing and its linkage with investment incentives. Telecommunications Policy, 27, 659–675. Weitzman, M.L. (1983). Contestable markets: an uprising in the theory of industry structure: comment. American Economics Review, 73, 486–487. Weizsäcker, C.C. von (1980a). A welfare analysis of barriers to entry. Bell Journal of Economics, 11, 399–420. Weizsäcker, C.C. von (1980b). Barriers to Entry: A Theoretical Treatment, Springer, Berlin. Willig, R.D. (1978). Pareto superior nonlinear outlay schedules. Bell Journal of Economics, 9, 56–69. Willig, R.D. (1980). What can markets control? In: R. Sherman (Ed.), Perspectives on Postal Service Issues. American Enterprise Institute for Public Policy Research, Washington, pp. 137–159.

PART II Trailblazers


Chapter 3 Chile: Where It All Started RICARDO RAINERI Departamento Ingeniería Industrial y de Sistemas, Escuela de Ingeniería, Pontificia Universidad Católica de Chile, Santiago, Chile

Chapter summary This chapter describes the pioneering deregulation and privatization process of the Chilean electric industry, which began in the 1980s. Over the years, Chile has had its share of success and failures, and offers an early example of the difficulties in dealing with three distinct segments of the value chain, namely generation, transmission, and distribution, along with a system operator. In the Chilean context, power generation was left to competitive investors, transmission was turned into an open access regime, and distribution was left as a regulated natural monopoly. The system was put to a test during a severe drought in the late 1990s and subsequent crises, including chronic shortages of natural gas from neighboring Argentina. After 25 years of fine-tuning, the sector still goes under continuous adjustments which respond to a learning process based on what have worked in the past and what have not, and thus it offers useful insights for policy makers and regulators around the world.

3.1. Introduction Chile is generally credited as the place where electricity supply industry (ESI) market reform started. With the enactment of DFL No. 11 in 1982, a new institutional framework for a decentralized and privately owned ESI was introduced. DFL No. 1 recognized three distinct segments: generation, transmission, and distribution. Power generation was considered as a competitive business, transmission was to be governed as an open access regime allowing all generators a non-discriminatory use of available transmission capacity, and distribution was left to be regulated as a natural monopoly. To coordinate the operations of competitive generators in an open access transmission network, an independent system operator – in this case Centro de Despacho Económico de Carga (Economic Load Dispatch Center, CDEC) – was created. These were considered radical ideas at the time, but now a standard practice.

1

General Law of Electric Services, Decree-Law No. 1 of 1982 from the Ministry of Mines (DFL No. 1).

77

78


The law defines three markets where generators can sell power and energy. The first market, between generators and distribution companies, is for small consumers, who pay regulated energy and power prices to their distribution company. The second market is for large consumers, who freely agree electricity supply contracts with generators or distribution companies. And a third market in the CDEC is the one where generators trade power and energy to fulfill their electricity supply contracts, in which case the prices of the energy transfer between the companies are decided by the CDEC according to the system’s marginal cost of generation, based on generators’ declared costs, while power transfers are valued according to the National Energy Commission (Comisión Nacional de Energía, CNE) regulated capacity price paid by small consumers in their electricity tariffs. Not only does Chile lead the world in the ESI market reform, but it also was among the first to experience natural and man-made crises with serious consequences. The first one, in 1998–1999, was caused by one of the worst droughts experienced by the country which affected a predominantly hydro-based system. The second, in 1999, resulted from an incomplete regulatory framework which caused a series of blackouts by the lack of coordination among power generators. And the third one, started in 2004 and continuing, resulted from political and economic phenomena in neighboring countries which severely affects a reliable supply of imported natural gas from Argentina, essential for an increasing natural gasdependent ESI in Chile. Other well known electricity crisis in a restructed electricity market is the one that suffered California (see Joskow, 2001a and 2001b). This chapter describes the fundamentals leading to the deregulation and privatization process of the ESI in Chile, the key crises that through these years have tested its normal operation, the policies and politics that have been adopted to address the problems, and the lessons we have learned from the whole process. Having been the first and battered a lot, Chile offers many useful insights for policy makers and regulators around the world. This chapter is organized into four sections that deal with the Chilean ESI. After the Introduction, Section 3.2 introduces the Chilean ESI, Section 3.3 describes the electric crises which affected the Chilean ESI, and Section 3.4 provides the final comments and key lessons which can be learned from the Chilean experience. 3.2. The Chilean ESI Chile is a thin and long country with a 756,950 km2 territory and a population of almost 16 million. In 2004, Chile had a total electricity consumption of 46,114 GWh and an installed generation capacity of 11,561.4 MW. Its ESI consists of four distinct interconnected systems (Fig. 3.1 and Table 3.1) that are isolated from each other: ●

●

Sistema Interconectado del Norte Grande (Greater North Interconnected System, SING) in the north. Sistema Interconectado Central (Central Interconnected System, SIC) covering the central part of the country including the country’s capital city, Santiago.

And two smaller systems in the south: ● ●

Sistema de Aysen (Aysen System). Sistema de Magallanes (Magallanes System).

The SING and the SIC account for the lion’s share of the installed generation and consumption, with the other two systems serving much smaller populations, as shown in Table 3.1.

Chile: Where It All Started

79

Fig. 3.1. ESI systems in chile. Source: Picture elaborated by the author with CNE data.

3.2.1. Key characteristics For the most part, this chapter is focused on the two large systems, the SIC and the SING, which are briefly described below. 3.2.1.1. Sistema Interconectado del Norte Grande The SING is a rather unusual system consisting of a predominantly mining/industrial load, in the order of 90%. It serves a number of large and isolated copper mines, some with individual loads as large as 300 MW. The ESI is mostly based on thermal generation located on the coastal edge, which experienced a significant growth between 1993 and 2004

80


Table 3.1. Electric systems in Chile (2004).

Installed capacity (MW)

Transmission Peak load lines (km) (MW)

% of population and consumption attributed to small users (residential, Demand commercial and (GWh) industrial)

SING

3596*

4889*

1567

11,240

5.6% of the 99.6% national population, thermal which represents 0.4% hydraulic 10% of the system’s consumption

SIC

7975*

8745

5431

34,602

93% of the national population, which represents 60% of the system’s consumption

41% thermal 59% hydraulic

Sistema de Aysen

34

–

18

82

20,000 customers

44% thermal 50% hydraulic 6% wind

Sistema de Magallanes

65

8.5

33

189

46,000 customers

100% thermal

Type of generation

*April 2005.

(Table 3.2). Today it counts with three distribution companies and three large independent generators. By international standards Chilean ESI market concentration appears high, far above for example to what happen in Argentina (see in this book the chapter by Isaac Dyner et al. “Understanding The Argentinean and colombian Electricity Markets”). The technological innovation in natural gas power plants, that remarkably increased their efficiency during the last two decades, and the increasing availability in the region of natural gas, together with the chances in the late 1990s to import natural gas from Argentina, makes possible the substitution of coal and petroleum power plants for natural gas combined cycle turbines. Given that natural gas power plants have smaller variable costs, they become the base of generation in the SING. The large investments in natural gas power plants implied that the SING installed capacity increased from 1277 MW in 1997 to 3596 MW in 2004, albeit peak load represents only 40% of the installed capacity. Figure 3.2 shows the growth that the installed capacity experienced with respect to the electricity consumption for the SIC and the SING. Although some of the new natural gas power plants enter the market without electricity supply contracts with the large mining companies or distribution companies, in spite of everything they were able to assure enough revenues to finance their investments. This is because the power plants’ dispatch is made according to the lowest variable cost, independently from the generators’ electricity supply contracts, which assures that more efficient power plants can sell their electricity at least at system’s marginal cost.2 2

Generators received energy and a capacity payment. The energy payment is determined by the power plant dispatched with the higher variable costs, and the capacity payment is defined by a function which considers the availability of the power plants and a regulated capacity price.

81

Chile: Where It All Started Table 3.2. SING. Thermal

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Sales (GWh)

Annual growth (%)

Capacity (MW)

Total (MW)

%

3040 3394 3989 4981 5749 6616 8120 8398 8991 9482 10,480 11,240

11.66 17.53 24.87 15.41 15.09 22.72 3.43 7.06 5.46 10.53 7.25

745 799 1157 1160 1277 1476 2637 3041 3441 3633 3641 3596

735 789 1144 1147 1264 1463 2624 3028 3428 3620 3628 3583

98.7 98.7 98.9 98.9 99.0 99.1 99.5 99.6 99.6 99.6 99.6 99.6

%

MW

%

Peak load (MW)

38 50 56 58 58 59

10 10 13 13 13 13 13 13 13 13 13 13

1.3 1.3 1.1 1.1 1.0 0.9 0.5 0.4 0.4 0.4 0.4 0.4

498 611 747 812 1021 1094 1154 1221 1360 1416 1567

Natural gas MW

1004 1519 1919 2112 2112 2112

Hydraulic

Source: CDEC SING, CNE.

3.2.1.2. Sistema Interconectado Central The SIC is the largest electric system in the country with 31 distribution companies, where 60% of its consumption corresponds to small regulated end users. In the period from 1990 to 2004, the SIC electricity demand almost tripled, from 12,512 to 34,602 GWh (Table 3.3 and Fig. 3.2), and installed capacity doubled, from 3195 to 7867 MW, where peak load have fluctuated between 62% and 79% of the installed capacity. Today the SIC have three large independent generators and few smaller companies or co-generators. Until 1996 the SIC depended on hydro generation which represented 75% of installing capacity, implying that electricity supply depends on what happen with hydrologic conditions. This situation changed somehow in 1997 when, motivated by private investors, the country starts importing natural gas from Argentina for the new power plants. With that and up to year 2004, the share of hydro capacity on total capacity decreased to 60%, while the share of natural gas power plants increased from 0% to 23%. Figure 3.2 shows that, contrary to that happened in the SING, the growth in generation capacity in the SIC has been more modest, and sometimes it lagged behind the growth in electricity consumption. Despite the rapid growth of thermal generation, the SIC still remains as a predominantly hydro-based system (Fig. 3.3), during the period 1985–2004 the average contribution of hydro generation to total generation is more than 75%. During unusually wet seasons, this figure could rise to 97% as in 1992, but it could be as low as 48% as in 1999 (Fig. 3.4). The volatility of hydrologic conditions is among the factors that complicate the Chilean ESI – not unlike other hydro-dominated systems such as the Nordic countries, Brazil, and Colombia. 3.2.2. Factors leading to the liberalization of the ESI The period between 1973 and 1990, the military government of Augusto Pinochet, inspired free market policies in Chile, leading to the reform of the ESI. Much of this was driven

82


550 500

SING capacity and demand growth index (SING 1993 ⫽ 100) SING demand SING capacity

450 400 350 300 250 200 150 100 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 (a) SIC capacity and demand growth index (SIC 1990 ⫽ 100) 280 260

SIC demand SIC capacity

240 220 200 180 160 140 120 100 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 (b) Fig. 3.2. SING and SIC capacity and demand growth indexes. Source: CNE, CDEC SING and CDEC SIC.

by ideologies that promoted competition through the active participation of the private sector in the economy, with the government retaining only a subsidiary role. To understand what happened during this period, one must put the ESI reform in a historic context. The change in the economic policy orientation during the military government reversed the previous 40-year trend of increasing government’s intervention in economic activities. It was during Allende’s Government (1970–1973) that State’s intervention in economic activity reached a maximum, where 100% of public services companies were owned by the government,3 with end users’ tariffs subsidized and the companies’ employment policy highly influenced by the political preference of the worker.

3

See Hachette and Lüders (1994).

83

Chile: Where It All Started Table 3.3. SIC. Thermal

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Sales (GWh)

Annual growth (%)

Capacity Total (MW) (MW)

%

MW

%

MW

%

Peak load (MW)

12,365 12,516 13,811 15,272 16,549 17,672 19,027 20,869 22,434 24,246 25,530 27,654 29,144 30,335 32,076 34,602

1.2 10.35 10.58 8.36 6.79 7.67 9.68 7.50 8.08 5.30 8.32 5.39 4.08 5.74 7.88

2942.1 3195.1 3831.1 3831.1 3889.9 3893.4 4083.6 4858.5 5266.8 6242.4 6695.1 6652.8 6579.2 6737.2 6996.2 7867.4

0.25 0.27 0.22 0.22 0.20 0.19 0.22 0.25 0.30 0.38 0.42 0.39 0.39 0.40 0.42 0.40

– – – – – – – – 379 1119 1119 1359 1359 1467 1467 1847

– – – – – – – – 0.07 0.18 0.17 0.20 0.21 0.22 0.21 0.23

2221 2334 2994 2994 3126 3153 3176 3667 3705 3892 3906 4030 4030 4055 4055 4695

0.75 0.73 0.78 0.78 0.80 0.81 0.78 0.75 0.70 0.62 0.58 0.61 0.61 0.60 0.58 0.60

2270 2273 2375 2632 2819 3070 3235 3497 3773 3991.4 4185.5 4516 4694 4878 5162 5430.8

721 861 837 837 764 741 908 1192 1562 2351 2789 2623 2549 2682 2941 3172

Natural gas

Hydraulic

Source: CDEC SIC, CNE. SIC installed capacity by source 5000 4500

Thermal Hydraulic

4000 3500

MW

3000 2500 2000 1500 1000 500 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Fig. 3.3. SIC installed capacity by source. Source: CDEC SIC and CNE.

After the military coup of Augusto Pinochet in September 1973, the Junta focused on reforming the ESI by introducing tariffs that allowed the firms to cover their legitimate costs while encouraging reductions in the number of employees to what can be considered efficient given their production level. At that time, pricing in the ESI was governed by DFL No. 4, enacted in 1959, by which tariffs were based on historical costs. By the end of the 1970s, the

84

Electricity Market Reform SIC generation by source

100.0 9

6

3

5

90.0

7

17

18

13

28

34

80.0

Hydraulic Thermal

14 25 32

40

30

37

41

35 43

52

70.0 60.0 50.0 91

40.0 30.0

94

97

95 84

82

93

87

86 72

66 60

76 68 63

59

71

65 57

48

20.0 10.0 0.0

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fig. 3.4. SIC generation by source. Source: CDEC SIC and CNE.

previous problems were solved to some extent, but some other important problems that needed further changes persisted:4 ● ●

●

●

Still a strong involvement of the State in the development of the industry.5 Increasing monopolization of the industry and its development in one State company, Empresa Nacional de Electricidad S.A. (ENDESA),6 damaging the likelihood to have a competitive industry and implying almost no chances for the government to be an effective counterbalance in the industry. Confusing entrepreneurial and regulatory roles of the State, with private investors at a disadvantage with respect to State companies. Discriminatory tariffs based on historical costs plus a rate of return, lacking from economic, uniform, and transparent procedures.

Until 1978 Chilean electric industry regulation was partially a responsibility of ENDESA; and because this was in conflict with the military government objectives, it created two agencies to take on the regulatory and supervisory roles of the State: ●

4

In 1978 the National Energy Commission (CNE): in-charge of the planning and policy design for the efficient operation of the energy sector and to advise the government on all the aspects of the industry.

See Comisión Nacional de Energía (1989). In 1979, the State participation in the electric sector reached 90% in generation, 100% in transmission and 80% in distribution, demanding more than US$ 200 million of public investments per year to satisfy demand increases. 6 ENDESA was founded by the government in 1943 under the custody of Corporación de Fomento de la Producción (CORFO). CORFO founded in 1939 was originally conceived as a State Development Bank to assist the private sector. But shortly CORFO became the vehicle by which the State becomes involved in the different areas of the economy. CORFO sheltered State companies in the areas of: electricity, telecommunications, chemistry, pulp and forestry, fishing, sugar, coal, computer services, and many others. ENDESA is the result of a national electrification plan developed in 1935 and endorsed in 1936 by the Chilean Engineers Institute. 5


●

85

In 1985 the Superintendence of Electricity and Fuels (Superintendencia de Electricidad y Combustibles, SEC): with a supervisory role, and the inspection and security of electric facilities.7

For the definition of regulated prices the responsibility was assigned to the Ministry of Economy, which considers the cost studies of an efficient model company made by CNE. The military government core reform imperatives were the promotion of higher efficiencies through open competition in generation and regulating the distribution tariffs. These ideas have been included in the electric law DFL No. 1 of 1982, and still embody that spirit. Table 3.4 summarizes the major milestones of the Chilean ESI reform since the late 1970s. 3.2.2.1. System operator The final objective of interconnected electric systems is to minimize the global cost of electricity supply by combining loads and generators of different nature, to arbitrage the excess of electricity supply within the different nodes, and sharing the margins of reserve capacity within all the nodes. In Chile, the system operator, CDEC, has been assigned the role to coordinate the operation of interconnected generation and transmission facilities. The SIC (CDEC-SIC) and the SING (CDEC-SING) system operators were created in 1985 and 1993, respectively. The law requires that the concessionaries of interconnected electric systems should coordinate their operation to: ● ● ●

preserve service quality; guarantee the less expensive operation of the system; and guarantee the right way of main transmission and sub-transmission systems established through a concession.8

The governance structure of the CDEC consists of a Board with representatives from generators, transmission companies, and co-generators. This Board is responsible for agreeing the bylaws that complement the law, required to accomplish the objectives of the CDEC. The Board is complemented by an Executive Direction9 which runs the system and determines electricity transfers and prices between the generators. With the existent generation–transmission facilities, the CDEC is responsible for its operation. The different characters of the SIC and the SING – the first with a large hydro capacity by which it can accumulate water from one year to the next, while the second an almost a 100% thermal system – have implied that the tools used for their planning and operation slightly differ but are inspired in the same principles of minimizing the long-term cost of electricity supply, given existent generation–transmission facilities and service reliability requirements. Each system dispatch is made independently of the ownership and the electricity supply contracts of each of the generators. Electricity transfers between generators result from the accounting of the efficient dispatch of the system and the supply contracts of each of the generators, where energy pricing is based on the cost of the less efficient unit being dispatched and capacity pricing is defined by CNE. 7

SEC was created in 1985 with Law No. 18,410 to replace the Superintendence of Electric Services, Gas and Telecommunications. Also, in 1999 with Law No. 19,613, SEC attributions are changed allowing it to impose stricter regulations and sanctions on the electric companies. 8 Law No. 19,940, March 2004, changed this article to “Guarantee the open access of main transmission (backbone) and sub-transmission systems, in compliance with this law”. 9 Divided in an Operations Direction and a Tolls Direction.

86


Table 3.4. Major events in Chilean electric industry. Event

Year

Comments

Creation of the CNE, Decree-Law No. 2,224

1978

A specialized institution to take on the regulatory and advisory roles on energy issues. The idea is to separate the regulatory, supervisory and entrepreneurial roles of the government.

Nodal prices

1980

First tariff decree to set nodal prices at the level of generation, transmission and distribution.

DFL No. 1

1982

The main reform is embodied in the General Law of Electric Services.

Creation of the Superintendence of Electricity and Fuels (SEC), Law No. 18,410

1985

Replaces the Superintendence of Electric Services, Gas and Telecommunications, to take on the supervisory roles of the government in the electric and fuel industry.

Decree-Law No. 11 of the Ministry of Economy

1984

Define fines and sanctions for electric companies.

Decree-Supreme No. 6 of the Ministry of Economy

1985

Define the norms to coordinate the operation of interconnected electric systems. Creates the CDEC and the rules for transmission pricing, rules modified later in 1990 and 2004.

Privatization

1986

Starts an extensive privatization process by which today almost 100% of the electric industry, generation, transmission, and distribution, are in private hands.

Decree-Law No. 119 of the Ministry of Economy

1989

Update fines and sanctions for electric companies.

Law No. 18,922 amends DFL No. 1 (Decree-Supreme No. 6) regarding electricity transmission pricing

1990

Improves the mechanism for electricity transmission pricing, introducing the concepts of area of influence, and a use and a capacity charge (tariff revenue and basic and additional tolls).

Decree-Law No. 327 of the Ministry of Mines that replaces Decree-Supreme No. 6 of the Ministry of Economy

1998

Improves the definition and exigencies of service quality and reliability, and the CDEC governance structure.

SIC and SING crisis

1999

Between 1998 and 1999 the SIC suffered a drought with a severe impact on hydraulic generation which has implied electricity rationing at the household level. In 1999 the SING suffered from severe blackouts because of the deficiencies in the coordination of the system.

Law No. 19,613

1999

Due to the severe drought that affected the SIC, the authority increases the control that SEC commands on the electric companies. Also, the droughts that happen to be more severe than the ones considered in the calculation of regulated prices are excluded to be considered as a cause of Force Majeure. (Continued )


87

Table 3.4. (Continued) Event

Year

Comments

Resolution from the Ministry of Economy R.M. No. 88

2001

Generators are forced to supply all the electricity required by distribution companies at a price equal to the node price. The last independently of the generators’ electricity supply contracts with the distribution companies.

Change to DFL No. 1 by Law No. 19,940

2004

Changes in regulation to incentive additional investments in electricity transmission. Also changes the criteria to define regulated nodal prices; is enlarged the market for free contracts; is defined a mechanism to set network access charges; and are explicitly recognized the ancillary services within the law.

Natural gas crisis

2004 onward

Economic crisis in Argentina implied natural gas export constraints to Chile, which left the Chilean ESI operating at the limit of its capabilities.

Changes to DFL No. 1 by Law No. 20,018

2005

As a reaction to Argentinean natural gas exports constraints to Chile, regulation is changed to incentive additional investments, and to increase the reliability and flexibility of the electric system.

Main concerns with respect to the CDEC refer to: ●

●

●

●

●

the dichotomy that exists between the CDEC’s planning and dispatch of the generation– transmission system, that is made independently of the ownership and the generators’ electricity supply contracts; the collective responsibility that the law consign to the generators and the transmission companies while being required to operate their facilities according to the CDEC instructions; the large number of conflicts that existed between generators and transmission companies with respect to pricing, and energy and capacity transfers; the deficient information that exists with respect to the CDEC’s criteria to dispatch the electric system; and the influence that some agents can exert on CDEC’s decisions.

So far some progress has been made to solve some of these issues. First, in 2004, is the novel conflict resolution mechanism specially created for the electric industry. It consists of an independent board with seven members, who sanction on many of the conflicts that emerge within the electric companies and/or the authority.10 Second, and also in 2004, is the change in the criteria to define electricity transmission prices, from agreed tariffs to regulated tariffs. And third, in 2005, is the enhanced independence of the system dispatch and its operation, achieved by increasing the CDEC Board voting quorum required to remove the Executive Director. 10

In Law No. 19,940, the board is known as the Electric Experts Panel, “Panel de Expertos Eléctricos”, which is elected by the Chilean Free Competition Court.

88


3.2.2.2. Pricing For end users, the Chilean electricity price system is based on a tariff system designed by the CNE in 1980, before the reform, and formalized in 1982 by the new law, DFL No. 1. Since 1980 the authority recognized a market for small and one for large consumers, where in the beginning small consumers were identified as the ones with a load of at the most 4 MW. Later, the 4-MW threshold level is reduced twice, first to 2 MW in 1982 with the approval of the DFL No. 1, and then to 0.5 MW in 2004 with Law No. 19,940.11 Small end users buy electricity at regulated energy and power prices, while large end users are allowed to freely negotiate their electricity supply contracts. Final prices for small end users result from the addition of different charges: power (capacity or peak load), energy, transmission, and distribution; and more recently, ancillary services.12 Some of the costs are borne by generators and the others by distributors and end users through specially designed tariff formulas. Generators can sell electricity in three different markets: ●

●

●

A generators’ market which takes place in the CDEC, where generators trade to complete their electricity supply contracts. A market for large users who freely negotiate energy and peak load or capacity prices with generators and distribution companies, generally through long-term contracts. A third market is for small users, where distribution companies buy generators’ electricity and sell it to small consumers at regulated energy and power nodal prices determined by the CNE.

Transfer prices between generators. Generators trade energy and capacity in the CDEC. Energy transfers that result from the coordination of the system made by the CDEC are valued at the marginal costs of the system, calculated by the CDEC on the information declared by the generators. The CDEC’ marginal cost is built upon the marginal cost of the less efficient generator being dispatched. For transfers of peak power between the companies, the transfers are made at the power cost calculated by CNE every 6 months in the “Indicative Works Plan for Generation and Transmission”, by which it determines the least expensive generation–transmission expansion plan for the system.13 It happens that the CDEC determines the operation of the system independently of the ownership and the generators’ electricity supply contracts, which implied that generators’ energy and power transfers are cleared after they have signed their electricity supply contracts with large end users, distribution companies, or other generators. Thus, a generator

11

With these consecutive reductions in the threshold level, the number of end users eligible to bid their electricity supply contracts has been increased. 12 Law No. 19,940, March 2004. 13 Given a demand forecast for the following 10 years and accounting for existent and in-construction generation–transmission facilities, the CNE’s “Indicative Works Plan for Generation and Transmission” determines every 6 months the least expensive generation–transmission expansion plan. Before 1997, each generator was required to satisfy in advance an energy balance between its contracted energy and its firm energy, firm energy which is determined based on the generator’s installed capacity and expected available energy. In the energy balance the sum of the contribution of energy from a power plant and energy purchase contracts with other generating companies constitute the company’s firm energy. In 1997 the concept of firm energy that previously existed in the 1985 Decree-Law No. 6 disappeared, where in advance only a global energy balance should be satisfied by the system. Power transfers are determined based on a company’s peak power demand and its firm power supply, where firm power surplus are mandatory traded. Decree-Law No. 327.


89

faces a risky future with no guarantee of being dispatch, as well as on the system’s marginal cost or capacity price at which the CDEC will clear the market. Here the computational model that determines the marginal cost of the system and the power plants’ dispatch comes as a fundamental piece of the puzzle;14 and during electric crisis or in circumstances when the system’s marginal cost abruptly increases controversies on the model and its parameter assumptions have provoked.15 Competitive prices for large consumers. By law, large consumers with a load above 0.5 MW contract electricity at prices, quantities, and conditions freely negotiated with the generators and distributing companies. The prices determined in these free contracts later play a key role in the small end users’ price; it is because by law they define an anchor for regulated energy prices for small end users, where small end users’ prices cannot differ more than 10% from the free prices.16 The main discrepancies with respect to free electricity supply contracts and pricing between large users and generators refer to service reliability and the responsibility of the different agents to ensure a secured electricity supply. Situations have been critical during energy crisis when the large end users have demanded from the generators and the transmission

14

In the 1980s and 1990s, the CDEC-SIC dispatch of the power plants used a model known as GOL (Gestión Optima del Laja), based on a marginal cost pricing criteria for electricity generation. The GOL model was inspired in Marcel Boiteux works what can be named as the “Electricité de France School” (see Boiteux, 1949, 1956, 1960); and also the GOL model follows the Jenkins and Joy (1974) model. Given water inventories, the expected hydrologies, hydro and thermal generation units, fuel and electricity rationing or failure costs, and projected demand, the GOL model minimizes the expected total costs, operation and rationing costs, to serve energy demand for a time period of 10 years. Until September 1991 the GOL model with quarterly periods was used to price energy transfers between generators. In October 1991 the model was replaced by the OMSIC model (OMSIC from its Spanish acronym for Operación Mensual del SIC, SIC’s monthly operation). This is a model with monthly periods that follows similar principles than the GOL model but that is specifically adapted to model the short-term operation of the system. More recently, the CDEC has replaced the OMSIC model used in the medium-term planning by the more sophisticated multi-dam multi-node model known as PLP. The CDEC-SING uses a comparable model, but without the complexity of having to model a hydro thermal system with an inter-annual dam. Rationing cost is understood to be the cost incurred in kWh, in average, by the end users when suffering from an energy shortage, and the energy shortage should be supplied with emergency generators, if it is so agreed. This rationing cost is calculated as a unique value, representative of the more frequent deficits than can show up in the electric system. 15 For example, controversies with respect to the criteria that are used to distribute capacity payments between generators: What is the contribution of each generator to supply energy during peak hours? What is the number of peak hours during which a power plant is eligible to receive a capacity payment? Should hydraulic generators receive capacity payments if they run out of water and cannot contribute with capacity? Should combined cycle natural gas power plants receive capacity payments if they run out of natural gas and cannot contribute with capacity? Is the capacity payment the right signal to assure enough peak load capacity for the system or it is given by the failure cost that the generators must paid to end users when they cannot satisfy the electricity demand? Of these questions, the number of peak hours that should be accounted for capacity payments was the first controversy solved by the Electric Expert Panel, “Panel de Expertos Eléctricos”, which was created in 2004. It define that for wintertime the number of peak hours that should be accounted for capacity payments is eight. 16 In 1982, DFL No. 1 circumscribes small end users’ prices to deviate no more than 10% from the free price. Afterward this band was changed twice. First in March 2004, Law No. 19,940 decreased the band from 10% to 5% and secondly in 2005, Law No. 20,018, the band changes in the opposite direction, by which the regulated energy price can differ from the free price up to 30%.

90


companies for additional investments and backup capacity; while the generators and transmission companies have demanded from the large end users, the installation of additional low-frequency (LF) relays to disconnect large blocks of consumption if a critical condition of operation in the electric system emerges. Regulated nodal prices. The DFL No. 1 established that, in electric systems whose installed generation capacity is above 1500 kW, the generation and transmission prices would be regulated when they supply electricity to small consumers or to distribution companies in the proportion that corresponds to small end users. The regulated nodal prices at the generation and transmission level are determined every 6 months by the CNE in April and October in the “Indicative Works Plan for Generation and Transmission”. This calculation is made based on a demand forecast of peak power and energy for the following 10 years, taking into consideration existent facilities and the ones in construction. The computational model is stochastic over a minimum of 40 hydrologies that until 1998 correspond to the 1941–1980 period. Under these set of assumption happened the worse hydrology corresponded to the 1968–1969 hydrologic year which later was exceeded by the drought that affected the country in 1998–1999.17 With these inputs, CNE determines the outcome that minimizes the present value of the expected cost of total supply, including expected rationing costs during the period of study. The basic energy price is determined as an average of the marginal cost of energy over the next 48 months of operation. The basic capacity price arises from the incremental cost of the more efficient generating units to provide additional power during the peak hours of the annual demand, increased in a percentage equal to the system’s theoretical reserve power margin calculated by CNE. The theoretical reserve power margin changes from time to time, and in April 2005, it was 11.76% for the SIC and the SING. Next and to account for energy and power prices for each node of the transmission network, the authority uses penalty factors which account for the energy and power losses of electricity transmission within the system. Finally, the regulated energy node price is checked against the competitive energy prices determined among generators and large end users, where previously calculated energy prices cannot differ more than 10% from the prices freely determined in the contracts between the companies and large clients. If the difference between the regulated prices and the free prices is greater than 10%, the CNE must multiply all the regulated node prices by a unique coefficient, higher or lower, to take the regulated prices within the band of 10% around the free prices. The CNE pricing model has been criticized because the parameters and assumptions do not reflect the conditions that frequently affect the industry. For example, in the late 1990s lack of investments in the SIC is attributed to the systematic reduction of nodal prices and the price rigidity has shown its inadequacy to signal the electricity crises. Both may help to explain the generators’ lack of interest to sign electricity supply contracts with distribution companies. Electricity prices. The path of regulated energy and capacity prices, and the CDEC spot marginal cost for the SIC and the SING are shown in Figures 3.5 and 3.6. 17

See Raineri and Ríos (1998); and Raineri (2000). The hydrologic data required was modified in 1999 with by Law No. 327 which requires a minimum of 40 hydrologic years, where on the April 2005 CNE Nodal Prices calculation, it used a total of 43 hydrologic years. The three additional hydrologies included in the model are: the first which roughly has 80% of the rain recorded in the 1968–1969 hydrologic year; the second with 90% of the rain recorded in the 1998–1999 hydrologic year; and the third that is more humid than the average sample, being determined in a way to keep the average of the hydrologic years sample rather constant.

91

Chile: Where It All Started SIC–SING energy price and CDEC marginal cost (million $ kWh, monthly average in April and October ) 160 140 120 100 80 60 40 20 0

October 2004

October 2003

October 2002

October 2001

October 2000

October 1999

October 1998

October 1997

October 1996

October 1995

October 1994

October 1993

October 1992

October 1991

October 1990

October 1989

October 1988

October 1987

October 1986

October 1985

October 1984

October 1983

October 1982

Energy SIC Energy SING MgC SIC MgC SING

Fig. 3.5. SIC and SING energy price and CDEC marginal cost. Source: CDEC, SIC, CDEC SING and CNE.

SIC–SING power price ($ kW month, monthly average in April and October) 18 Power SIC Power SING

16 14 12 10 8 6

October 2004

October 2003

October 2002

October 2001

October 2000

October 1999

October 1998

October 1997

October 1996

October 1995

October 1994

October 1993

October 1992

October 1991

October 1990

October 1989

October 1988

October 1987

October 1986

October 1985

October 1984

October 1983

October 1982

4

Fig. 3.6. SIC and SING power price. Source: CDEC SIC, CDEC SING and CNE.

In Figure 3.5 we see that since the 1980s exists a decreasing trend in the regulated energy price, decreasing trend which is steeper in the SING than in the SIC, where the regulated energy prices have come closer to the CDEC’s marginal cost. In the case of the SIC, the marginal cost depends on the hydrologic conditions, whereas in the case of dry conditions or a severe drought the system is forced to dispatch the more expensive thermal units. This explains the sudden increases in the system’s marginal cost for the 1996–1997 and 1998–1999 periods. In the case of the SING, the system’s marginal cost depends on the variable cost of its thermal power plants, which until the late 1990s depended on oil and coal prices, and since 1999 it also depends on the less expensive imported natural gas.

92


The idea of having stable end users’ prices was originally conceived in the DFL No. 1, for that we have, from Figure 3.5, that regulated energy nodal prices have followed a smooth and decreasing trend coming closer to the path that may well be predicted by the system’s dispatch marginal cost, which can be considered as the long-term average energy price. But besides the original intention of the authority, it happens that afterward and from the producers’ side, the regulated nodal energy prices resulted insensible to the short-term supply conditions that affect the systems; being a useless instrument to adjust supply and demand, a critical condition of operation happens in the system. The SIC and the SING regulated capacity prices are depicted in Figure 3.6. Conceptually capacity prices should be determined by the marginal cost of increasing the installed capacity, the marginal turbine which is determined by standard or novel technologies as was the case of the new natural gas power plants in the late 1990s. In both systems, the SIC and the SING, after the entry of natural gas, the capacity price decreased. But exist other events which can affect the regulated capacity price: the interconnection of isolated electric systems as it happens in the SING in the late 1980s; and changes in the theoretical reserve margin determined by the authority. The CDEC’s marginal costs of energy is shown in Figure 3.5, are monthly average, but the CDEC calculates hourly marginal costs, meaning that actual cost fluctuations are even larger, and are larger than the regulated energy price fluctuations determined by the biennial calculation of the CNE. After all and under the new regulatory framework which boosted competition between generating companies, in Chile we have decrease in electricity prices. To these also contributed the fine-tuning and learning process of the new regulatory model that followed its introduction since 1982. Chilean regulations have tried to promote private decisions based on economic signals through prices. Consequently, the developments of the electric and natural gas industry have been made by private agents, attracted in a competitive way toward the investment opportunities offered by the market. The new regulatory model and privatization of the most important companies at the end of the 1980s make possible large private investments in generation. It is in this new competitive atmosphere where the companies began to invest in thermal generation to meet the increasing demand of energy as well as the new environmental constraints. Based on these facts is that in the early 1990s the chances to import natural gas from Argentina emerged as the solution to the country’s energy needs. Electricity distribution pricing. The 36 distribution companies that exist in Chile have a nonexclusive public service concession over a geographic area and are mandated to serve the electricity demand at a regulated price called distribution value added or VAD. The VAD is a multiple-part tariff for using the distribution infrastructure and it is formed by: a fixed fee for managing, billing, and servicing the consumer; average energy and power losses; and a fixed fee per unit of power to pay for the operation, maintenance and investment costs. In VAD estimates, annual investment costs are calculated considering the replacement cost of efficient facilities for a projected demand, the facilities’ lifetime, and a 10% real return on assets. VAD is determined every 4 years for each distribution company based on a yardstick competition model where tariffs are defined according to an efficient model company that distributes electricity in the distribution company’s geographic area. Each distribution company’s VAD is determined by a cost study made by the CNE. The distribution company can contrast the CNE study with its own study, in which case VAD parameters result by weighting the CNE study on two-thirds and one-third for the company study. With the tariffs thus determined and the regulated nodal prices, the consistency of the distribution tariffs is


93

revised for all distribution companies, where on the aggregate the average return on assets should be within a band of 10 ⫾4%. If the average return on assets is outside this band, tariffs should be adjusted such that the average return for all distribution facilities reaches the closer upper or lower limit of the band.18 Next, electricity price for small end users with a peak load of at most 0.5 MW is obtained by adding to the different VAD components, the regulated energy and power prices.19 The main problems experienced with electricity distribution pricing are as follows: ●

●

●

First, the weighting of the results, two-thirds for the parameter values obtained in the CNE VAD study and one-third for the parameter values obtained in the distribution company study, has a perverse incentive to create artificial costs structures which tend to diverge in the cost structure of the efficient model company. For example, past tariff processes have taken to differences in parameter values beyond 100% or even 200%.20 Second, the discount rate used to calculate tariffs is a fixed 10% real discount rate which may not reflect the opportunity cost of investments, and it should be replaced by one that reflects the opportunity cost of investments, as is currently done in other regulated industries in Chile where it is used as the capital asset pricing model (CAPM). Until 2004 electricity distribution pricing lacked a clear mechanism to set network access fees (access charges) for other agents who wants to sell electricity to large end users located in the distribution companies networks. Thus, only a few electricity supply contracts for large users located in the distribution companies’ networks have been served by a company different to the distribution company. This situation is expected to change with the new norms approved in 2004 that provide a clearer procedure to define network access fees.21

Electricity transmission pricing. The SIC developed to transport electricity from the hydro power plants located typically in the south of the country to the large consumption centers located in the center of the country;22 and the SING to connect the large mining 18

See Schleifer (1985) and Rudnick and Raineri (1997a). Before the regulatory change of 2004, Law No. 19,940, electricity transmission was paid by generators, where no explicit electricity transmission charge existed for small end users. Law No. 19,940 modifies that, where end users will pay 20% of electricity transmission and generators pay the remaining 80% through pricing formulas. Thus, in the near future, the structure of the prices for small end users will additionally consider a unique charge for transmission access in proportion of their consumption of energy, in such a way that the resultant price corresponds to the end user allocated cost at the level of production, transportation and distribution. 20 Some proposals have been made to replace the current mechanism (2/3 and 1/3 weighting factors) by one where, for example, the Electric Expert Panel, “Panel de Expertos Eléctricos”, chooses one of the tariff studies. See note. 21 Article No. 71-43 of Law No. 19,940 defines the criteria to calculate electricity distribution access charges for other agents who want to sell electricity in the distribution companies’ network. Basically, the fee is calculated in such a way that when the large user buys energy and power at regulated nodal prices, his final price, including the electricity distribution fee, should be equal to the price that a regulated small end user pays. The access charge is determined as an efficient component pricing rule for the distribution company. Today it is expected that with these access fees, competition to supply electricity to large end users within a distribution companies network increases as large end users start renewing their electricity supply contracts, and also because the number of end users who qualify to participate in the large customers market has increased. 22 In the same way as it happens with Brazil’s electric transmission system. In that respect, see Joao Lizardo R. and Hermes de Araújo (2006). 19

94


companies with the cities and power plants on the Northern coast. In both systems exists one private company (both Hydro-Québec subsidiaries) which owns the main transmission lines, frequently with a radial configuration. Until recently, revenues for transmission lines owners were determined by a multiple-part tariff approach defined by Law No. 18,922 of 1990, a change to DFL No. 1 in relation to transmission payments. According to it electricity transmission is paid by the generators as per their location within the network and as a function of their electricity injections and withdrawals from the network. Transmission charges consist of three components: ● ● ●

A “tariff revenue” (“Ingreso Tarifario”). A “basic toll” (“Peaje Básico”). An “additional toll” (“Peaje Adicional”).23

The “tariff revenue” is a variable fee associated to the differences in power and energy nodal prices within the different nodes and paid according to the generator injections and withdrawals of power and energy; while the “basic toll” and the “additional toll” are fixed fees calculated as a supplement to the “tariff revenue” to complete the payments that the transmission owner must receive, and they are paid depending on the generator’s area of influence and the location where the generator withdraws electricity.24 For a generator, the payment of the basic toll gives the right to withdraw electricity, without additional payments, in all the nodes of the system located within its area of influence. In addition, it gives the right to withdraw electricity, without additional payments, in all the nodes from which, in typical operation conditions of the system, net physical transmissions take place toward the area of influence.25 If the generator wishes to withdraw electricity in other nodes different to its area of influence, it must agree additional tolls with the transmission lines and substations’ owner.26

23

See Rudnick and Raineri (1997b). The total payment of electricity transmission facilities is defined as the annuity of the replacement value plus operation and maintenance costs. The area of influence should be understood as the set of lines, substations and other facilities of the electric system, direct and necessarily affected by the injection of power and energy by a generating unit. The basic toll is defined as a fixed fee that must be paid to complete the annuities corresponding to operation and maintenance costs and investment in lines, substations and other transmission facilities in the area of influence, after subtracting the tariff revenue. To this end, the basic toll is calculated in proportion to the maximum power transported by each user within the area of influence and must pay: the investments at their replacement value, assuming a lifetime of 30 years, the annual operation and maintenance costs, and a 10% real return on investment. The fees are revised every 5 years. 25 The net transmission, for these effects, is defined as the average energy transmission throughout a calendar year. This right will subsist in as much the net transmission stays toward the area of influence. 26 The additional tolls are calculated in the same form as the basic toll, and the payments give the generator the right to withdraw electricity in all the nodes located within the facilities. It also grants the generator the right to withdraw electricity, without additional payments, in all the nodes from which in typical conditions of operation of the system, net physical transmissions take place toward the nodes covered by the additional tolls. This last right exists in as much the net transmission condition indicated is satisfied. 24


95

The main problems with electricity transmission have been the difficulties between transmission companies and generators to agree the basic and additional tolls, which can be explained because of: ● ● ●

the ambiguity of the concept of area of influence, a key ingredient to define the tolls; changes in the direction of net physical transmissions or energy flows; the vertical integration that existed in the generation–transmission segment.27

Together these have implied that some segments of the transmission system have been left unpaid, causing a negative effect on the incentives to invest in the additional transmission capacity required for the reliable operation of the system.28 In the year 2000, the vertical integration that existed in generation–transmission ended, but despite that, the disagreements about electricity transmission fees remained. To end with the controversies that existed on transmission fees the authority approved a change in the law in 2004, Law No. 19,940, which defines a complete new order for electricity transmission. Basically, it changed the model from one of the decentralize transmission industries with negotiated fees to one of the centrally planned segment or monopoly with regulated fees. The reform conveys: ● ●

●

27

the independence of transmission from generation and distribution; a change in the legal status of main transmission lines and sub-transmission lines to one of public service, which determines stricter service obligations;29 mandatory transmission investments defined in an electricity transmission expansion study supervised by CNE, which recognizes two types of expansions, new facilities and extensions of existent facilities;30

In the SIC until year 2000, ENERSIS S.A. controls: the largest generator ENDESA S.A., with 63% of generation capacity; the largest distribution company CHILECTRA S.A., with almost 100% of electricity distribution in the region where the capital city is located; and TRANSELEC S.A. owner of 98% of 500-kV transmission lines. The vertical integration ended in the year 2000 after ENDESA Spain obtained the control of ENERSIS holding, and it sold TRANSELEC S.A. to Hydro-Québec (see Raineri, 1999, 2003). 28 Since the early 1990s, many arbitrages have taken place to define basic and additional tolls; arbitrages which frequently resulted in contradictory results. The lack of technical and economic concepts of what is a generator’s area of influence explains most of the disputes. What was agreed in one arbitrage is not compulsory to what can be agreed in other arbitrages, implying that the concept of area of influence was defined in different and eventually contradictory manners in the different arbitrages. 29 The change in the legal status mean that the development and planning of electricity transmission turns out to be an activity in the benefit of end users and generators, and not as a generators’ development strategy or an extension of the generator’s activity to reach the consumption centers as was the case with the former law. 30 The study is made every 4 years. With that study as reference, every year the CDEC’s Operations Direction revises the consistency of the transmission facilities and makes a recommendation to CNE on the additional facilities needed. CNE, with the study and the CDEC recommendations, defines the expansion plan for the following 12 months. This expansion plan defines investments that are mandatory for transmission companies when they are an expansion of an existent facility, or investment projects to be publicly bid if they are a new development. The identification of the additional facilities, new or expansion arises from a regulated process with the representatives of all the players of the electric sector, and with regulated end users being represented by government authorities. For new investment projects, the execution and operation is carried out by the company that wins in an international public bidding. In the case of a development that is considered as an expansion of an existent major installation, the owner of the existent facility is the one that has the responsibility to carry out the works and operate the facilities in compliance with the law.

96

●

●

Electricity Market Reform the value to be paid for existent transmission facilities, which is determined every 4 years; and criterias to pay for the investments that finally are approved as necessary for the development of the electric system and to preserve its security.

The new tariff criterias for new investments discriminate between new facilities and extensions of existing facilities. For new facilities the tariffs consider the value of investment plus the operation, management and maintenance costs, and are applied during five tariff periods (for a total of 20 years), after which they are updated by the value determined in the main transmission study supervised by CNE. For investments projects identified as expansions of existing facilities the tariffs are revised every 4 years, which implies a larger uncertainty to recover investments as well as the operation, management, and maintenance costs.31 By law, the study which defines the expansion of the transmission system should consider existing generators’ facilities as well as the generation facilities that the companies declare in construction. Thus it happens that when a company declares the construction of a new power plant, it conditions the future development of electricity transmission, but in the end the company is not forced to build the power plant. What this means for the development of the electric industry is an open question, but so far the number of generators who has announced the construction of new power plants has increased compared to what was seen in the past years.32

3.3. Electric Crises The Chilean ESI has suffered three major crises since 1998: ●

●

31

In 1998 and 1999, the SIC experienced a drought that severely affected the availability of hydro resources. In 1999, the explosive growth of the SING installed capacity by the entry of new natural gas power plants led to a sequence of blackouts and brownouts whose origin was the lack of regulations which does not allow the proper coordination of a fast growing system, among them, and most importantly, rights were not defined for the appropriate provision of ancillary services.

With the new law, the trunk system is divided between the common area of influence and the rest of the trunk system. The common area of influence is paid in 80% by power plants injections and in 20% by consumption withdrawals (contracts). The rest of the system is paid by injection or withdrawal, completely, depending if the power flow gets in or out of the common area of influence. From this, it happens that, on average, the trunk system is paid 70% by energy injections and 30% by energy withdrawals. Tolls are defined according to the investments’ replacement value average costs determined by the CDEC Tolls Direction, according to the law and the proportional use that each generator’s injections make of the system. For sub-transmission, the tariffs are determined by the CNE every 4 years bearing in mind efficient investments at their replacement value. 32 With Law No. 20,018, approved in early 2005, that imply an increase in node prices, the generators’ interest to invest in power plants increased, where more than 5,600 MW of additional capacity for the next 10 years have been announced, demanding US$ 4 billion on investments. These announcements finally were included in the October 2005 CNE Indicative Works Plan for Generation and Transmission, where 5,700 MW of additional investments in generation are considered. This happened after a long period of time where generators expressed no interests to invest in generation.


●

97

In 2004, the political instability in Argentina has exposed the excessive dependence of Chile on Argentinean natural gas, where the Argentinean natural gas export constraints have left the Chilean electric industry operating at the limit of its capabilities.

The causes of these crises, their effects, and efforts to address the problems are briefly described below. 3.3.1. First crisis The SIC electric energy crisis was triggered because in 1998–1999 the Central zone of Chile suffered the most extreme drought of recent years, which was complemented by numerous delays in the entry of a new natural-gas-fired combined cycle power plant (named Nehuenco), which would add 5% of additional generation capacity. The drought and the delay of Nehuenco, due to major technical failures, left the SIC with generating reserves below the minimum operating conditions. Under these circumstances the authorities approve three Electricity Rationing Decrees allowing electricity distribution companies to effectively ration electricity consumption in the months of March, April, and May of 1999. Fortunately by mid-1999, the hydrologic conditions change favorably and with it, the need to extend the third rationing Decree disappeared. The electric energy deficit engender both end users complain questioning the regulatory model and the generators’ disagreement on the price that should be used to pay for energy transfers between the generators: the price should be the system’s marginal cost or the system’s failure cost?33 Also, additional concerns existed with the use of the water that was stored in dams before the crisis, and that because 1998 ended with a rain deficit which was ignored. Regulated nodal prices are calculated with a stochastic model which consider a set of parameter values and assumptions, among which stands the range of 40 hydrologic years.34 On range of the 40 hydrologic years hydraulic generators made the case that droughts more severe than the ones considered in regulated nodal prices are circumstances of Force Majeure, which exonerates them from paying compensations to small end users as well to value the energy transfers that take place in the CDEC at the failure cost. Their main argument was that the applicability of the mathematical models should be restricted by the range of the parameters being used and model assumptions. Thus, the prices determined by the mathematical models should be applicable only if the system operates within the range of the parameter values being used. Their argument was backup because in a normal operation condition, small customers pay a risk premium within the regulated node price by which they will be compensated for the non-served energy. It is understood that the risk premium should be applicable only in the cases previously foreseen within the range of parameter values considered for the nodal prices calculation.35 Thus if the model is required

33

Traditionally the failure cost is far above the marginal cost of the less efficient power plant being dispatched, and the larger the energy deficit, the larger the failure cost. See Raineri and Ríos (1998); and Raineri (2000). 34 See Footnote 17. 35 The risk premium paid, the failure cost, forces the generators to compensate small end users for the non-served energy whenever electricity rationing occurs, in circumstances foreseen in the model and that cannot be invoke as Force Majeure. Because of that, the companies have the incentive to invoke as Force Majeure or Act of God those circumstances that were not considered in the calculation of node prices, as droughts more severe than the ones considered in the calculation of node prices.

98


to give price signals for all foreseeable and unforeseeable events, the range of the parameters considered should be wide enough to include also the extreme and less probable events that may occur. In the debate, thermal generators with surplus energy favor the idea that energy transfers should be paid at the failure cost as well as that the energy deficit should be equitably distributed among all the small end users of the distribution companies. Their incentives were to forward electricity to fulfill other generators – distribution companies’ contracts, being paid for that a price equal to the failure cost, well above the price at which they had signed electricity supply contracts with distribution companies. In addition, as the failure cost is increasing in the size of the energy deficit, it happens that the larger the deficit, the higher the price at which generators with surplus energy could sell their electricity. Hydro generators have been forced to take actions to diminish the negative adverse effects of the drought. In the case of ENDESA holding a predominantly hydro generator, it: ●

● ●

invest more than US $200 million in turbines, with an additional contribution of 651 MW of power; contracted 99.8% of the cogeneration capacity in operation; extended the transmission capacity from north to south to evacuate the generation capacity surplus in the North zone of the SIC.36

Colbún, another hydraulic generator and owner of Nehuenco, made significant efforts with plant builders to repair and put in operation the new natural-gas-fired combined cycle power plant as soon as possible. The reaction of the authority to the crisis was to change the law. In particular from these changes resulted the highly controversial change introduced in Article 99bis of DFL No. 1, where droughts, more severe than the ones considered in the calculation of node prices have been ruled out as causes of Force Majeure.37 Other changes in the law are: ●

●

●

energy transfers that take place in the CDEC should be valued at the system’s marginal cost, which for rationing hours was defined equal to the failure cost; the socialization of energy losses requiring the energy deficit to be equally distributed within the small end users of all the distribution companies; and a substantial increase on the sanctions for generators that fail to satisfy electricity supply contracts for distribution companies, and in the case of a Rationing Decree, rationing has to be executed proportionally to the commitments of each company.

The changes significantly increase the generators’ risk. They change the generators’ responsibility in the contracts with distribution companies as well as in the price that they have to pay for energy transfers that take place in the CDEC, risk that was not equally recognized within the regulated price. The flaws of the regulatory change undertook in Article 99bis became evident in the year 2001 when, after many purchase announcements, few distribution companies have fail in their intention to contract electricity from a generator, as required by law. Up to May 2005, the distribution company Saesa had made eight unsuccess purchase announcements; and CHILECTRA, the largest Chilean distribution company, also failed three times, and only 36

See Juan Eduardo Vásquez (1999). Article 99bis of DFL No. 1 was changed through Article 2 No. 2 of the Law No. 19,613 from June 2, 1999, published in the Official Newspaper on June 8, 1999. 37


99

succeeded in 2003 after an additional change in the law was implemented, which constrained the generators’ liability for the non-served energy only to the true deficit. In 2001, the reaction of the government to solve the generators’ lack of interest to sign long-term contracts with distribution companies came through the Ministry of Economy, which on April 30, 2001 enacted the R.M. Exempt No. 88 by which it requires from the generator members of the CDEC-SIC to jointly satisfy all the distribution companies’ demand for electricity, independently of the distribution companies have or not an electricity supply contract.38 The decreasing trend of regulated nodal prices and the negative effect of Article 99bis underestimated the generators’ risk to sign electricity supply contracts, and that become apparent in the CNE’s 2001 SIC “Indicative Works Plan for Generation and Transmission” which did not project the entry of new power plants in the short term.39 At the time, the date of entry of the next power plant was predicted for January 2003, a natural-gas-fired power plant that later was not built, and in July 2003, a 570-MW hydroelectric power plant that finally began operating in 2004 with 640 MW. In terms of transmission projects, the “Indicative Works Plan for Generation and Transmission” expected for January 2004 an interconnection line between the SIC and the SING, which would contribute 250 MW to the SIC, and in July 2006 an interconnection line with Argentina which would contribute 400 MW to the SIC. So far, none of the transmission lines have been built. The SIC delay of investments in generation capacity results from the small capacity growth compared to the electricity demand growth (see Fig. 3.2). The lack of investments and the critical electricity supply situation expected for the years 2002–2004 concerns the authorities,40 who, in the desperation of avoiding a potential crisis, threatened the industry to purchase 250–500 MW of additional capacity if the private investors do not compromise these additional investments. Afterward the authority desisted when it was criticized by taking a role that is beyond its normative and supervisory roles. The main energy policy lessons from this crisis are as follows: ●

●

●

●

38

The lack of flexibility of regulated energy prices isolates small end users from large short-term fluctuations that happen in the market. The parameter values of the mathematical model can be exceeded in extreme conditions of operation, and require defining norms to operate under these conditions. Conditions which are excluded from being a cause of Force Majeure should be defined in advance and recognized in the prices. The industry should design a larger number of instruments which allows for a smooth balancing of supply and demand.41

In R.M. exempt No. 88 determines that the energy withdrawn by distribution companies without a contract should be paid at the regulated node price, as well as that the energy consumption should be proportionally distributed among all the generators. 39 The CNE Indicative Works Plan for Generation and Transmission is made taking into account all the power plants and transmission lines that the different companies have declared under construction. Thus, if the Indicative Works Plan for Generation and Transmission does not expect in the short- to medium-term new facilities, it must be because private agents have no interest to invest in the industry. 40 Between years 1999 and 2003, the installation of new power plants was expected to be inferior in 50% to what the authority anticipates that the demand will grow. See Magazine Qué Pasa, June 16, 2001. 41 Demand side management tools like tariff incentives and price–quality menus have been almost absent in Chilean ESI. The importance of price – quality menus for electricity supply contracts is analyzed in Raineri and Rudnick (1997).

100

●


The need to have a well-defined ancillary services market and its pricing mechanism, which may also consider payments for investments in backup capacity.42

3.3.2. Second crisis During 1999 the operation of the system faced two large disturbances on July 25 and on September 23, taking the system to blackouts, additionally to partial outages. Unlike the problems of the SIC which were caused by nature, the SING’s problems resulted from a sharp growth in the installed capacity (Fig. 3.2) that lacked the adequate change in norms and rules to guarantee the appropriate coordination of the large generating units and the large consumption of mining companies. The SING is a system where the large mining companies explain nearly 90% of the total energy consumption, which also occurs in large blocks with respect to total demand (the largest single mining consumption is 300 MW, the system’s peak load is 1567 MW, and new natural gas-combined cycle power plants are as large as 400 MW). A system of these characteristics becomes very unstable if occurs a sudden exit/entry of a large block of consumption and/or of a large power plant. Thus and given the particular characteristics of the system, in the heart of the crisis was the weak provision of the ancillary services being required to guarantee a reliable electricity supply.43 Contrasting with the SIC, the imbalances that cause the SING problems are not explained by the lack of energy, but have their origin in: ● ●

●

● ●

the uncoordinated entry and exit of large generating units; constraints in power plants’ static and dynamic properties (speed at which they can take or drop loads); constraints on the power plants’ technical minimum of operation and the system’s stability if it is operated with small loads with respect to the installation capacity; failures in the transmission system; and uncoordinated entry and exit of the large blocks of consumption from mining companies.

The large mining companies criticize the deficiencies of electricity supply and the ambiguity of the contracts, and inquire the generators to bear the costs of a deficient supply by: ● ● ●

●

investing in backup capacity; increasing spinning reserves; setting load shedding at 48 Hz to avoid the collapse of the system (in a 50-Hz system); and setting a limit on the maximum power that can be dispatched by each generator.

Although large mining companies recognized that in the year 2000 the incidence of blackouts diminished, they argued that the system is one that allows to fail and not to pay.44 As a consequence of the crisis, and to reduce the risk of future blackouts with a costly effect on large mining companies, and with the cooperation of member companies as well 42

Payments to backup capacity today are made through reserve margin payments. The margin of 15% was effective before 1998 and lowered to 6% in 2001 for the SIC and 5% for the SING, and lately they were increased to 11.76% according to the April 2005 Nodal Prices Decree for the SIC and the SING. 43 See Raineri et al. (2005) and Ríos et al. (2005). 44 See Electricidad Interamericana, December 2000 – January 2001.


101

as large end users, since November 30, 1999, the CDEC-SING implemented a Short-Term Security Plan, which has had later revisions. In it, the CDEC defined the actions undertaken to mitigate the effects of a sudden disconnection of a large generating unit or a large consumption. On January 2001 this plan considered: ● ● ●

the installation of LF relays for a maximum of 340 MW; a margin or spinning reserves of 15%;45 and a limit on the maximum capacity to be dispatched by each power plant that today is of 250 MW.46

The above limit on the maximum capacity to be dispatched happens despite some of the newest power plants of the SING have an installed capacity of 400 MW. Thus, the question that come up is how the contribution of the different agents to the security of system will be paid, recognizing the fact that the new power plants have an installed capacity up to 400 MW but are restricted only to dispatch 250 MW. The SING’s crisis was essential to recognize within the Chilean legislation, the role that ancillary services have in the security and stability of the electric system, services that at the time of the crisis were voluntarily provided by few companies of the SING, as it also happens in the SIC. The more typical ancillary services required in the SING are voltage control, frequency regulation, primary and secondary injection of active and reactivate power, spinning reserves, and the backing up of power units when they are fine-tuned and run setup tests. Regarding the character of these services, some of them benefit all the agents in the industry, as a public good, and others only benefit a few, as a private good.47 The main energy policy lessons from this crisis are as follows: ●

●

●

The need to have suitable measures to allow the right coordination of all the agents who participate in the industry. The problem of the SING was not of an energy deficit, but rather it was a problem of a deficit of instantaneous power to absorb the sudden entry/exit of large blocks of consumption and/or large generation units. The need to have norms to backup the process of testing and fine-tuning of the new power plants. Particularly, in the current crisis this happened with new power plants whose capacity is from 20% to 52% of the peak load of the system. The need to have a well-defined ancillary services market and a pricing mechanism which adequately recognizes the character of each of the services. The effort must be placed on finding an incentives structure where the contribution to preserve the quality and the reliability of the electric service is a private and not a public good to which the different agents of the sector contribute in an unequal and non-remunerated basis.

Thanks to the SING’s Short-Term Security Plan, the failures of the system decreased from 192 in year 2000 to 72 in year 2003, where the ones that finally have had an effect on end users decreased from 65 in year 2000 to 13 in year 2003.

45

On October 2000, in the calculation of nodal prices, this theoretical power reserve margin was previously established in 5%. 46 First this limit was set at 180 MW, then it was increased to 200 and 220 MW in peak hours, and today it is in 250 MW. 47 See Raineri and Rudnick (1997); Raineri and Ríos (2001); and Galetovic (2001).

102


3.3.3. Third crisis From the early 1990s, the idea to import natural gas from Argentina matured. This country had important reserves and a growing infrastructure for its transportation, also complemented by the large natural gas reserves that were found at the end of the decade in Bolivia. Thus, the perception was one of the abundant natural gases in the region for many years. In the early 1990s the deregulation and privatization of the natural gas and the oil industry in Argentina facilitated in 1995 the natural gas integration with Chile, which allows the free import and export of this product between both countries. By the agreement, exports are allowed as long as there is no compromise to the internal supply of the exporting country, and in case of rationing it would be equally distributed within all the consumers independently of their country of origin.48 Argentina, as the exporting country, required that any company interested to export natural gas to Chile should prove that it had enough reserves to satisfy its domestic contracts as well as a surplus to satisfy its export contracts. Under this agreement, seven international gas pipelines have been constructed between Argentina and Chile. Under this situation, Chile began to import increasing quantities of natural gas from Argentina, since 1997 in the SIC and 1999 in the SING, whereas in 2003 Chile imported an average of 22 Mm3/day of natural gas from Argentina. In Chile 33% of natural gas imports are used for electricity generation, to fuel five thermal plants in the SING, with a capacity of 1443 MW49, and seven in the SIC, with a capacity of 2097 MW. The SIC natural gas power plants reach 23% of the installed capacity (see Table 3.3); and in the SING they reach 59% of the installed capacity (see Table 3.2). The SIC and the SING gas pipelines have fueled power plants, which between 1996 and 2004 represent 72% of the total increase in generation capacity (see Tables 3.2 and 3.3). The remaining natural gas imports are to supply the industrial consumption of a methanol plant (37%), and residential, industrial and State Oil Refineries consumption (30%). The introduction of natural gas has implied that the Chilean electricity sector became highly dependent on Argentinean natural gas, where in 2004, 28% of the SIC’s generation and 61% of the SING’s generation was based on natural gas. In this situation the system’s efficient dispatch significantly depends on the availability of natural gas, and any shortage of it can severely affect the cost of the systems and the availability of useful generation capacity required to satisfy electricity demand. Also, any shortage of natural gas puts in a risky condition; the payments being required to recover the large investments that have been made for its use. The success introduction of natural gas in Chile has been glimpsed in 2004 when Argentinean economic problems affect its capacity to satisfy the natural gas exports contracts. The economic imbalances developed in Argentina during the 1990s became unsustainable in 2002 with the devaluation of the local currency, which since 1992 was artificially set at one Argentinean peso per American dollar. With the devaluation, in a short period of time, the Argentinean peso dropped to a rate of 3 pesos per American dollar. The devaluation led to large changes on relative prices with significant adverse effects on income and employment. According to World Bank’s figures, between October 2000 and October 2002, the percentage of Argentinean population under poverty increased from 33% to 58%.50 To 48

Protocolo No. 2 de Integración Gasífera de 1985 (Natural Gas Integration Protocol No. 2 of 1985), as part of the Chile and Argentina Economic Complementation Agreement. 49 These power plants have been complemented with a 643-MW natural-gas-fired power plant (Salta), built in the northwest of Argentina, which through a 345-kV transmission line injects electricity into the SING. 50 World Bank (2003).


103

diminish the short-term adverse effects of the devaluation on the population, government authorities pesified (fixed in pesos terms) the tariffs for basic utilities and transportation services, what differs to what have happened to other basic goods whose costs have risen by 30% on average. The pesified tariffs for natural gas consumption – household, transportation, electricity, and industry – become severely distorted with respect to the price of the closest substitute fuel. The artificially low natural gas price increased the domestic demand and also has replaced the use of the alternative but more expensive fuel. The pesified tariffs for the producers have a negative effect on the incentives to increase natural gas extraction, and to invest in exploration and the construction of additional transportation and distribution facilities. Thus sooner than later, production and transportation capacity constraints have become binding in a context of declining proven natural gas reserves subject to distorted energy prices. Argentinean natural gas exports mostly go to Chile, which represents 13% of Argentinean natural gas production. Thus, for Argentinean authorities, and if there is any shortage on natural gas supply, cutting exports to Chile appear as the less costly choice taking into account internal political constraints. Even though domestic tariffs in Argentina have been pesified, natural gas exports to Chile are under long-term contracts expressed in American dollars, meaning that natural gas arrives in Chile at prices roughly of US $2.5 MBtu (approximately US $1.6 MBtu at wellhead price), far above the Argentinean regulated natural gas prices for electricity generation of US $0.5 MBtu.51 As a result of the price distortions together with a modest recovery of the Argentinean economic activity after the 2002 crisis, Argentina has experienced a significant growth in domestic natural gas demand. However, due to the artificially low prices received by natural gas producers, their incentives to invest in exploration and the development of new gas fields and facilities has been eroded. From 2004 and with an increased emphasis in 2005, the Government of Néstor Kirchner in an effort to assure its domestic consumption has imposed severe constraints on natural gas exports to Chile, where exporting companies are mandated to fulfill domestic consumption prior to their export contracts. The natural gas deficit in Argentina that started in 2004 has constrained natural gas exports to Chile, meaning a deficit for Chile with respect to the consumption in a normal supply condition. Figure 3.7 shows that daily reductions on natural gas have reached levels of 50%, where the sectors most affected by these reductions are electricity generation and industrial processes. The thermal power plants that were able to substitute natural gas with an alternative fuel (diesel oil) have seen the fuels cost growing fourfold, affecting their relative efficiency in the CDEC dispatch and the system’s marginal cost. The industrial sector, when feasible, invested in backup capacity to continue with its operations. Household consumption was unaffected because Chilean authorities mandate that available natural gas should first favor households, hospitals, and small stores, and then electricity supply over large industrial customers. Looking back, the SIC and the SING benefit by having Argentinean natural gas at low prices. From its introduction and until March 2005 this implies savings close to US $3.5 billion on the fuel bill.52 However, since 2004, the natural gas deficit has left the electric system at the limits of its operational capabilities; the CDEC was forced to plan a more expensive

51

Crisis Energética, in Acontecer, Publication of the Instituto Tecnológico de Buenos Aires, May 1, 2004. Other benefits are the reduction in air pollution in Santiago, city that in the early 1990s was one of the most air-polluted cities in the world. 52

104


60

Natural gas injection constraints to Chile requested by Argentinean government respect to natural gas injections under unconstrained conditions. Total constraints between 05-05-2004 and 05-23-2005: 1328 Mm3.

50

%

40 30 20 10

2005-5-18

2005-4-6

2005-2-23

2005-1-12

2004-12-1

2004-10-20

2004-9-8

2004-7-28

2004-6-16

2004-5-5

0

Fig. 3.7. Imported natural gas constraints on Chile. Source: CNE.

dispatch using all the natural gas that is available at any moment, saving water in the case of the SIC for those periods when the natural gas deficit could be even larger. With this, the SIC’s reliability of electricity supply is left to what happened with the rain and hydro power generation. Natural gas supply in the SING depends not only on the decisions of the Argentinean authorities, but also on the Bolivian authorities’ decisions. This is because since August 2004 Argentina receives an important amount of gas from Bolivia in the north of the country (6.5 Mm3/day), through an agreement that expires in late 2005. The current situation implies a risky condition for the SING, because if Bolivia curtails natural gas exports to Argentina, the natural gas deficit in the northwest basin of Argentina would force its authorities to suspend all the gas delivered from that basin to Chile. In such a case, the SING remains with the generating capacity of other fuels, as coal, fuel oil and, diesel oil may be in fact not enough to satisfy the current SING’s electric demand. This situation illustrates that even though there are reserves and production capacity of gas in Argentina and Bolivia to satisfy the needs of the region, the geopolitical relationships are seriously affecting the Chilean electric market in the north of the country. The reactions in Chile to the Argentinean natural gas export constraints have been diverse, among which stand out the Chilean authorities lead to invest in a plant to imported liquefied natural gas (LNG). However, given the long lead times nothing is expected from this project until 2008, at the earliest. Also stands outs the changes in regulation that were approved on May 2005 which imply that:53 ● ●

53

natural gas shortage in thermal power plants cannot be invoked as a case of Force Majeure; generators are allowed to offer incentives to small end users to reduce electricity consumption when an energy deficit arises; Law No. 20,018, May 2005.


●

●

●

105

distribution companies are allowed to sign long-term contracts, where the bid price will be considered in the nodal prices;54 a mechanism is defined to provide price incentives for investments in renewable energy; and until 2008, the generators who supply electricity to distribution companies (regulated end users) without contracts will receive a price which follows the system’s marginal cost.55

Even though some of these regulatory changes will not solve the short-term problem, they are expected to encourage new investments on a diversified energy matrix. At the same time regulatory changes have provided the generators with some instruments by which they can persuade small end users to reduce their consumption, with regulated nodal prices responding in a more flexible way to short-term supply conditions. The main energy policy lessons from the current crisis are: ● ● ● ●

●

an excessive dependence on natural gas imported from Argentina; the need to have end users’ prices responding to supply conditions; the need to have a diversified energy matrix; the need to have flexible tools or incentive instruments to balance electricity supply and demand in situations of extreme operating conditions; the need to have binding bilateral agreements.

3.4. Final Comments In the 1980s, Chile has pioneered the modernization and deregulation processes of the electric industry worldwide, where the deregulation of the Chilean electric industry responded to a change in the beliefs of the role of the State and the private sector on the economic activity. The Chilean electric industry followed the road of a market-oriented model. After more than two decades of experience with a deregulated and privately owned electric industry, the sector still goes under continuous adjustments which respond to a learning process of what have worked and what have not, as well as to the industry crises that have had significant impacts on the society. In this chapter we described the most distinctive features of Chilean ESI reform and three most important crises that it experienced since the late 1990s. The first crisis was in 1998–1999 when the SIC, a mostly hydraulic system, suffered a severe drought which took to an electricity rationing at the residential and industrial level; the second crisis was in 1999 that affected the SING, a thermal system, which experiences a fast growth on demand and generation capacity which was not complemented by the adequate norms to coordinate the operation of large new power plants and the large consumption of mining companies; and the third crisis which affected the Chilean ESI since 2004 occurred

54

The long-term contracts should not exceed 15 years, and the bid price has a 20% cap above the regulated nodal energy price. If in a contract a distribution company obtains a price which exceeds the average price of the industry by more than 5%, the difference that exceeds that 5% is socialized in the tariffs of all the distribution companies. 55 Basically, this refers to generators being mandated on April 30, 2001 by the R.M. exempt No. 88, by means of which it is ordered that generator members of the CDEC-SIC must supply all the electricity demand of distribution companies independently if they voluntarily have signed an electricity supply contract. The price would be equal to the nodal price plus some fraction of the difference between the nodal price and the CDEC dispatch marginal cost.

106


as Argentina has severely constrained natural gas exports to Chile, while the Chilean ESI has become increasingly dependent on natural gas imports from that country. The three crises imposed political pressures which ended up with changes in the law, changes that in some cases have been a response to the social concern that created the crises instead of a long-term vision of the rules needed to develop an efficient ESI. Looking back to what has happened and what has been made, it should be said that among the main energy policy lessons that can be derived from Chilean experience are: ●

● ●

● ●

●

●

● ● ●

the need to have flexible prices which do not isolate end users from short-term market conditions; the need of demand side management tools, like tariff incentives and price–quality menus; the mathematical models used to set regulated and simulated prices as well as for the system operation, reflect a circumscribed set of conditions, and this requires the definition of norms to operate under extreme conditions which exceed the ones considered by these models; the need to have a well-defined ancillary services market; the need for an adequate coordination among all the agents and a clear definition of rights and duties of each one; a clear definition of the reliability level that the ESI must have, and how much will be paid for that as well as for backup capacity; conditions which are considered as causes of Force Majeure should be defined in advance and recognized in the prices; the need to have diversified sources of energy supply; the need to have binding bilateral agreements; and the need to avoid regulatory uncertainty and to have stable rules with the flexibility to respond to the changes needed by the industry.

The Chilean ESI was among the first ones to be privatized and deregulated and after being battered for more than two decades, it offers many useful insights for policy makers and regulators around the world. Although Chilean electric industry have taken the road of market-oriented mechanisms, there is still some way to go to have a complete free market driven electric industry. Acknowledgments The author specially thanks the valuable comments made by Sebastian Berstein, Francisco Courbis, Juan Manuel Cruz, Raúl Espinosa, Luis Hormazabal, Eduardo Soto and the editors Fereidoon P. Sioshansi and Wolfgang Pfaffenberger.

References Decree-Law No. 327, 1998. Berstein, S. (1986). Tarificación eléctrica a costo marginal en Chile: Aspectos conceptuales, metodológicos y practices. IV Seminario Latinoamericano y del Caribe sobre Tarifas de Energía Eléctrica, Lima, Perú, November. Boiteux, M. (1949). La tarification des demandes en pointe: Application de la théorie de la vente au coût marginal. Revue générale de l’électricité, 58, 321–340.


107

Boiteux, M. (1956). Sur La Gestion Des Monopoles Pubics Astreints AL’Equilibre Budgetaire. Econometrica, 24, 22–40. Boiteux, M. (1960). Peak load-pricing. Journal of Business, 33, 157–179. Comisión Nacional de Energía (1989). El Sector Energía en Chile, December. Decree-Law No. 6, 1985. Decree-Supreme No. 219, 1999. Decree-Supreme No. 287, 1999. Decree-Supreme No. 640, 1998. Electricidad Interamericana, December 2000 – January 2001. Galetovic, A. (2001). Asignación de costos debidos a centrales que operan para satisfacer restricciones técnicas en el SING, CEA, Departamento de Ingeniería Industrial, Universidad de Chile. Hachette, D. and Lüders, R. (1994). La Privatización en Chile. CINDE. Instituto Tecnológico de Buenos Aires, May 1, 2004. Crisis Energética, in Acontecer, Publication of the Instituto Tecnológico de Buenos Aires. Isaac, D., Santiago, A. and Eric, R.L. (2006). In this book the chapter Understanding the Argentinean and Colombian Electricity Markets. Jenkins, R.T. and Joy, D.S. (1974). Wien Automatic System Planning Package (WASP) – An Electric Utility Optimal Generation Expansion Planning Computer Code. OAK Rige National Laboratory, ORNL4945. Joao Lizardo R. and Hermes de Araújo (2006). In this book the chapter The Case of Brazil: Reform by Trial and Error. Joskow, P.L. (2001a). California’s Electricity Market Meltdown, April. Joskow, P.L. (2001b). California’s Electricity Crisis, March. Juan Eduardo Vásquez M. (1999). Energy Managing and Planning Director, ENDESA S.A., presentation at Pontificia Universidad Católica de Chile, September. Law No. 20,018, May 2005. Law No. 24,076, 1992. Regulates transport and distribution of natural gas in Argentina, authorizes natural gas exports subject to a government approval. Law No. 19,940, March 2004. Magazine Qué Pasa, June 16, 2001. Manifesto on the California Electricity Crisis, January 26, 2001. Institute of Management, Innovation, and Organization at the University of California, Berkeley. Ministry of Mines (1982). General Law of Electric Services, Decree-Law No. 1 of the Ministry of Mines (DFL No. 1). Natural Gas Integration Protocol No. 2 (Protocolo No. 2 de Integración Gasífera), 1985. Protocol That is an Element of the Chile and Argentina Economic Complementation Agreement. Public Emergency and Exchange Rate Regime Reform Law 25,561 (Ley de Emergencia Pública y Reforma al Régimen Cambiario 25,561), February 2002. Argentina. Raineri, R. (1999). Buscando el Control Corporativo: El Ingreso de Endesa España a la Propiedad de Enersis. Ediciones Universidad S.A. y M.N. Consulting Ltda., Santiago Chile, p. 138. Raineri, R. (2000). Comentarios al documento: Anatomía de una crisis eléctrica. Working Paper No. 101, Departamento de Ingeniería Industrial y de Sistemas, Pontificia Universidad Católica de Chile. Raineri, R. (2003). Becoming global: the entry of ENDESA Spain in the property of ENERSIS and the DUKE energy contest, Working Paper No. 142, Departamento de Ingeniería Industrial y de Sistemas, Pontificia Universidad Católica de Chile. Raineri, R. and Ríos, S. (1998). Costo de Falla y Precios para Valorizar las Transferencias de Energía en el CDEC, Working Paper No. 87, Departamento de Ingeniería Industrial y de Sistemas, Contract DICTUC S.A. – ENDESA S.A. Raineri, R. and Ríos, S. (2001). Contratos y Calidad de Suministro Eléctrico en el SING, Working Paper, Departamento de Ingeniería Industrial y de Sistemas. Raineri, R. and Rudnick H. (1997). Analysis of service quality standards for distribution firms. In L. Felipe Morandé L. and B. Ricardo Raineri (eds.), (De)Regulation and Competition: The Electric Industry in Chile. Ilades-Georgetown University, pp. 259–294.

108


Raineri, R., Ríos, S. and Vasquez, R. (2005). Business opportunities and dynamic competition through distributed generation in primary electricity distribution networks. Energy Policy, 33(17), 2191–2201. Ríos, S., Raineri, R. and Roca, M. (2005). Dynamic performance and resource mix modification in competitive environment. IEE Proceedings in Generation, Transmission and Distribution, 152(6), 770–779. Rudnick, H. and Raineri, R. (1997a). Chilean distribution tariffs: incentive regulation.In L. Felipe Morandé and B. Ricardo Raineri (eds.), (De) Regulation and Competition: The Electric Industry in Chile. Ilades-Georgetown University, pp. 223–257. Rudnick, H. and Raineri, R. (1997b). Transmission pricing practices in South America. Utilities Policy, 6(3), 211–218. Scheleifer, A. (1985), A theory of yardstick competition. Rand Journal of Economics, 16(3), 314–327. SEC Resolution 754, 2004. Subsecretaría de Combustibles de Argentina, 2004. Disposición 27/2004, GAS NATURAL. Approves the Rationalization Program for Natural Gas Exports and Use of Transport Capacity. World Bank (2003). Argentina crisis and poverty. World Bank Report No. 26127-AR, July 24.

Chapter 4 Electricity Liberalization in Britain and the Evolution of Market Design DAVID NEWBERY Faculty of Economics, University of Cambridge, Cambridge, UK

Britain was the exemplar of electricity market reform, demonstrating the importance of ownership unbundling and workable competition in generation and supply. Privatization created a de facto duopoly that supported increasing price–cost margins and induced excessive (English) entry. Concentration was finally ended by trading horizontal for vertical integration in subsequent mergers. Competition arrived just before the Pool was replaced by New Electricity Trading Arrangements (NETA) intended to address its claimed shortcomings. NETA cost over £700 million, and had ambiguous market impacts. Increased competition caused prices to fall, companies withdrew plant, causing fears about security of supply, but price–cost margins then increased and plant was returned to the system.

4.1. Introduction Britain’s electricity supply industry (ESI), among the first to introduce radical liberalization and re-organization, is one of the most studied models in the world. The initial England and Wales market design has been enlarged, has gone through several distinct phases since 1990. The Electricity Pool set up in 1990 was replaced by the New Electricity Trading Arrangements (NETA) in 2001, which was extended to cover Scotland under the British Electricity Trading & Transmission Arrangements from 2005, and continues to evolve as new rules are discussed and accepted. Figure 4.1 shows on a map of Britain the grid and main power stations connected at December 2004 (although the grid has hardly changed since restructuring in 1990, a large number of old coal- and oil-fired stations have closed and new gas-fired stations have opened). This chapter provides an overview of the developments of the Britain’s market, its evolution, and provides insights which may be useful to market designers in other parts of the world. The standard model of the ESI in almost every country before liberalization was an effectively vertically integrated franchise monopoly under either public ownership or cost-ofservice regulation. Investment in generation and transmission were (in theory) chosen to deliver the least-cost expansion plan (subject to government energy policy on fuel mix and 109

110


Fig. 4.1. Map of the UK with the location of the plants and transmission network. Source: NGC Seven Year Statements 2004/2005.

plant choice), financed by low-cost borrowing underwritten by the franchise revenue base. Britain was no exception, with the entire ESI under state ownership since nationalization in 1947. The Central Electricity Generation Board (CEGB) owned all generation and transmission in the whole of England and Wales, selling bulk power to 12 Area Boards, responsible for distribution and supply (retailing). In Scotland, the North of Scotland Hydro-Electric Board

Electricity Liberalization in Britain and the Evolution of Market Design

111

(NSHEB) and the South of Scotland Electricity Board (SSEB) each held regional franchises that included generation, transmission, distribution, and supply. The Government set the annual External Financial Limit restricting (publicly provided) borrowing, which in some years could be negative, implying a net dividend payment to the Treasury. The tariff structure was moderately sophisticated, with a two-part zonal Bulk Supply Tariff charging for capacity (of both generation and transmission), and variable costs (energy and regionally differentiated losses). Area Boards offered a variety of tariffs, with various forms of peak-hour capacity charges. While the pricing may have been sophisticated (although subject to government macro-economic considerations), investment planning, and particularly investment delivery, was poor, slow and costly, and there were few incentives to deliver cost efficiency. Liberalizing and restructuring the ESI was intended to replace this command and control structure with its cost-based (and often politically influenced) charges by a decentralized market-driven system that would nevertheless deliver secure, reliable electricity efficiently and at competitive prices. At the time the Government decided to restructure and privatize the ESI, there were few models available. The USA had evolved the contractual form of investorowned franchise monopolies under state cost-of-service regulation that was criticized for poor incentives, stranded investments and, in some states, high prices. Chile, on the other side of the planet, had been reforming and restructuring its ESI from 1978, and started gradually privatizing the sector from 1987. Norway already operated a spot market for wholesale energy for electricity generating and supply companies, but these remained publicly owned. With no obvious model to follow, considerable political pressure to deliver a competitive outcome (in contrast to the earlier privatizations of telecommunications and gas), and a tight timetable, the challenge was to design a set of markets and institutions to deliver these objectives. Of comparable importance, the design had to allow a smooth and predictable transition to a market-based system not just for electricity, but for the nationalized coal industry, three-quarters of whose (largely uneconomic) output was sold to the ESI, which in turn depended on coal for three-quarters of its output. The generation and distribution companies were to be sold to the general public and therefore needed predictable revenues on which they could be valued.

4.2. Restructuring and Privatization The debate on how best to restructure the ESI was vigorous, and hinged on the degree and nature of competition to be introduced. The CEGB unsurprisingly wished to remain monolithic, and argued that it would facilitate competition by new entry, but the Minister in charge of the process, Cecil Parkinson, was clear that he wanted to introduce more competition (Parkinson, 1992, Chapter 13). This required unbundling transmission from generation, splitting generation into several companies, and creating a wholesale market. In England and Wales this was achieved by separating out National Grid from the CEGB to create a transmission company covering England and Wales, but in Scotland the existing structure of two vertically integrated regional companies was retained, using the English wholesale market to introduce competition.1 Northern Ireland (restructured later) was different again in adopting the single buyer model as more suited to its isolation from the British Grid and very small size. 1

The Scottish argued that their system was smaller, with inadequate links to England, and worked well, so they argued for retaining the original structure.

112


The question of how best to introduce competition then resolved into how many generation companies to create, and this hinged on how to privatize nuclear power (and its investment program), an over-riding consideration of the Prime Minister, Margaret Thatcher.2 Privatizing nuclear power appeared to be inconsistent with more ambitious plans to divide the CEGB into as many as five to six generating companies. This led to a proposal to place the 12 nuclear stations in England and Wales with 8 GW into a large company, BigG, with the bulk of the fossil generation. The hope was that its resulting size would reduce the commercial risks attached to the nuclear assets and make the whole financially viable. It was then considered most sensible to group the remaining stations into a single company (LittleG) to countervail the might of BigG, rather than split them into several smaller companies. At a late stage the financial advisors made it clear the nuclear stations were not saleable at a positive price. The nuclear stations were transferred to Nuclear Electric and kept in public ownership. The Government thus divided the CEGB, with its 74 power stations and the National Grid, into four companies. Sixty percent of conventional generating capacity (30-GW capacity) was placed in National Power, and the remainder (20 GW) was placed in PowerGen. The high-tension grid, together with 2 GW of pumped-storage generation useful as rapid reserve,3 was transferred to the National Grid Company (NGC). The Electricity Act 1989 created the post of the Director General of Electricity Supply (the DGES), to regulate industry. This included enforcing and in due course modifying the price controls on the natural monopoly wires businesses of the NGC and the Regional Electricity Companies (RECs). He had a duty to ensure that reasonable demands for electricity were met, and that licence holders were able to finance their activities, to promote competition in generation and supply, to protect customer interests, and to promote efficiency. The Office of Electricity Regulation, Offer, was set up by the Government as an independent body under the Electricity Act, headed by the DGES. The four companies created out of the CEGB were vested (i.e. created) as public-limited companies (plcs) on March 31, 1990, at the same time as the 12 distribution companies, known as the RECs. NGC was transferred to the joint ownership of the RECs, and the RECs were sold to the public in December 1990. Sixty percent of National Power and PowerGen was subsequently sold to the public in March 1991, with the balance sold in March 1995. The pumpedstorage generation of NGC was separated and sold to Mission Energy at the end of 1995, and the RECs sold their shares in NGC in a flotation on the Stock Market, also at the end of 1995. The modern (English and Scottish) nuclear stations were finally floated as British Energy in 1996, with the rump of the original Magnox stations remaining in public ownership until decommissioned. Competition in generation was introduced, and all generators (public and private) were required to sell their electricity in a wholesale market, the Electricity Pool. The Scottish system, with about 10-GW capacity, was also restructured on March 31, 1990, when the NSHEB became Scottish Hydro-Electric, and the non-nuclear assets of the SSEB were transferred to Scottish Power. Both were privatized as vertically integrated utilities in June 1991, regulated on broadly the same basis as the industry in England and Wales. They were free to sell into the English market (and English generators were able to sell into Scotland and were entitled to access generation from the Scottish companies at the English Pool price, but in the early years prices were higher in England than in Scotland). Figure 4.2 shows the evolution of the UK electricity supply by fuel type (Britain accounted for 97.3% 2

Henney (1994) provides useful blow-by-blow account of events leading up to the restructuring decisions of 1988–1989. 3 Turbines pump water up to a hill-top reservoir during off-peak periods, allowing generation in peak periods or to provide rapid response to meet short falls in generation.


113

350

300

250

TWh

200

150

100

50

0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 Net imports

Hydro⫹other

CCGT

Other steam

Coal

Nuclear

Fig. 4.2. UK generation by fuel type 1990–2004. Source: DTI (2005).

of UK capacity in 2003, with Northern Ireland making up the rest), while Figure 4.3 shows the capacity by fuel type of England and Wales (84% of total UK capacity). In 1990, the fuel mix was overwhelmingly coal (68%) and nuclear (19%), continuing the pattern of the previous decade, with no contribution from combined cycle gas turbines (CCGTs). Other noncoal steam (mostly oil) rose to 12% in 1991, but has rapidly diminished. Coal decline rapidly until 1999 when it fell below 30%. Gas-fired CCGT’s share rose rapidly, reaching 40% by 2004. Nuclear output peaked in volume (91 TWh) and share 28% in 1998, but has since fallen to 74 TWh and 21%. Clearly there have been dramatic changes in British electricity generation over the period since privatization. The prospectus on which the companies were sold (and their licences) also set out a timetable for introducing competition into supply. At privatization, the 5000 consumers with more than 1-MW demand were free to contract with any supplier (who could buy directly from the Electricity Pool), but all other consumers had to buy from their local REC, which had a franchise monopoly. In 1994 the franchise limit was lowered to 100 kW, and another 45,000 customers were free to choose their supplier. Starting in late 1998, the remaining 22 million customers had that right, and by mid-1999 the REC franchises finally ended. Table 4.1 below provides a timetable of major milestones in the British electricity industry (as does DTI, 2005). 4.2.1. Market and institutional design One of the most interesting institutional change in restructuring the British ESI was the creation of the Electricity Pool, a compulsory bulk electricity day-ahead market that determined the merit order and wholesale spot price of electricity in Britain. This operated as a compulsory

114

Electricity Market Reform 70,000

70,000

60,000

60,000

50,000

50,000 Peak demand 40,000

30,000

30,000

20,000

20,000

10,000

10,000

MW

40,000

0

Ap

ril Ap 199 ril 0 Ap 199 ril 1 Ap 199 ril 2 Ap 199 ril 3 Ap 199 ril 4 Ap 199 ril 5 Ap 199 ril 6 Ap 199 ril 7 Ap 199 ril 8 Ap 199 ril 9 Ap 200 ril 0 Ap 200 ril 1 Ap 200 ril 2 Ap 200 ril 3 20 04

0

Oil Coal Other OCGT CCGT others CCGT National power ⫹ PowerGen Nuclear

Import

Fig. 4.3. Plant capacity connected in England and Wales, 1990–2005. Source: NGC Seven Year Statements, various years, and data from Bower and Humphries.

day-ahead last-price auction with non-firm bidding, capacity payments for plant declared available (proportional to the loss of load probability (LOLP), and thus a negative exponential function of the reserve margin), and firm access rights to transmission (with generators compensated if transmission constraints prevented their bids being accepted). Each day generators bid their plant into the pool before 10 a.m. of the day ahead, and received their dispatch orders and a set of half-hourly prices by 5 p.m. for the following day. Bids had to be valid for the 48 half-hourly periods, although generators could specify various technical parameters (minimum load, ramp rates, etc.) in some detail to force a particular pattern of use over the day, and also influence whether the plant would set the price. The half-hourly system marginal price (SMP) was the cost of generation from the most expensive generation set accepted (including start-up costs where appropriate), based on a forecast of demand and ignoring transmission constraints. Generators declared available received capacity payments and, if dispatched, the SMP, which together made up the pool purchase price (PPP). All companies buying electricity from the pool paid the pool selling price (PSP) whose difference from the PPP was the uplift, which covered a variety of other payments made to generators. The Transmission System Operator (TSO, National Grid) used the same (rather ancient) software GOAL to dispatch plant as the former CEGB. As the successor companies had copies of GOAL, they could shape the rather complex individual plant bids (start-up, no-load, and three incremental prices plus various technical parameters) to optimize their revenue, rather than bidding the true parameters. The various institutions required to manage the decentralized system were codified in the Pooling and Settlement Agreement (PSA), a multilateral contractual arrangement signed by generators and suppliers which provided the wholesale market mechanism for trading


115

Table 4.1. Major events in the British ESI. Event

Date

Comments

Electricity Act

1989

Provides legal framework for restructuring and regulation. Stephen Littlechild appointed first regulator (DGES) and sets up Offer.

Sale of RECs

1990

British ESI restructured, CEGB split up, Electricity Pool created, NGC transferred to RECs, RECs sold to public. 5000 sites above 1-MW customers free to buy in Pool.

Sale of generation

1991

60% NP and PG sold to public.

Coal Review White Paper

1993

Initial Coal contracts extended/replaced, 1990–1993 “dash for gas”, pit closures.

Sale of British Coal

1994

Coal industry sold to single private buyer.

Second-tier market

1994

100-kW market contestable (additional 45,000 eligible). Generators agree to divest plant within 2 years and accept a pool price cap.

End of Golden share

1995

Remaining 40% of NP and PG sold. Government Golden share in RECs expired, RECs subject to acquisitions.

First Price Control

1995

RECs subject to new price control, reopened after take-over wave.

NGC floated

1995

RECs float NGC, NGC’s pumped storage sold to Edison Mission.

British Energy

1996

Privatized.

Divestiture, bids

1996

Eastern leases 6 GW from NP and PG, NP’s and PG’s initial attempts to buy RECs denied.

Windfall tax

1997

£5.2 billion “excess profits tax” of privatized utilities levied by incoming Labour Government, value added tax (VAT) on energy reduced from 8% to 5%.

Regulation

1997

NGC’s price control reset, Government announces review of utility regulation.

Coal contracts end

1998

Contracts put in place in 1993 backed by REC contracts end, more pit closures, Government moratorium on gasfired generation.

Third-tier market

1998–1999

All 22 million customers contestable starting May 1998.

EU Electricity Directive

1999

EU Electricity Directive effective February, DGES report on pool prices (one of several). Merger of Offer and Ofgas into Ofgem.

Utilities Act

2000

Gas moratorium lifted November, continued generation sales.

NETA

2001

NETA introduced March 27.

British Energy collapse

2002

Low prices cause British Energy to enter administration.

Plant closures

2003

Plant mothballed following low prices, NGC predicts shortage, prices rise, plant returned.

Energy Bill, White Paper

2003

British Electricity Trading & Transmission Arrangements BETTA proposed, White paper on low carbon future.

ETS

2005

EU Emissions Trading starts January 1

BETTA

2005

British Electricity Trading & Transmission Arrangements go live April 1.

NP: National Power; PG: PowerGen.

116


electricity. It defined the rules, and required almost all parties wishing to trade electricity in England and Wales to do so using the Pool’s mechanisms. It provided the supporting financial settlement processes to compute bills and ensure payment, but did not act as a market maker. NGC owns and controls high-voltage transmission, and as the TSO was responsible for scheduling and despatch. Combining transmission with system operation has the advantage that the TSO can be provided with incentives for efficient operation that an Independent System Operator (ISO) would find too risky. With the sale of pumped storage, NGC no longer had any generation that could present conflicts of interest. NGC also acted as the Ancillary Services Provider, the Settlement System Administrator, and the Pool Funds Administrator, though again the provision of these services can be and often are separated from the provision of transmission services. In addition to the Pool, which acted both as a commodity spot market producing the reference price and a balancing market, most generators and suppliers signed bilateral financial contracts for varying periods to hedge the risk of pool price volatility. The standard contract was a Contract for Differences (CfDs) which specified a strike price (£/MWh) and volume (MWh), and was settled with reference to the pool price, so that generators were not required to produce electricity in order to meet their contractual obligations. These CfDs could be one or two sided, offering different hedging possibilities.4 Partly because the market structure was so concentrated, and partly because of the pass-through nature of the franchise contracts, other markets were slow to develop and remained very illiquid. The Electricity Forward Agreements market emerged as a screen traded over-the-counter market that allowed contracts to be traded anonymously and portfolio positions balanced. It failed to evolve into a futures market, partly because of the illiquidity caused by the large number of products (4-hourly periods for working and non-working days, for SMP, PPP, and uplift), but mainly because the underlying market was so uncompetitive. Contracts are not only important for risk sharing but were also critical in managing the transition from a franchise monopoly able to pass all its costs through to its captive customers to a market-based industry in which customers were free to buy from the cheapest supplier. The two major transitional problems facing the designers of restructuring were that British deep-mined coal was considerably more expensive than imported coal (and was soon to be revealed uncompetitive against gas), and that the CEGB had failed to set aside definable funds for decommissioning nuclear power plants. The surplus available to build up a decommissioning fund after paying for operating and fuel cycle costs were likely to be far too low given the likely equilibrium pool price. The first problem of transition was handled by a set of take-or-pay contracts between the generators and British Coal for the first 3 years at above world market prices. In parallel, the generators agreed contracts to supply the RECs for almost all their franchise output, for up to 3 years, that allowed the costs of the coal contracts to be recovered from these contract sales.5 There was the additional and very important benefit that the profit and loss accounts of the generators and RECs could be better projected, and these provided helpful financial assurance for the privatization to proceed. The second problem was dealt with by imposing a Non-fossil Fuel Obligation (NFFO) on the RECs (to buy electricity generated from non-fossil fuels, overwhelmingly nuclear 4

Over time the market developed quite sophisticated hedging instruments, for example hedging against prices above a specified strike price for the six most expensive half-hours in a month. The ability of the market to devise suitable hedges is relevant to the discussion whether regulators should insist on Reliability Options to protect consumers and reward rarely run plant (Vazquez et al., 2002). 5 The details of the various contracts required are set out in more detail in Henney (1994, pp. 120–124).


117

power), and imposing a Fossil Fuel Levy (FFL) on all fossil generation (initially at the rate of 10.8% of the final sales price). This levy was paid to Nuclear Electric to build up a fund to meet its liabilities (of about £9.1 billion, which can be compared with the privatization proceeds from selling off the CEGB of just under £10 billion). There are two routes to effective competition in generation. The first and more satisfactory route is to ensure that capacity is divided between sufficiently many competing generators that no one generator has much influence over the price.6 This option was ruled out by the government’s initial decision to create one large company in order to privatize nuclear, and by the tight Parliamentary timetable which gave too little time to reconsider plans and to divide the generation companies further once it became clear that nuclear power was unsaleable. In the early years after privatization, the two fossil generators set the pool price over 90% of the time (the balance being mostly set by pumped storage, which arbitraged a limited amount of electricity from the off-peak to the peak hours). Nuclear Electric, Scotland, and France supplied base-load power that hardly ever set the pool price. Green and Newbery (1992) calculated that a duopoly unconstrained by entry would have significant market power and would be able to raise pool prices to very high levels (shown in Fig. 4.6). The second and indirect route to competitive pricing is to induce generators to sell a sufficiently large fraction of their output under contract, and expose them to a credible threat of entry if the contract price (and average pool price) rises above the competitive level. A generator that has sold power on contract only receives the pool price for the uncontracted balance. If this is a small fraction of the total (and it is usually about 10–20%), then there is little to gain from bidding high in the pool. High bids run the risk that the plant is not scheduled, leading to the loss of the difference between the SMP and the avoidable cost. The trade-off between lost profit on uncontracted marginal plant and higher inframarginal profits becomes increasingly unattractive as contract cover increases. Contracts and entry threats are complementary – entry threats encourage generators to sign contracts, and contracts facilitate entry (Newbery, 1998a). The advantage of creating sufficiently many companies for competition is that it does not need to rely on the continued contestability of entry, and it works well even when the competitive price is well below the entry price, in periods of excess capacity. As this route was not chosen, contracts and entry threats were all that remained, at least if price regulation was to be avoided. On vesting, the three generating companies were provided with CfDs for virtually their entire forecast franchise output, most for a period of 3 years. This both managed the transition to a free market and initially reduced their incentive to exercise spot market power to negligible levels,7 though not their ability to take advantage of transmission constraints and to game capacity availability. 4.2.2. Regulation of domestic suppliers, entry, and the “dash for gas” Initially only one-third (by volume) of the market was free to buy power in the Pool or by contract from competing suppliers, and the captive customers required regulatory assurance that their prices would be reasonable. This was assured by imposing price controls on the RECs. These allowed them to pass through the regulated charges for transmission and 6

One of the best ways to measure the extent of market power is the proportion of time that one generator is pivotal, i.e. whose supply is essential to meet demand, given all other sources of supply including imports. On this basis the two fossil generators were pivotal almost all the time for the first 5 years. 7 The obligation to take large amounts of coal under take-or-pay contracts also encouraged low bidding.

118


distribution, and the costs of purchasing power subject to a licence condition to purchase this “economically”. This raised the question of how own generation should be dealt with. Many of the RECs were keen to enter into generation on their own account, or to encourage it by others, not least to compete with the incumbent generators. If restructuring to deliver lower concentration had been ruled out at Vesting, then entry was even more critical to secure eventual competition. Indeed, one of the RECs, Norweb, had already begun to build a CCGT before Vesting, and the licences allowed the RECs to enter into generation up to a specified limit (15%) of total demand within their own area. Entry of new merchant independent power producers (IPPs) would be helped by the existence of long-term contracts for gas and electricity. Although there was a preference for truly independent merchant plant,8 there was now a precedent for RECs to enter into long-term power purchase agreements (PPAs) with IPPs, and to hold some equity in the IPPs. The PPAs allowed the IPPs to sign long-term contracts for gas (usually take-or-pay) and to issue comparable duration bonds. The economic purchasing requirement was intended to reduce the risk of “sweetheart deals”,9 and, critically, the franchise would end in 1998, limiting the possible damage to captive customers. Over the next 3 years substantial entry occurred. Before the end of 1990, contracts had been signed for 2.5 GW of plant (generally of 15-year duration) and by 1993 contracts had been signed for some 5 GW of gas-fired CCGT plant. This, in addition to the incumbents’ planned 5 GW of similar plant, would displace about 25 million tonnes of coal, or nearly half the 1992 generation coal burn of 60 million tonnes. The new CCGT capacity amounted to about one-sixth of existing capacity, which was in any case more than adequate to meet peak demand (although much of it was obsolete and was planned to be replaced by the CEGB with new nuclear stations). The “dash for gas” and the switch from coal more than halved the size of the remaining deep coal mining industry. The coal labor force had fallen from nearly 200,000 at the time of the 1984–1985 coal miners’ strike to about 70,000 by 1990, but further pit closures reduced numbers to 20,000 by 1993 and less than 10,000 by 1998. Figure 4.3 shows the rapid entry of gas-fired generation, and the resulting evolution of capacity connected to the National Grid. The decline in CCGT owned by PowerGen and National Power reflects industrial restructuring discussed below. Capacity payments were made to each generating set declared available for despatch, and were equal to the LOLP multiplied by the excess of the value of lost load (VOLL, initially set at £2500/MWh and indexed to the retail price index, RPI) over the station’s bid price (if not despatched) or the SMP (if despatched). This was set the day ahead and proved manipulable by declaring plant unavailable, and then re-declaring available on the day to collect the now raised payment. This practice was investigated by the regulator and new audit procedures were agreed to reduce the incentives for mis-reporting unavailability (Offer, 1992), together with new Pool rules for computing LOLP. This was now determined by the highest declared or re-declared capacity in the current and 7 previous days, so that there was an 8-day lag between declaring a plant unavailable and its impact on LOLP. A somewhat perverse implication was that the actual LOLP could be unity (certain power

8

Enron’s 1875 MW Teeside plant was proposed in 1989, and started construction in November 1990 (Henney, 1994, p. 222). 9 The DGES issued a statement to the Secretary of State on October 17, 1990 stating (inter alia) that “where a REC had taken equity in generation projects, I would need to be assured in addition that other generators, both existing and potential, had not been overlooked or put at a disadvantage.” (Reprinted in the privatization prospectus at p. 44.)


119

cuts) while the value used to reward capacity could be almost zero. Newbery (1998c) argued that the computation of LOLP seemed excessive, given the high level of reliability over the first decade, and its overestimate may have contributed in part to the high-capacity payments. On the other hand, the VOLL seemed rather low, as Patrick and Wolak (1997) found that large consumers in one area were charged £7153/MWh in each of the three peak (or “triad”) half-hours in 1994/1995 for grid connection charges.10 As it was the product of VOLL and LOLP that determines capacity payments, these two possible errors may have been offsetting. In the 1994–1995 financial year, the generators earned £1421 million from capacity payments, or £24.5/kW/year, compared to £5821 million from selling at the SMP. Capacity payments were thus 20% of total payments for generation (excluding other ancillary services supplied by generators to the pool), and far higher than in earlier years when plant was more fully contracted. These capacity payments would have been sufficient to build 3 GW of new plant, or nearly 6% of total capacity. In the period 1995–1997, the annual average capacity payment was over £30/kW/year. During this period, the annual grid connection charge varied from £8/kW to ⫺£10/kW. The cost of keeping a new open-cycle gas turbine to provide reserve power might be £20/kW/year in interest and depreciation, and perhaps £6/kW/year for operating and maintenance (O&M) (MMC, 1996), so capacity payments should have been more than enough for security of supply. The exponential relationship between capacity payments and the tightness of demand (measured by the reserve margin) could also provide incentives for large generators to withhold plant, as Newbery (1995) demonstrated. Depending on the contract cover and plant margin, generators with a market share of about 30% might have an incentive to withdraw plant, exactly the opposite incentive to that intended. Green (2004a) examined the evidence and found that this strategy did not appear to have been significant. Later, dissatisfaction with capacity payments would be one of the factors causing the DGES to review the workings of the Pool and recommend the changes that resulted in the NETA of 2001, after which capacity payments were abolished. In addition to dispatching stations, NGC as TSO also had to resolve transmission constraints by paying out-of-merit generators to run if required (“constrained on”) or not to run (“constrained off”) in an export-constrained zone even if in the unconstrained dispatch. Under the vertically integrated CEGB the dispatch schedule automatically determined the security-constrained efficient dispatch, and the grid appeared adequately sized for such organizational form, but in the market-driven unbundled industry, the costs of resolving constraints rapidly increased (to £255 million in 1993–1994 or £4.3/kW/year). NGC offered an incentive deal to the RECs to share the benefits of reducing these and other costs, an idea that taken up by Offer, who was able to negotiate a better deal. The subsequent price control for NGC contained incentives to reduce constraint (and other ancillary service) costs, essentially by sharing the costs with a cap and collar. NGC proved adept at contracting for some plant behind constraints, making minor reinforcements to the grid, and scheduling maintenance to minimize these costs, reducing these constraint costs to less than 10% of their peak value.

10

Admittedly, these charges are not known accurately until after the peak, but large customers subscribe to moderately accurate forecasting services that can predict when prices are likely to be very high. The very low observed price response suggests that consumers value not adjusting the load in response to high prices, and by implication attach an even higher value to not losing the load.

120


Another criticism of the Pool was that it was only half a market, lacking any demand-side bidding. That was not quite correct, as NGC operated an annual tender auction for the provision of standing reserve to assist in its system management function. Standing reserve was provided by open-cycle gas turbine and pumped-storage plant, but also by demand reductions and non-centrally despatched small generators, though all had to offer amounts in excess of 3 MW. Large consumers could therefore specify their availability and willingness to reduce demand in various seasons and at various times of day, and NGC then accepted bids for which the total cost of providing load reductions were less than VOLL. In 1997/1998, 1809 MW of centrally despatched generation and 458 MW of demand modification and small-scale generation were contracted (NGC, 1997). The offer curve of such bids suggests that while there was some moderately cheap demand-side flexibility, beyond a quite modest level consumers needed a higher value than VOLL to be willing to curtail load, again suggesting that VOLL may have been underestimated, and that short-run demand elasticities for electricity were very low (with current control and metering devices).11 In addition, the Pool developed a less successful form of demand-side bidding directly into the Pool, and again its failure was an additional source of pressure to reform the Pool. The PPP determined the price of raw (unconstrained) energy and capacity, but generators and consumers are interested in the price at their location. It was appreciated that the theoretical solution to efficient spatial pricing is locational marginal pricing (LMP) developed by Bohn et al. (1984). As NGC, encouraged by Offer, explored a more satisfactory solution to the rather hastily designed system put in place at privatization, it was recognized that LMP faced a number of potentially serious drawbacks, not least of which was that its performance in the presence of considerable market power was untested (and largely unknown). The additional basis risk of trading at a large number of grid points whose price could diverge considerably from the pool price would require a large number of potentially illiquid contracts to cover risk. The possible gain in allocating the costs of transmission constraints and losses more precisely was not thought worth the loss in transparency and market liquidity. NGC therefore retained zonal access charges based on the incremental costs of reinforcing the grid to meet demands and supplies in that zone. NGC also publishes annual Seven Year Statements (looking ahead 7 years) which update predictions of demand and supply by zone, and indicate where new generation might best locate. The transmission charges are paid by consumers based on demand at the three half-hours of system maximum demand separated by 10 days (the “triad”), and by generators based on transmission entry capacity (or output in the triad if facing a negative grid charge). The more serious weakness in locational pricing was that, in contrast to the CEGB period, transmission losses were not borne by generators, distorting the merit order, while firm access rights rewarded, rather than penalizing, generators in export-constrained zones. Scotland was the obvious example of both problems, and two successive attempts by Offer to introduce transmission losses were successfully appealed to the courts. Green (2004b) estimates that nodal rather than uniform pricing in the presence of market power would have raised welfare by possibly 1.8%, which is high compared to the gains of restructuring discussed below. 11

Enthusiasts continue to believe that low-cost ICT (Information and communication technology) will enable even domestic consumers to time-shift loads such as freezers, hot water and storage heaters, and air conditioners where their thermal inertial allows electricity to be stored for modest periods in the form of heat (or cold). Evidence that this is cheaper than carrying generation reserves remains sparse, although Italy has now installed over 23 million automated meters which have the capability to manage domestic load (see Automated Meter Management Presentation to CRE, Michele Mazola, Paris, June 16, 2005, IBM).


121

4.2.3. Performance after privatization Privatization and restructuring the CEGB delivered substantial improvements in efficiency, as Newbery and Pollitt (1997) document. They estimated that after the first 5 years, costs were permanently 6% lower than under the counterfactual continued public ownership, with a present discounted value at the public sector discount rate of 6% equal to a 100% return on the sales value of £10 billion. Labor productivity doubled, real fuel costs per unit generated fell dramatically (even in the publicly owned nuclear company), and substantial new investment occurred at considerably lower unit cost than before privatization. The contrast with Scotland was striking, where a similar social cost–benefit study by Pollitt (1999) found negligible efficiency improvements. One reason was undoubtedly that the two Scottish companies were not restructured, and remained vertically integrated, making it more difficult for competitors to gain access to their home market, even though nominally Scotland was able to trade in the English Electricity Pool. Scotland was an exporter through a severely constrained interconnector that was not efficiently priced, and had only two local generators, reducing the prospects of competition. Figure 4.4 shows the average price of domestic electricity in Edinburgh, Scotland, and London, England. Initially, London was 10% more expensive than Edinburgh, but by 2001 Edinburgh was almost 10% more expensive than London. This raises the question on how access to these scarce interconnectors should be determined and priced (NGC, 2004). If full nodal pricing is thought problematic, then “market splitting”, in which the System Operator (SO) determines when constraints isolate markets, and then sets market clearing prices in each zone, as in Norway, would seem attractive. In particular, it would allow English generators to contract with Scottish consumers, and this counterflow would release more export capacity from Scotland, as the constraint only applies to net electricity flows. English generators would effectively be paid to export to Scotland an amount equal to the excess of the English marginal price over the Scottish marginal price

12 Edinburgh London

Pence/kWh (2003 prices)

11

10

9

8

7

04

03

20

02

20

01

20

00

20

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

19

19

90

6

Fig. 4.4. Domestic electricity prices at 2003 prices excluding value added tax (VAT). Source: DTI Energy Prices, various issues. Figures are averages for credit customers taking 3300 kwh/year.

122


(which should include transmission losses), and would thus be able to compete effectively in that market. Under the existing system Scottish generators could price locally up to the English Pool price, as the shadow price of the export constraint was not made explicit and they did not pay for the quite substantial transmission losses.12 In the event the British Electricity Trading and Transmission Arrangements (BETTA) took the simple (if not efficient) approach of just absorbing the interconnectors into the regulated transmission system, appointing NGC as the GB System Operator, and treating the whole of Britain as a single market (which, as noted above, in contrast to nodal pricing, is costly).13 The lesson that vertical unbundling (at least legal, and preferably of ownership) is essential for effective competition has been accepted in the new EU Electricity Directive, and in the consultation for BETTA that started in 2003. Privatization, combined with unbundling and a transparent wholesale market, provided incentives for considerable efficiency improvements, but the concentrated market structure initially allowed the incumbent generators to retain these cost reductions as enhanced profits. The social cost–benefit analysis of Newbery and Pollitt (1997) found that while the overall simple sum of net benefits of privatizing the CEGB was nearly £10 billion, consumers lost relative to the counterfactual in which fuel prices fell and the CEGB had set prices as in the past, while the owners of the generation companies gained very substantially. Unbundling the natural monopoly transmission and distribution functions also allowed incentive regulation. Transmission and distribution companies are now subject to price controls, reset every 5 years. The price cap is indexed to the RPI with a projected productivity gain, X, hence the short-hand description of RPI-X, meaning that an index of regulated prices can increase by no more than the percentage increase in RPI less X each year of the control period. The initial level of the price index, P0, is also periodically reset by the regulator, Offer (subsequently Ofgem). In contrast to tradition cost-of-service regulation, price caps offer the incentive to cut costs and keep (at least until the next price control) the resulting increased profits. Recent price controls have employed benchmarking to establish the efficient frontier, further improving incentives. Improvements in the first 5 years under the initial price controls were modest as the X factor set at privatization was unambitious. Most of the price cuts, efficiency gains, and transfers to consumers took place after the price controls were reset (and after much merger and acquisition activity). After an initial increase in controllable costs of some 15% in the first 4 years after privatization, these costs fell by 30% from 1994 to 1998, and labor productivity doubled. The efficiency gains through cost reductions and investment savings amounted to £7.1 billion at a real discount rate of 6%, assuming a counterfactual productivity growth of 2% p.a. This was the target set in the public sector although performance in the 4 years before privatization was closer to 1% p.a. The restructuring was costly (at £1.1 billion) so the net gain was £6.1 billion (rounding), comparable to the gains of restructuring the CEGB. Customers gained £1.1 billion, the Government lost about £4.5 billion in future dividends but sold the companies for £8.2 billion, while the buyers gained £9.5 billion in future dividends for their purchase (all discounting at 6%) (Domah and Pollitt, 2001). 12

Marginal transmission losses from Northern generators to the load centers were often greater than 10%. Despite various attempts and judicial review Ofgem “is of the opinion that it is not legally possible for it to approve this Modification Proposal” (to introduce cost-reflective charging for transmission losses). Ofgem’s Information Note of January 30, 2004. 13 Under the Energy Act 2004 the Secretary of State amended transmission license conditions in September 2004 to create a single GB-wide set of arrangements for trading energy and for access to and use of a single GB transmission system. This came into effect on April 1, 2005.

123

Electricity Liberalization in Britain and the Evolution of Market Design 7000

35 Restraint

30

Profit maximizing Tacit collusion

6000 Plant withdrawal

£(2001)/MWh

25 20

5000 4000

NETA

3000

15 Electricity

10

2000

Coal cost Gas cost

5

HHI

Price control

1000

Coal HHI 0 04

03

20

02

20

01

20

00

20

99

20

98

19

97

19

96

19

95

19

94

19

93

19

92

19

91

19

19

19

90

0

Fig. 4.5. Real wholesale electricity and fuel prices 1990–2005. Source: Pool data and UKPX RPD data, DTI Energy Prices for fuels, HHI from Bower and Humphries.

Figure 4.5 summarizes a long and turbulent period of pricing in the England and Wales wholesale market, during the entire life of the Pool until 2001, and under the NETA thereafter. Hourly and daily price volatility was very considerably higher than the smoothed figures shown. Averaged over the year 1997/1998, for example, the average spread between the highest and lowest half-hourly PPP prices on a day is 180% of the average price on that day, and the standard deviation of half-hourly prices over the year is 78% of the average PPP.14 Figure 4.5 shows the fuel cost of generating electricity from coal at 36% thermal efficiency and from gas at 50% gross efficiency (55% net efficiency), and hence the margin between the yearly moving average wholesale price and avoidable cost. The line with diamond markers gives on the right-hand scale the Herfindahl Hirschman index (HHI) of market concentration of coal-fired plant (for most of this period the price-setting plant). This is the sum of the squared percentage shares of available capacity, so that the initial value of just over 5000 represents the equivalent of a duopoly.15 The evolution of prices is divided into periods identified by Sweeting (2001). To understand them it is first necessary to discuss the determinants of imperfectly competitive equilibrium prices in an electricity pool. 4.3. Characterizing Market Equilibrium in a Pool Modeling price formation to understand market power and market efficiency is a challenging problem that is not yet fully solved. Green and Newbery (1992) modeled the English Electricity Pool by adapting Klemperer and Meyer’s (1989) supply function equilibrium 14

Some of this variability is predictable and can therefore be hedged. Characterizing the unpredictability of prices, which is a measure of risk, requires correcting for predictable time variations over the day, week and year. The standard deviation of the difference between the actual half-hourly price and the moving average for that hour was £14/MWh or 55% of the average price. 15 The number of equivalent firms is 10,000/HHI.

124

Electricity Market Reform Feasible supply functions Duopoly and Quintopoly 100 2-firm range 5-firm range

GBP/MWh

80

D

Demand variation

Maximum demand C

Cournot line

60

B

40 20

Marginal cost

X

A

0 0

5

10

15

20

25 GW

30

35

40

45

Fig. 4.6. Equilibrium price range ignoring entry threats and contracts. Source: Calibrated for England 1990.

(SFE) model. The model is difficult to solve and typically gives a continuum of equilibrium prices. Figure 4.6 reproduces their calibrated model for England and Wales, ignoring contracts and entry threats. This approach is attractive and appears to be supported by companies’ claims that they bid supply schedules. It assumes a single-price gross pool with bids that hold for a reasonable period of time over which demand varies, as with daily bidding in the English Pool. In its simplest form it assumes that the supply functions bid are continuous and differentiable, and that demand is linear with constant slope but varies over the 48 half-hours. Each generator chooses a supply function that maximizes his profits given the residual demand he faces, made up of the variable total demand less the total supplies bid by other generators. As his bid has to be valid over the whole daily range of residual demands, instead of choosing a single quantity to submit to the market that would determine a single price (as under the Cournot assumption), he has to choose a continuous function relating the quantity that he is willing to offer at each price realization. The set of feasible solutions will be Nash equilibria in supply functions. There are, however, difficulties with this approach. Most pools (and the English Pool in particular) restrict bids to a single price for each quantity offered, producing a step function or ladder rather than a continuously differentiable function. The Amsterdam Power Exchange is a good example, where their web site provides the bid and offer ladders for each hour, their intersection providing the market clearing price for that hour. Fabra et al. (2004) argue that this radically alters the nature of the equilibrium, and requires modeling the market as a last-price auction, following on the earlier article of von der Fehr and Harbord (1993). They solve this if there is a single period and a known inelastic demand (up to a binding price cap), but cannot characterize the solution for bids that must hold for many periods (48 in Britain) with uncertain or varying demand. Hortacsu and Puller (2004) use data that is in step function form, which they then smooth to determine the marginal revenue of the residual demand facing each generator, and demonstrate that at least for the larger companies their bids appear to be profit maximizing against this smoothed schedule. Newbery (1992)


125

suggested that if generators randomized over the positions of the steps in a step function, they could replicate a differentiable supply function, but it remains an open question whether this would be an optimal response to such behavior on the part of other generators. Standard Cournot oligopoly models are simpler, can be defended in tight market conditions, but suggest a more deterministic outcome than supply function models with their range of indeterminacy. Increasingly, consulting companies are developing price-formation models, the best of which capture the strategic aspects of supply function models with more careful modeling of the non-convexities of start-up costs which can dramatically influence the cost of providing additional power for short periods. Despite this apparent diversity of approach and the rather unsatisfactory theoretical foundations of bidding models, the evidence from various markets is consistent with the SFE story. Competition is more intense (closer to Bertrand) and prices closer to avoidable costs with spare available capacity, but as the margin of available capacity decreases, competition becomes less intense and outcomes closer to Cournot (as in the SFE). However, there remain two additional considerations before we can understand the English price evolution shown in Figure 4.5. First, while the possession of market power is legal, abusing it is not, and dominant generators need to be aware of the threat of competition references. That is the simplest explanation of incumbent bidding behavior from 1990 to 1994, where an acceptable level of prices at which to aim was arguably the entry price. The second important feature of the Pool is that it is a repeated auction, repeated every day and with the evidence of bids and outcomes available with a relatively short lag to the participants. The European Commission provides a characterization of collective dominance as a situation in which the market characteristics are conducive to tacit co-ordination and such co-ordination is sustainable, that is it is profitable and deviations can be deterred. The market characteristics that are conducive to tacit co-ordination include concentration, transparency, maturity, with a homogenous product produced by companies with similar costs and market shares, facing an inelastic demand, and with barriers to entry. Evidence supporting such a finding would include excess price–cost margins, profits, and an insensitivity of prices to a fall in cost. With the important exception of barriers to entry, the English Pool appeared to have all these defining characteristics and behavior. Tacit co-ordination was therefore to be expected, and market surveillance should clearly take account of this possibility. 4.3.1. Tacit co-ordination in the Electricity Pool Sweeting (2001) tested for tacit co-ordination by looking at individual company bids, subtracting all other bids from total demand to determine residual demand, and asking whether the bids were profit maximizing given the residual demand (but ignoring contract positions). He finds that in the first period up until 1994 both incumbents bid less aggressively than would be (short-run) profit maximizing, and that the prices were on average around the level at which entry was just profitable. If we were to conjecture on what strategies collectively dominant incumbents might co-ordinate, given close regulatory scrutiny, then keeping the price at the entry level while dividing the market in proportion to some objective criterion (such as plant capacity) would be plausible. It would also explain why both companies were keen to build new CCGTs even when their economics were marginal,16 for 16

While it is true that IPPs also entered, they did so on rather favorable long-term contracts not available to the incumbent generators. Certainly given the early gas prices and CCGT efficiencies, and compared to the opportunity cost of coal, the economics of investment were very marginal, as the House of Commons (1993) argued.

126


this would allow them to justify an increased market share (or in practice, in this prisoners’ dilemma, maintain market share in response to investment by the other company). As Figure 4.5 shows, fuel costs continued to fall away from electricity prices, but the generators nevertheless made several major increases in electricity prices, to the point that the regulator claimed that they were excessive (compared both to avoidable costs and the cost of new entry). After discussions with the regulator (and threats of a reference to the Monopolies and Mergers Commission) the companies agreed to sell 4–6 GW of plant within 2 years, in order to facilitate competition, and agreed to price caps (on both the annual average demand- and time-weighted pool price) for that period. The companies sold 6000 MW to Eastern (later TXU) with an earn-out clause of £6/MWh,17 ostensibly to compensate for the sulfur permits transferred with the plant and to reduce the buyer’s risk, but with the additional consequence of raising their rival’s marginal cost when bidding into the pool. Sweeting found that during the period from 1996 (after divestiture when the price cap ended) to 1998, each company’s bids seemed to be best (i.e. individually profit maximizing) response to those of the other companies and thus each firm was non-collusively maximizing profits. The price–cost margin increased as the regulatory threat of market abuse was replaced by (rather relaxed) competitive pressure, and the incumbents were probably quite happy to have sold plant at prices reflecting market power, in a market that was continuing to experience rapid entry. The ability to sustain a high price–cost margin depends on the volume of excess capacity, which was threatening to increase rapidly unless more coal plant were withdrawn or scrapped. At the same time Offer and Parliament (through the selection committee that investigated the energy industries) were becoming increasingly convinced that the Pool was not working well, and that the detailed rules of the dispatch algorithm were being manipulated to increase profits (Offer, 1998a–c). Henney (2001) notes other sources of discontent, notably the Labour Party’s belief that the Pool used “an operating and pricing system that was not competitive and was weighted against coal” (Robinson, 2001).

4.4. The NETA In October 1997, the Minister for Science, Energy, and Technology asked the DGES to review the electricity trading arrangements and to report results by July 1998. Offer’s terms of reference, agreed with the Government, were to consider whether, and if so what, changes in the electricity arrangements would best meet the needs of customers with respect to price, choice, quality, and security of supply; enable demand to be met efficiently and economically; enable costs and risks to be reduced and shared efficiently, provide transparency; respond flexibly to changing circumstances; promote competition in electricity markets, facilitating entry and exit from such markets; avoid discrimination against particular energy sources; and be compatible with Government policies (Offer, 1998d, pp. 83–84). The process that led to the eventual ending of the Pool and its replacement by NETA have been extensively described and criticized elsewhere (e.g. Newbery, 1998b, c). Shuttleworth (1999), writing after the publication of Offer’s Interim Conclusion (Offer, 1998d), noted that “it is difficult to find any rigorous analysis to underpin the reform proposals”, while Newbery (1998c) concluded that “(T)he present review appears to have relied mainly upon

17

That is, Eastern paid £6/MWh to the selling company for all electricity generated, increasing the effective marginal cost by that amount.


127

unsubstantiated claims, inappropriate analogies, unquantified criticisms, and a remarkably uncritical assessment by the participants of the debate, without commissioning the kind of detailed analysis one might have expected from a regulatory agency claiming industry expertise.” The Pool Review (Offer, 1998e) argued that the complexities of price formation in the Pool allowed generators to exercise more market power than would have been possible had the market been structured more like a classic commodity market. It criticized the opaque method of determining price based on a scheduling algorithm (GOAL) devised for the vertically integrated CEGB, as well as the capacity payments, and the concept of a single-price auction. It also criticized the PSA for blocking desirable changes, because as a contract between parties it could only be changed with their agreement, and, given the voting arrangements, it was rare for any change to make all parties better off. The recommendations of the Pool Review were accepted after extensive industry consultations and NETA went live on March 27, 2001. NETA replaced the PSA by a Balancing and Settlement Code with a more effective method of making modifications, giving Ofgem more influence in the process.18 The Pool ceased to exist. Electricity was now to be traded in four voluntary, overlapping and interdependent markets operating over different time scales. Bilateral contract markets cover the medium and long run, while forward markets offer standard contracts (base-load, peak hours) for periods up to several years ahead. A short-term “prompt” bilateral market (OTC and exchange), operating from at least 24 hours to Gate Closure (described below, initially 31⁄2 hours before a trading period, subsequently reduced to 1 hour in July 2002), allowed parties to adjust their portfolio of contracts to match their predicted physical positions. This short-term market would yield information to construct a spot price for each half-hour (e.g. the UKPX Reference Price Data). At Gate Closure, the official end of the bilateral markets, all parties had to announce their Final Physical Notifications (FPNs) to the SO. The SO would then accept bids and offers for balancing the system. These bids and offers would be fed into the Balancing Mechanism (BM) to produce cash-out prices for clearing imbalances between traders’ FPNs and their actual (metered) positions. The most obvious difference between NETA and the Pool is that under the Pool all generation was centrally dispatched while under NETA plant is self-dispatched. The obligation to balance output with demand is now placed on each generator, with the SO’s task confined to ensuring system stability. The Pool, that previously acted as both a wholesale market for all electricity and allowed NGC as SO to balance the system, was replaced by a bilateral market and a BM (also operated by NGC as SO) for the residual imbalances of parties that fail to self-balance. Whereas the Pool operated as a uniform single-price auction for buying and selling all power (including that needed for system balance), under NETA the market participants themselves determine the prices for the great bulk (about 97%) of all power traded. The BM is run as a discriminatory (pay-as-bid) auction for the residual amount (about 3%). Participants pay the cost of remedying any imbalances in their own positions, and NGC charges for system balancing through the Balancing Services Use of System charge. Elexon determines two cash-out prices: the weighted average of accepted offers determines the system buy price (SBP) and that of bids the system sell price (SSP). Any party found to be out-of-balance when metered amounts are compared with FPNs is charged either the SBP (if they are short, that is the FPN is more than the metered output (for a

18

Ofgem, the Office of Gas and Electricity Markets, was set up in 1999 as the successor to Offer.

128


generator) or less than metered consumption (for a consumer), or they receive the SSP if they are long (and have to spill power). The critical feature of the original design of the BM is that these prices are normally different (SBP ⭓ SSP),19 and charge each party’s imbalances, regardless of whether or not they amplify or reduce the system imbalance as a whole. Figure 4.8 below gives an indication of this volatility. As a result of the initially extreme volatility of the balancing prices and the high values taken by the SBP, a considerable number of modifications were made. One of the more important modifications (P78, shown in Fig. 4.6) made the reverse balancing price (i.e. the price facing parties who were in the opposite position to the overall market, e.g. long when the market was short, and hence aiding balance) would revert to the spot price, and hence not penalize those helping balance the system relative to their selling in the spot market. The ideas of moving to a single balancing price, or a balancing price based on marginal rather than average cost, have has been mooted but so far rejected by Ofgem. Note that there are two distinguishing characteristics of the BM, either of which could be changed independently. The first is that there are (normally) two different prices for being short or long. The second characteristic is that these prices are determined from a discriminatory auction in which bids and offers pay or are paid as bid, and the average cost of securing the services is then charged out.20 One consequence of this combination is that it is more risky for a generator to offer balancing services. If a generator has an accepted offer to increase output, and then suffers a loss of output, he is likely to have to pay more than he is paid. He may therefore prefer to retain the spinning reserve for his own insurance. A single final balancing price would make such an offer never any worse than self-insuring and normally better, and would thus promote a more liquid balancing market.

4.4.1. The evolution of market structure During the period in which NETA was under discussion, National Power and PowerGen had to decide on their future strategy in the face of considerable uncertainty about market developments and impending excess capacity. Until 1995, the RECs had been protected against take-over by Golden shares, but these lapsed and in the following few months eight of the 12 RECs were targeted by bidders. Six were successfully acquired, two by other UK regulated utilities, one by the vertically integrated Scottish electric utility, one by Scottish Power, and two by US utilities. The bids by National Power and PowerGen for two RECs were referred to the Monopolies and Mergers Commission, and then blocked by the Secretary of State. One of these RECs was subsequently bought by another US utility group. The logic of combining risky generation with the offsetting risks of supplying downstream customers was amplified by the very favorable low-debt position of these regulated utilities, and made them irresistible to the duopoly generators, but the market power of the

19

The prices were equal by about 25% of the time, and SSP exceeded SBP very occasionally (0.1% of the time) in the first 18 months. 20 The Dutch balancing market is at the other extreme. It operates a uniform price auction to determine a single price for those 15-minute periods in which the system is either long or short for the whole period, and charges those who are short while rewarding those long. There is the potential (not yet used) to add a penalty of 1 euro/MWh to both imbalances. If the system is both short and long within the 15-minute period it determines two prices, effectively one for each sub-period in which the imbalance is in one direction.


129

duopoly generators had already led the Secretary of State to reject their proposed vertical integration on one occasion.21 The obvious solution was for the companies to divest generation so that they could pass scrutiny when they bid for supply companies, and that in preparation for full retail liberalization were being unbundled from the REC distribution businesses. The urgency of achieving this objective was increased by the uncertainty over the new trading arrangements. Thus on November 25, 1998 PowerGen entered undertakings with the Secretary of State to sell 4000 MW of plant and to end the earn-out clause on its 1996 power station sales in return for clearance to acquire East Midlands Electricity’s distribution and supply business. Similarly National Power agreed to sell the 4000-MW Drax station in order to buy the supply business of Midlands Electric. The delicate task facing National Power and PowerGen was to sell the plant for attractive prices into a market that was in danger of being oversupplied with increasing gas generation. Here the new Labour Government may have inadvertently helped by imposing a moratorium on building new gas-fired plant in 1997 to assist the coal mining industry during the period of sorting out the Pool. (The Government also imposed a so-called Climate Change Levy that was actually a tax on energy rather than carbon, again protecting coal, if not British Coal.) The DTI estimated that the moratorium delayed the building of 5800 MW of gas-fired capacity (some indefinitely). The solution for the duopolists was to ensure that the price–cost margin remained high while plant was offered for sale, and Sweeting (2001) identifies the period from 1998 to early 2000 as one in which National Power and PowerGen could have increased their individual profits if they had bid lower prices. That is consistent with co-ordinating on a higher-price equilibrium than short-run myopic profit maximization would deliver. During this period plant was profitably sold, indicated by the falling HHI in Figure 4.5, and the changing shares in Figure 4.7. The companies buying the plant were warned by Ofgem that there was no guarantee that prices would remain high, particularly given the impending arrival of NETA, which Ofgem was claiming would itself lead to prices at least 10% lower than otherwise. Nevertheless, Edison Mission paid £1.3 billion for the 2000-MW stations at Fiddler’s Ferry and Ferrybridge in July 1999, or £314/kW, and increased the plant output by more than 30%. With the new buyers keen to improve the returns on their purchases by increasing plant output, Figure 4.5 shows that the earlier co-ordinated duopoly equilibrium was no longer sustainable and the price–cost margin collapsed before NETA went live, but after the fall in concentration (HHI). Edison Mission subsequently sold its two stations in October 2001 for less than half the purchase price (incurring a balance sheet impairment of $1.15 billion on the $2 billion purchase cost). Plant sales have continued apace, and DTI (2004) notes that during the Summer of 2004 a number of US firms, including AEP and Edison Mission Energy, exited, while major electricity companies (SSE, Scottish Power, Centrica) bought stations from US owners or their creditors. Nearly 8000 MW of generating capacity, around 10% of GB generating capacity, changed ownership, with the majority going to vertically integrated companies with interests in both electricity generation and supply.

21

Risks could have been hedged by long-term contracts between generation and supply companies, but the transaction costs of writing long-term contracts to cover all contingencies (such as the ending of the Pool, the Emissions Trading System, Climate Change Levy, Renewables Obligation Certificates) might make vertical integration more attractive. Supply companies also suffer from credit risk as they are typically under-capitalized unless combined with generation or distribution.

130


Offer encourages sale of 6000 MW

NETA live

35,000 PG and NP trade horizontal for vertical integration

30,000

MW

25,000 20,000 15,000 10,000 5000

April 1990 October 1990 April 1991 October 1991 April 1992 October 1992 April 1993 October 1993 April 1994 October 1994 April 1995 October 1995 April 1996 October 1996 April 1997 October 1997 April 1998 October 1998 April 1999 October 1999 April 2000 October 2000 April 2001 October 2001 April 2002 October 2002 April 2003 October 2003 April 2004 October 2004

0

ALCAN Innogy International Power

National Power (NP) Edf TXU/Eastern

British Energy Independent AES AEP Edison Powergen (PG) SS&E

Fig. 4.7. Capacity ownership of coal generation, 1990–2005. Source: Data supplied by Bower and Humphries.

4.4.2. The impact of the new trading arrangements on market performance The intellectual case for replacing the Pool through which all energy was traded by a voluntary BM covering rather less than 3% of energy was that this would force both buyers and sellers to haggle over the price of electricity without the clear and transparent signals delivered by the Pool. If parties were forced by a risky, opaque, and potentially penal imbalance market to contract ahead, consumers would have time to shop around for better deals, making the market more competitive. This argument ignores two important facts. The first is that about 90% of electricity traded before NETA was under contract, and the annual contract round was a period of intense haggling. NETA did not change that. The second is that the relative bargaining strength of generators and consumers depends more on the ability of generators to hold the market to ransom than the fine details of the spot or balancing market. This strength is best measured by the extent to which consumers can meet their total demand from all other generators, and the number of generators that are pivotal (i.e. essential) for meeting demand. The loss to a generator of not selling is the difference between the price and variable cost (shown in Fig. 4.5), whereas that to a consumer of not being able to buy power is potentially the difference between the VOLL and


131

the price, which may be hundreds of times as large. Pivotal generators therefore have very considerable bargaining power in any market design. Three factors influence this bargaining power: the number of competing generators, the reserve margin, and (in the longer run) the ease of entry. Entry was extremely easy in the Pool, but, after vertical integration and with the removal of a guaranteed market of final resort, considerably riskier under NETA. The Pool reserve margin was normally quite adequate as a result of earlier entry, while competition had just become intense as the Pool ended, and was already demonstrating its impact on spot and contract prices. The claim that NETA therefore was necessary (and sufficient) to mitigate generator market power is unsubstantiated. There was a rather more confused claim that replacing a single-price auction (like the Pool) by a pay-as-bid or discriminatory auction would obviously lower the average price, ignoring the auction literature on revenue equivalence. A more sophisticated claim was advanced by Currie (2000), who argued that repeated single-price auctions encouraged collusion more than discriminatory auctions. Newbery and McDaniel (2003) argued that the theoretical, empirical, and experimental evidence on auction design applied to the electricity market was ambiguous. Fabra et al. (2004) developed simple models comparing the two auction designs and were able to demonstrate that with predictable and unchanging demand, a discriminatory auction would yield lower (short-run) prices than a single-price auction, but would typically lead to a less efficient use of plant. They were not able to produce results for multi-period and repeated auctions. Offer (1999) estimated the costs of switching to NETA at about £700 million (spread over a 5-year period) followed by additional annual costs of £30 million.22 Offer justified this cost by claiming that prices would fall 10% as a direct result (although this would be a transfer from generators to consumers, not a net social benefit). Removing the gas moratorium helped, but in the event prices fell so far that CCGT entry was put on hold indefinitely. Ofgem was able to claim that the “Evidence of the first year of NETA shows wholesale prices around 40% below those under the former Electricity Pool.” (Ofgem, 2002). While this may exaggerate the fall, Figure 4.5 shows that compared to 1999, prices at the time of NETA were indeed substantially lower. Newbery and McDaniel (2003) argued that the price fall was due to competition, not NETA, as prices fell before NETA, and as electricity cannot be stored, future prospects of changed trading should have had no impact on pre-NETA prices. Bower (2002) (who supplied the plant data used in Fig. 4.6) demonstrated this more rigorously using formal econometric tests. Evans and Green (2003) confirmed this finding for Bower’s specification, but raised the question whether the announced end of the Pool would unravel the collusive equilibrium by backward induction for the point at which the Pool (and transparent pricing) would end. They found support for a variable “para-NETA” that takes the value 0 until October 2000, and 1 thereafter, using data up until September 2002. Prices and margins have subsequently risen, and these regressions do not perform as well on the longer time series now available.23 An alternative and simpler explanation is that indeed collusion ended, not by the anticipated arrival of NETA, but because of the actual fall in concentration, an event carried forward by the desire of the incumbents to integrate forward into supply, expressed as early as 1995. 22

“The costs of implementing and operating the new trading arrangements are estimated to be between about £136 to £146 m per annum, for a 5-year period. Thereafter the operating costs are expected to be of the order of £30 m per annum.” (Offer, 1999, p. 14). These continuing costs almost certainly understate the extra costs of maintaining 24–27 trading floors for balancing. 23 Personal communication regarding work in progress from Green.

132


4.4.3. Security of supply in an energy-only market The Pool paid increasing amounts for capacity as the reserve margin declined, to encourage generators to maintain an adequate reserve margin, either by keeping old plant open longer or by investing in new plant. Its determination was described above and implements the solution proposed by Vardi et al. (1977), itself the result of a long debate on peak-load pricing dating back to Steiner (1957), Turvey (1968), and Boiteux (1960). NETA abandoned this principle, and left generators to ensure that they contracted and bid to cover not just energy costs but also capacity costs. The central question faced by market designers is whether a liberalized competitive energy-only electricity will deliver adequate security of supply, or whether some additional mechanism, such as a capacity payment or capacity obligation, is required to support an adequate reserve margin for generation adequacy and system reliability. In the medium run the average price in any electricity market will be determined by the conditions and costs of entry and exit (and possibly on the threat of price caps or other regulatory interventions that might reduce expected profits). The first publication of the Joint Energy Security of Supply (JESS) Working Group in June 2002 (DTI, 2002) stated that “Capacity margins are healthy and are expected to remain so.” Their next report in February 2003 indicated no change. Shortly thereafter, generating companies started to experience financial distress and some went into administration. British Energy, the privatized nuclear company, had to be bailed out by the Government, and surviving companies started to scrap or mothball plant. The plant’s reserve margin fell to below the NGC’s target margin of 20%, and in the Summer of 2003, NGC’s forecasts suggested a rapidly deteriorating situation. In response, Winter 2003/2004 forward peak prices rose from £25/MWh to over £35/MWh and some mothballed plant was returned to the system. Some of the Winter month forward prices rose even more dramatically as Figure 4.8 shows. In the event the Winter was mild (only 15% of Winters in the past 75 years were as mild), demand was lower than the previous year, and there were no capacity shortages. The Third JESS Report in November 2003 (DTI, 2003) claimed that forward markets were delivering the appropriate signals and participants were responding as they should, although the scare revealed a worrying lack of information about the status and likely time needed to return mothballed plant. The impact of plant removal and narrowing reserve margins can be seen in Figure 4.5, where the price–cost margin has returned a considerable way toward a more sustainable equilibrium (although one that in 2004 was still too low to justify new build). The Fifth JESS Report of November 2004 (DTI, 2004) forecast the plant margin for Winter 2004/ 2005 at 20%, and noted that there was a further 1.2 GW of mothballed plant that could return within this Winter period if required, which would raise the plant margin to over 22%. Ofgem has argued that markets worked well and signaled impending scarcity in adequate time for the system to respond by returning plant that had been withdrawn primarily in response to the collapse in price as the markets equilibrated to their newly competitive structure. Sceptics argue that until NGC announced the impending scarcity little happened, and that 6 months’ warning is inadequate for delivering anything other than recently mothballed plant (as the longer they are idle, the more likely it is that they will be cannibalized for spares and will take longer to return to service).24 In addition, following pressures from Ofgem and analysis of its likely reserve requirements for Winter 2003/2004, NGC initiated 24

Britain may be relatively unusual in the quantity of mothballed plant that has been potentially available for return to the system. In other jurisdictions the opportunity cost of maintaining such plant, rather than securing tax deductions for abandoning plant and releasing the site for replacement construction, may militate against mothballing.


133

60 Base load November 2003 Base load January 2004 Peak November 2003 Peak January 2004

Forward price (£/MWh)

55 50 45 40 35 30 25

03 gu st 20 27 03 Au gu 16 st 20 Se 03 pt em be r2 06 00 O 3 ct ob er 20 26 03 O ct ob er 15 20 N ov 03 em be 05 r2 D 00 ec 3 em be 25 r2 D 00 ec 3 em be r2 00 3

20

Au

ly

07

Ju 18

28

Ju

ne

20

03

20

Fig. 4.8. Winter 2003/2004 forward electricity prices (£/MWh). Source: Heren data.

a Supplemental Standing Reserve Tender on October 14, 2003 to increase its reserve capacity. The tender closed on October 27, 2003, and NGT procured a total of 852 MW of Supplemental Standing Reserve at a total cost of £18.87 million or £22/kW (Ofgem, 2004a). The majority of this volume was provided from plant that had previously been mothballed and a significant amount came from the demand side. What can we conclude from this expensive change in market arrangements and to what extent has security of supply been affected? First, the costs of balancing are now directed on those that cause imbalances, and to that extent should induce more efficient responses. On the other hand, these balancing charges may now be excessive, to the detriment of nonportfolio generators (i.e. new entrants and British Energy) and intermittent suppliers like wind.25 The balancing prices are considerably more volatile and unpredictable than the pool prices that served as a more liquid balancing market. Figure 4.9 shows monthly moving averages of the buy and sell prices as well as the spot prices. The weekly or daily prices show considerably greater volatility, and a measure of this is provided in Figure 4.10, which gives the standard deviation of the difference of the spot and balancing prices for a period of a week (i.e. for the 336 observations) and a quarter (i.e. 4368 observations). The effect of the P78 rule change is very clear where the difference between the SBP and SSP falls sharply in March 2003 and remains moderately low thereafter. The impact on volatility is somewhat less pronounced but statistically significant. 25

Although the net surplus of the BM is recycled, there are transfers between different types of participants, while there are extra real costs in the dual cash-out prices which fail to confront participants with the correct costs of centrally provided balancing, especially in maintaining extra spinning reserve.

134

Electricity Market Reform 50 SBP UKPX RPD SSP SBP⫺SSP

45 40

P78 rule change

£/MWh

35 30 25 20 15 10 5

January 2002 March 2002 May 2002 July 2002 September 2002 November 2002 January 2003 March 2003 May 2003 July 2003 September 2003 November 2003 January 2004 March 2004 May 2004 July 2004 September 2004 November 2004 January 2005 March 2005 May 2005

0

Fig. 4.9. Spot and cash-out monthly moving average prices January 2002 to June 2005. Source: Elexon price data: UKPX is the Reference Price Data for the day-ahead spot market. 90 Monthly SD of SBP ⫺ RPD

80

Quarterly SD of SBP ⫺ RPD Quarterly SD of RPD ⫺ SSP

70

Incident at Damhead Creek SSP goes to ⫺£5870

£/MWh

60 50

P78 rule change

40 30 20 10 1 May 2005

1 March 2005

1 January 2005

1 November 2004

1 July 2004

1 September 2004

1 May 2004

1 March 2004

1 January 2004

1 November 2003

1 July 2003

1 September 2003

1 May 2003

1 March 2003

1 January 2003

1 November 2002

1 September 2002

1 July 2002

1 May 2002

1 March 2002

1 January 2002

0

Fig. 4.10. Standard deviation of difference of balancing price and spot price 2002–2005. Source: Elexon price data: UKPX is the Reference Price Data for the day-ahead spot market.26 26

On May 19, 2004 NGC engineers took action to prevent an unsafe situation and issued its first ever emergency instruction to stop Damhead Creek Power Station. This in turn triggered the acceptance of

135


110 NETA BM 03-4

BM 03-4 Pool 1998–1998

35

90

Pool 1997–1998

70

30

50

25 6

0

4.

5

4.

9

3.

4

2.

8

2.

3

1.

7

1.

2

0.

20

0.

0

30 0.

Cost net of avoidable cost £/MWh

40

Percent

15 10 5 0

0

10

20

30

40

50

60

70

80

90

100

Percent time cost higher than marginal cost by amount on y-axis Fig. 4.11. Cost of 24-hour failure under the Pool and NETA. Source: Pool data and UKPX RPD data.

The costs of balancing will depend on whether the participant is a generator or supplier. As contestability is a key issue, the relevant question is whether the BM unreasonably raises the risk to a small entrant. This can be estimated as the risk of having to pay the buy price (SBP) after a generator suffers a forced outage, in order to meet an assumed contract position. If a generator fails at a random moment and stays off-line for 24 hours, the cost will be the 24-hour average of the SBP from that moment. In the year before the P78 rule change indicated above, the expected cost of such an outage (relative to an assumed variable cost of £12/MWh) was £17/MWh or £0.4/kW/event compared to £13/MWh or £0.32/kW/event under the Pool for 1997–1998. The variance was, however, twice as high as under the Pool. In the year following P78, the average cost had fallen to £11/MWh or £0.3/kW/event and the variance had also fallen to 150% that of the Pool. Figure 4.11 illustrates the cost duration curve for balancing under NETA from April 1, 2003 to March 31, 2004 compared to the Pool in 1997–1998. Thus 5% of the time the cost would be £30/MWh for the following 24 hours in both the Pool and the recent BM, and 1% of the time it would be £70/MWh in the BM compared with £44/MWh under the Pool. The risks in the early days of NETA were very much higher and led to claims that plant was inefficiently part loaded to avoid penal imbalance costs, at considerably higher cost. One should interpret this finding with some care, as the Pool required bids to remain valid for 24 hours while bids and offers to the BM can be changed on a short time scale and in response to a perceived tightening of the market when a large unit goes off-line, making it more risky for generators to handle outages. Even if we ignore such responses, if a large plant were to go down, the demand in the BM might be such as to considerably increase the 26

(continued) a large proportion of bids at ⫺£9999/MWh, causing the SSP to go to ⫺£5,728 in HH 28 at a cost of £3.55 million (Elexon Circular EL01201 of August 20, 2004). This had a huge impact on the standard deviation of prices for the whole of the 30-day window of the moving average, swamping all other prices, as can be seen.

136


short-run cost, but without knowing the shape of the bids and offers it is hard to estimate by how much. The effect on suppliers is that they would over-contract on average as the SBP is more costly than spilling the surplus at the SSP, and this would (slightly) raise the cost and risk of selling. This may have the desirable effect of encouraging contracting, which tends to mitigate generator market power in the spot market, although at the expense of increasing demand and hence market power in the contract market. The low liquidity of most electricity markets (and certainly the British markets) makes the cost of rebalancing contract positions high and again acts as an entry barrier. Second, the BM mutes scarcity signals by paying generators their bid price and not the marginal price (in order to mitigate market power and reduce possibly politically unacceptable price spikes). This, together with the lack of integration with spot, forward and contract markets, and the lack of any capacity payment, makes the entry decision more uncertain and risky, and may lead to lower reserve margins. If so, then the lower reserve margins and the extra entry costs will allow a higher-average wholesale price, refuting Ofgem’s claim that NETA alone (i.e. regardless of market structure) would reduce wholesale prices by 10%. Again, this claim needs to be examined carefully, for example, for a peaking generator that can offer balancing services of a half-hour duration at very short notice. Figure 4.12 shows the annual profits that a peaking unit (with fast start-up able to offer half-hourly supplies and selling into the balancing market at the SBP) would earn at different prices for its variable costs. As an indication, the June 2005 cost of distillate at 32% efficiency would have been about £75/MWh, ignoring the (high) fuel excise, but less than half that running on natural gas. One obvious problem with this calculation is that it requires skilled bidding to achieve the SBP in a pay-as-bid market. The graph demonstrates that profits are sensitive to the fuel cost and balancing prices (which are likely to be related). For example, taking the daily British gas prices (spot at the National Balancing Point), the profit of a 32% efficient opens-cycle gas turbine in 2002/2003 would have been £248/kW/year. The following year when gas prices rose and balancing prices were less volatile, the profit would only have been £21/kW/year, but in 2004/2005 the profit would again have risen to £145/kW/year. The annual fixed costs would include grid charges (which vary across the

200 2002/2003 2004/2005

180 160 £/kW/year

140 120 100 80 60 40 20 0 20

30

40

50

60

70

80

90

100

110

120

Variable cost £/MWh Fig. 4.12. Profit of a peaking plant selling into the BM at SBP. Source: Elexon price data.


137

country) from ⫹£9/kW/year to ⫺£7/kW/year) and other fixed costs (perhaps £6/kW/year), as well as capital charges of perhaps £36/year, suggesting a requirement of £50/kW/year. This suggests that peaking investment can be remunerative under NETA, but is risky, partly because of fuel and electricity price uncertainty, and partly because of the problem of bidding. This could be circumvented by offering such services to NGC on contract to bid into the BM, and/or making the BM a single marginal price market. NGC simulated the effect of moving to a marginal price balancing market for 2003/2004, although ignoring the likely response of participants to such a significant change in market rules (NGT, 2004). The average SBP would have increased by £2.78/MWh and SSP decreased by £1.80/MWh. The standard deviation of SBP would have increased by £14.39/MWh indicating an increase in volatility, whilst the increase in standard deviation of SSP was much smaller at £1.23/MWh.27 This increased volatility would have improved the profits of peaking plant (while perhaps increasing its risk somewhat). Critics of energy-only markets argue that the new system has not been properly stress tested yet, and that reliability either requires a sufficiently concentrated market to ensure high enough prices to cover full costs, or some supplementary mechanism to reward capacity and encourage investment in a timely manner in advance of possible system scarcity. Thus the US Federal Energy Regulatory Commission (FERC) in its 2002 Notice of Proposed Rule Making on the Standard Market Design stated: “FERC is concerned that the spot market alone will not signal the need to bring development of new supply resources in time to avert a shortage … .” The British approach is to argue that provided markets are allowed to clear and are not subject to an unreasonable price cap,28 and provided the TSO is incentivized to maintain system reliability, a competitive energy-only market should work well without additional mechanisms.29 The argument is that the spot and balancing market should signal short-run scarcity and force consumers to contract ahead, thus feeding scarcity signals back into forward and contract markets. As already argued, this would be better achieved by a marginal price balancing market than the British design of an average price of the offers as paid. The main concern is whether investors will look far enough ahead to make timely investments. In a tight oligopoly, competition for market share and/or high entry-inducing prices should provide adequate incentives. If the market is as competitive as in Britain then market risk is more likely to lead to delays in exercising the real option of investment as the preemption effect will be lower. Some (e.g. de Vries and Hakvoort, 2002; de Vries and Neuhoff, 2004) argue that generation investment requires long-term contracts like the earlier PPAs in the Pool, and/or a captive franchise market on which to write them. Others (e.g. de Luze, 2003) are concerned that generating companies’ recent financial distress (and banks who might provide debt) will be reluctant to invest until profits significantly improve.

27

NGT simulation’s results should be interpreted with care as they rely on the strong assumption that generators and suppliers would not have changed their bidding behavior into the market; there is a large literature demonstrating the impact of market rules on market players bidding strategies. 28 It is reasonable to cap spot markets by the VOLL, and the BM in Britain can only enter four digits, making the maximum allowed price £9999. 29 Part of the problem in the USA is that the ISOs are prone to take “reliability actions”, including out-ofmarket calls, as supplies get tight that have the effect of depressing prices. It is admittedly difficult to ensure that balancing prices (and local spot prices) efficiently reflect scarcity and not market power. There is also the different institutional history of managing privatized electricity utilities in Britain and the US that may partly explain differences in the confidence which regulators place in market solutions to generation adequacy.

138


However, most generating investment in Britain is likely to be undertaken by the large Continental companies already present – EdF, RWE, and E.On, who are also vertically integrated into supply and hence well hedged in the wholesale market. Such companies are unlikely to rely on short- to medium-run forward contracts in making investment decisions that last 30–40 years, and will base their decision on market fundamentals. Although the British market is competitive, these players have ambitions to retain their Continental position with high levels of output, and seem likely to invest if the fundamentals warrant. The early complaints of wind generators and combined heat and power (CHP) that they were discriminated against have not been adequately tested against more recent market conditions – certainly CHP output dropped dramatically but that was arguably because of the adverse spark-spread (electricity less gas cost). Wind power now typically sells on contract to supply companies who can better manage the imbalance risk within their entire portfolio, and is in any case massively rewarded by Renewable Obligation Certificates that increase the electricity price from around £25/MWh to £65/MWh. Perhaps the greater concern is that such a high level of renewables support may not be credible in the longer run, and this policy uncertainty may make investment in all forms of generation riskier.

4.5. Retail Competition Full retail competition was envisaged right from the start of restructuring, with a phased opening, initially for the 5000 or so customers of above 1-MW demand, then in 1994 for the 50,000 sites with above 100-kW demand, and finally the aim was complete liberalization by 1998. By offering customers to freedom to switch supplier, suppliers would be put under pressure to cut margins, offer tariffs and contracts suited to the customer, and in turn bid more aggressively in the wholesale market, and the ability of the regulator or government to influence contracting and generation decisions would be reduced. Large customers mostly switched from their local RECs to the two incumbent fossil generators, National Power or PowerGen. The size of the competitive market increased in April 1994, when the 50,000 sites with demands of between 100 kW and 1 MW were allowed to change their supplier, and required to install a half-hourly meter. The RECs were initially slow to enter the second-tier market to supply large customers,30 but thought they could (and would have to) compete more effectively for smaller customers. One-quarter of the RECs entered the 100-kW second-tier market in the first year, and the share increased steadily thereafter. By 1996, they (collectively) supplied more than half the sites taking a competitive supply, with greater success in the 100-kW market. Offer’s early survey of suppliers indicated that competition had been effective even among the smallest sites in this market, those with demands of under 300 kW, for one-quarter of them bought from a second-tier supplier in 1995/1996. By 2002, 62% of half-hourly metered customers had switched at least once (Ofgem, 2003, Table 5.1). Price controls in this market were lifted in 1994, when competition was introduced, and the ability to change supplier seems to have been sufficient protection for these consumers. If competition was successfully introduced, the process itself was fraught with difficulty, as significant changes to systems and procedures were needed to cope with the larger number of customers (Green and Newbery, 1997). Many important decisions were not made 30

First-tier customers buy from the local (originally franchise-holding) REC, second-tier customers buy from other suppliers, so a REC selling out of area is a second-tier supplier. This distinction was removed in the Utilities Act of 2000.


139

early enough to allow systems to be developed and tested. Installing the required meters and communications link was chaotic, and many were either not installed in time, or installed but not properly registered. It took more than a year to sort out some of the problems, amid acrimonious debates over who had caused the chaos.31 Since then competition appears to be developing well, measured by periodic surveys published on the Ofgem web site (e.g. Ofgem, 2003), despite recent consolidation and some increased concentration in the supply market (with an HHI of 1670 and the top six suppliers having 90% of the market at the end of 2002). Ofgem’s (2000) review noted concerns about potential entry barriers caused by a lack of unbundling of distribution and supply, but legal unbundling was subsequently required by an amendment in the Utilities Act 2000. Full retail liberalization started in September 1998 and was completed by June 1999, with a near-90% consumer awareness of the opportunity to choose (Littlechild, 2001). Ofgem (2004b) reviewed the experience, noting that about half of domestic customers had switched at least once (of whom 80% moved to a supplier with a “dual fuel” offer). British Gas, the original gas franchise holder, had 44% of the dual fuel market despite being the most expensive (for gas). Incumbents maintained prices in their original franchise area substantially higher than they offered to new customers out of area, and genuinely new entrants (those without an existing franchise business) collectively secured less than 1% of the market, most of them going bankrupt or selling out to incumbents. The incumbent Eastern/TXU also went bankrupt and many RECs sold their supply businesses. Prices fell, but not as much as costs, with incumbent supply margins widening substantially (up to 26% of domestic bills), first under generous price caps on within-area retail electricity designed not to stifle competition, and subsequently, when price controls were removed, because of the stickiness of customers. Salies and Waddams Price (2004) estimated that the incumbency effect raised prices 4–13%. By 2002 a domestic customer cost about as much on the stock exchange as 2 kW of generation (rather more than the capacity needed to supply him/her). Ofgem (2004b) noted that the market was “competitive but not yet mature”, and by comparison with most other examples of electricity liberalization, Britain has been judged a success, notably as a high proportion switched suppliers. Nevertheless, Green and McDaniel (1998) noted that retail liberalization had cost £276 million to set up with on-going costs estimated at £36 million/year, while Salies and Waddams Price (2004) concluded that overall the net social benefits of liberalization were negative. Taking a broader view, retail liberalization allows the regulator to stop interfering in this part of the market, and transfers the risks of contracting to suppliers, who can no longer pass through “sweetheart” deals to a captive supply base. Ministers and politicians would find it harder to protect the coal industry with coal-backed electricity contracts passed through to the franchise market, as in 1993–1998, and the ability of the Government to protect renewables at the expense of consumers has required an Act of Parliament. Pessimists (de Vries and Neuhoff, 2004) have argued that the resulting lack of long-term contracts backed by a captive market makes entry into generation more difficult, and might reduce security of supply. On the other hand, the high profits in retailing have compensated 31

Blame was heaped on the electricity companies, for delaying decisions; the regulator, for introducing competition in metering services at the same time; and the meter operators, for failing to install meters when promised (even those not wishing to change supplier required a half-hourly meter and communications equipment). None of the parties faced financial penalties for failure to meet their obligations (although this was apparently considered by Offer), and the industry recovered part of the additional cost of sorting out the disruptions from a levy on consumers.

140


to some extent for the losses caused by the margin collapse in the wholesale market (and vice versa when wholesale prices rise), and vertically integrated companies are effectively automatically contracted when customers are as sticky as domestic households. Vertical integration, provided it is combined with adequate competition, may provide the financial stability and market assurance needed to ensure adequate investment in generation. Most domestic customers would probably be better off with a regulated supply margin and benchmarked contract costs passed through under regulatory supervision. In short, it is difficult to see why retail liberalization attracts so much support in the new Electricity Directive that compels full market opening by 2007 (other than the reasonable assumption in some jurisdictions that regulators would be less effective at protecting consumer interests than competition). 4.6. Conclusions The British experiments have demonstrated a number of important lessons for electricity market liberalization. First, ownership unbundling of transmission from generation helps support a competitive wholesale market, which in turn puts pressure on companies to reduce costs. Scotland failed this test and failed to improve its performance. Second, efficient pricing of scarce interconnector capacity and charging correctly for losses might have allowed the Scottish market to import English competition, but was blocked by the courts. Third, while competition drives down costs, concentrated markets can sustain inefficiently high price–cost margins. Pivotal generators retain market power that is best addressed by reducing concentration, although entry that increases the reserve margin also helps. Tacit co-ordination is likely given electricity market characteristics, and is best addressed by encouraging contracts and entry, and reducing concentration. Fourth, investment in generation can be facilitated by a transparent Pool, domestic franchises, and wholesale market power. Britain replaced nearly one-third of its (already adequate, if rapidly obsolescing) capacity by such means. Entry (including returning mothballed plant to service) is responsive to price signals (the forward spark-spread). Fifth, unbundling and liberalization increase risk for generators and encourage them to seek vertical integration with suppliers. This offers the opportunity for the regulator and competition authorities to trade horizontal for vertical integration and to reduce concentration, at the cost of increased entry barriers. A better alternative is to start from a more fragmented structure. That would allow one to consider legal restraints on such vertical integration to encourage more contracting and market liquidity, but we lack evidence on the costs and advantages of such enforced competition. Sixth, retail liberalization has delivered significant benefits to large and medium customers, although managing the market opening for more than the largest customers is challenging and can be expensive if all customers need interval meters and communications equipment. There is little evidence that full retail liberalization delivers positive net social benefits in well regulated jurisdictions, although it has enabled vertically integrated companies to hedge their wholesale market position with sticky downstream customers, enabling them to continue to finance investment when needed. Seventh, the British approach to liberalization that requires licences to be held by both potentially competitive and natural monopoly segments has worked better than many of the Continental alternatives (and arguably the USA’s onerous duty on regulators to deliver “just and reasonable” prices). Licences require the holders to provide the regulator with the information needed for adequate market monitoring, and allow market abuses to be swiftly and cheaply addressed.


141

Finally, such apparently basic issues as the desirability or not of capacity payments (or obligations) and the design of the wholesale and balancing markets remain unresolved. NETA abolished capacity payments, and relied upon the BM to signal shortages of supply, which would feed back into spot and contract markets. The ideal of a Pool with adequate competition, capacity payments, and a better governance structure for rule changes was never tried, and might have worked as well or better than NETA, with its emphasis on bilateral contracting and dual cash-out pricing in the BM. On balance, NETA replaced the Pool’s flawed governance structure by one more susceptible to incremental improvement (though at the cost of greater regulatory uncertainty) failed to increase either the liquidity of markets or the participation of the true demand side, increased trading costs, and cost over £700 million. Once it settled down and the obvious rule changes were made, NETA probably delivers similar outcomes as the Pool from existing generation. Entry is now more difficult than before, but that is not solely due to NETA. Vertical integration has reduced the demand for suppliers to contract, the end of the domestic franchise has removed the logical counterparty to contracts with new independent generators, but the removal of the Pool as a market of last resort almost certainly raises entry costs. Just at the time that FERC has embraced the concept of a Pool (with LMP) as the benchmark for the Standard Market Design, Britain has abandoned a model whose main failing was its poor market structure and governance. Nevertheless, NETA (or its successor BETTA) may not be very different from the Pool in sustaining a reliable and secure electricity market, provided the TSO is charged to deliver that reliability, if necessary by contracting ahead and making sure that unbiassed and trustworthy information about future demand and supply is fed to the markets.

Acknowledgments I am indebted to Paul Joskow, Richard Green, Alex Henney, Karsten Neuhoff, and Fabien Roques for their help, and particularly to Stephen Littlechild for his detailed comments, even where I have not accepted them, with the usual disclaimer. This chapter is a modified version of a article originally published in the 2005 Special Issue of the Energy Journal on European Electricity Liberalization. I am indebted to the IAEE for permission to reproduce the article here, and to the UKPX for their spot price data.

References Bohn, R.E., Caramanis, M.C. and Schweppe, F.C. (1984). Optimal pricing in electrical networks over space and time. Rand Journal of Economics, 15(3), 360–376. Boiteux, M. (1960). Peak load pricing. The Journal of Business, 33(2), 157–179. Bower, J. (2002). Why did electricity prices fall in England and Wales? Market mechanism of market structure?, Oxford Institute for Energy Studies Working Paper EL O2. Currie, D. (2000). The new electricity trading arrangements in England and Wales: A review. The Beesley Lectures on Regulation. The Institute of Economic Affairs and the London Business School, 10 October 2000. de Luze, G. (2003). Investment in power generation: a banker’s perspective. Presentation at the 25th March 2003 IEA/NEA Conference in Paris. de Vries, L. and Hakvoort, R. (2002). The question of generation adequacy in liberalised electricity markets. Delft University Working Paper. de Vries, L. and Neuhoff, K. (2004). Insufficient Incentives for Investment in Electricity Generation, Cambridge, CMI EP 42, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm

142


Domah, P.D. and Pollitt, M.G. (2001). The restructuring and privatisation of the regional electricity companies in England and Wales: a social cost benefit analysis. Fiscal Studies, 22(1), 107–146. DTI (2002). Joint Energy Security of Supply Working Group First Report, June, available at http://www.dti. go.uk/energy/jess DTI (2003). Joint Energy Security of Supply Working Group Third Report, November, available at http:// www.dti.go.uk/energy/jess DTI (2004). Joint Energy Security of Supply Working Group Fifth Report, November, available at http://www.dti.go.uk/energy/jess DTI (2005). Digest of UK Energy Statistics 2004, available at http://www.dti.gov.uk/energy/inform/ energy_stats/electricity/index.shtml Evans, J. and Green, R. (2003). Why Did British Electricity Prices Fall After 1998?, CMI WP26, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm Fabra, N., von der Fehr, N.-H. and Harbord, D. (2004). Designing Electricity Auctions, CSEM WP 122, available at http://www.ucei.berkely.edu/pubs-csemwp.html von der Fehr, N.-H. M. and Harbord, D. (1993). Spot market competition in the UK electricity industry. Economic Journal, 103, 531–546. Green, R. and McDaniel, T. (1998). Competition in electricity supply: will ‘1998’ be worth it? Fiscal Studies, 19(3), 273–293. Green, R. and Newbery, D.M. (1997). Competition in the electricity industry in England and Wales. Oxford Review of Economic Policy, 13(1), 27–46. Green, R.J. (2004a). Did English Generators Play Cournot? Capacity Withholding in the Electricity Pool, CMI WP41, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm Green, R.J. (2004b). Electricity Transmission Pricing: How Much Does it Cost to Get it Wrong, CMI WP63, available at http://www.econ.cam.ac.uk/electricity/publications/wp/index.htm Green, R.J. and Newbery, D.M. (1992). Competition in the British electricity spot market. Journal of Political Economy, 100(5), 929–953. Henney, A. (1994). A Study of the Privatisation of the Electricity Supply Industry in England and Wales. EEE Ltd, London. Henney, A. (2001). The illusory politics and imaginary economics of NETA, Power UK, March, 16–26. Hortacsu, A. and Puller, S.L. (2004). Testing Strategic Models of Firm Behavior in Restructured Electricity Markets: A Case Study of ERCOT, CSEM WP 125, available at http://www.ucei.berkely.edu/pubscsemwp.html House of Commons (1993). British Energy Policy and the Market for Coal: Report HC 237. HMSO, London. Klemperer, P.D. and Meyer, M.A. (1989). Supply function equilibria in oligopoly under uncertainty. Econometrica, 57(6), 1243–1277. Littlechild, S. (2001). Electricity regulation (Chapter 2). In P. Vass (ed.), CRI Regulatory Review 2000/ 2001. Centre for the Study of Regulated Industries, Bath, pp. 25–62. MMC (1996). National Power Plc and Southern Electric Plc: A Report on the Proposed Merger. Monopolies and Mergers Commission Cm 3230. HMSO, London. Newbery, D.M. (1992). Supply function equilibria, mimeo, Cambridge University. Newbery, D.M. (1995). Power markets and market power. Energy Journal, 16(3), 41–66. Newbery, D.M. (1998a). Competition, contracts and entry in the electricity spot market. RAND Journal of Economics, 29(4), 726–749. Newbery, D.M. (1998b). Pool reform and competition in electricity (Chapter 5). In M. Beesley (ed.), Regulating Utilities: Understanding the Issues. Institute of Economic Affairs, London, pp. 117–166. Newbery, D.M. (1998c). The regulator’s review of the English Electricity Pool, Utilities Policy, 7(3), 129–141. Newbery, D.M. and McDaniel, T. (2003). Auctions and trading in energy markets – an economic analysis (Chapter 10). In P. Vass (ed.), CRI Regulatory Review2002/ 2003. Centre for the Study of Regulated Industries, Bath, pp. 195-–234. Newbery, D.M. and Pollitt, M.G. (1997). The restructuring and privatisation of the CEGB – was it worth it. Journal of Industrial Economics, XLV(3), 269–303. NGC (1997). Report on the Tender for the 1997/98 Standing Reserve Service. National Grid Company plc.


143

NGC (2004). National Grid Company Consultation on Options for Allocating GB Transmission Access Rights under BETTA, 21 May 2004, Warwick. Available at http://www.ofgem.gov.uk/ NGT (2004). Response to OFGEM Consultation, Electricity and Gas Cashout Review, Executive Summary. National Grid Transco, Warwick. Offer (1992). Review of Pool Prices. Office of Electricity Regulation, Birmingham, December. Offer (1998a). Review of Electricity Trading Arrangements: Background Paper 1 – Electricity Trading Arrangements in England and Wales. Office of Electricity Regulation, Birmingham, February. Offer (1998b). Review of Electricity Trading Arrangements: Background Paper 2 – Electricity Trading Arrangements in other Countries. Office of Electricity Regulation, Birmingham, February. Offer (1998c). Review of Electricity Trading Arrangements: Working Paper on Trading Inside and Outside the Pool. Office of Electricity Regulation, Birmingham, March. Offer (1998d). Review of Electricity Trading Arrangements – Interim Conclusions. Office of Electricity Regulation, Birmingham, June. Offer (1998e). Review of Electricity Trading Arrangements. Office of Electricity Regulation, Birmingham, July. Offer (1999). The New Electricity Trading Arrangements, Vol. 1. Office of Electricity Regulation, Birmingham, July. Ofgem (2000). A Review of the Development of Competition in the Industrial and Commercial Electricity Market. Office of Gas and Electricity Markets, London. Ofgem (2002). Annual Report. Office of Gas and Electricity Markets, London. Ofgem (2003). Review of Competition in the Non-domestic Gas and Electricity Supply Sectors. Office of Gas and Electricity Markets, London. Ofgem (2004a). Security of Supply April to September 2003, Six Month Retrospective Report. Office of Gas and Electricity Markets, London, February. Ofgem (2004b). Domestic Competitive Market Review 2004: A Review Document. Office of Gas and Electricity Markets, London, April. Parkinson, C. (1992). Right at the Centre. Wiedenfield and Nicolson, London. Patrick, R.H. and Wolak, F.A. (1997). Estimating the customer-level demand for electricity under realtime pricing. Paper Presented to the POWER conference, March, Berkeley, University of California Energy Institute. Pollitt, M.G. (1999). The Restructuring and Privatisation of the Electricity Supply Industry in Scotland, mimeo, Cambridge, available at http://www.econ.cam.ac.uk/electricity/people/pollitt/esiscotland.pdf Robinson, G. (2001). The Unconventional Minister: My Life Inside New Labour. Penguin, London. Salies, E. and Waddams Price, C. (2004). Charges, costs, and market power: the deregulated UK electricity retail market. The Energy Journal, 25(3), 19–37. Shuttleworth, G. (1999). The Electricity Industry 1997–1998 (Chapter 2). In P. Vass (ed.), Regulatory Review 1998/1999. CRI, University of Bath, pp. 19–34. Steiner, P. (1957). Peak loads and efficient pricing. Quarterly Journal of Economics, November. Sweeting, A. (2001). Market Outcomes and Generator Behaviour in the England and Wales Wholesale Electricity Market 1995–2000, mimeo, MIT. Turvey, R. (1968). Optimal Pricing and Investment in Electricity Supply. MIT Press, Cambridge. Vardi, J., Zahavi, J. and Avi-Itzhak, B. (1977). Variable load pricing in the face of loss of load probability. Bell Journal of Economics, 8(Spring), 270–288. Vazquez, C., Rivier, M. and Perez-Arriaga, I.J. (2002). A market approach to long-term security of supply. IEEE Transactions on Power Systems, 17(2), 349–357.


Chapter 5 The Nordic Electricity Market: Robust by Design?1 EIRIK S. AMUNDSEN,1 LARS BERGMAN,2 AND NILS-HENRIK M. VON DER FEHR3 1

Department of Economics, University of Bergen, Bergen, Norway and Institute of Food and Resource Economics, The Royal Veterinary and Agricultural University, Copenhagen, Denmark; 2 Stockholm School of Economics, Stockholm, Sweden; 3Department of Economics, University of Oslo, Oslo, Norway

One of the desirable characteristics of a competitive market is how well it can handle stress, arising from natural or man-made causes. In the case of electricity markets, natural causes, including abnormal weather patterns and/or droughts, particularly in systems, which are predominantly hydroelectric. The Nordic market experienced and survived a severe hydro shortage during 2002–2003. This chapter provides an overview of the Nordic market, focusing on its resilience to handle natural calamities, in contrast to markets such as California, which did not behave well during a similar stressful situation in 2000–2001. Main conclusions are that fears regarding supply security and adequacy are likely to be unfounded. Nevertheless, as inherited over-capacity is eroded, and new market-based environmental regulation takes effect, tighter market conditions are to be expected. It is then crucial that retail markets are fully developed so as to allow consumers to adequately protect themselves from occurrences of price spikes. 5.1. Introduction The Nordic electricity market – encompassing Denmark, Finland, Norway and Sweden2 – is well established by now. Starting in Norway in 1991, regulatory reform gradually spread to Sweden (1996), Finland (1997) and Denmark (2002). In all countries, separation of competitive and monopolistic activities, establishment of independent Transmission System 1

This chapter is an updated version of an article titled: The Nordic market: signs of stress, published in 2005 in the Special Issue of The Energy Journal on European Electricity Liberalization (von der Fehr et al., 2005). The authors are indebted to the International Association for Energy Economics for allowing reuse of this material in the current chapter. The contributions of Perry Sioshansi and Erling Mork, formerly with Nord Pool, are gratefully acknowledged. Thanks are due to Elforsk AB, Market Design and to Renergi, The Research Council of Norway for financial support of work resulting in parts of the present chapter as well as in the aforementioned article. 2 The fifth Nordic country–Iceland–is not interconnected with the others. The term “Scandinavia” is inappropriate, as it only encompasses Denmark, Norway and Sweden.

145

146


Table 5.1. Background data of the Nordic electricity market. Denmark Installed capacity (MW) Total Hydropower Nuclear power Thermal power Wind power

12,830 11 0 9704 3115

Generation mix Hydropower (%) Nuclear power (%) Thermal power (%) Wind power (%) Consumption and population Total consumption (TWh) Total generation (TWh) Population (million)

Finland

Norway

Sweden

Sum

16,893 2978 2640 11,225 50

28,081 27,676 0 305 100

33,361 16,143 9441 7378 399

91,165 46,808 12,081 28,612 3664

0 0 76 24

18 16 66 0

99 0 1 0

48 28 22 1

51 13 31 4

35.2 43.8 5.387

84.9 79.9 5.220

115.0 107.1 4.565

145.5 132.5 8.976

380.6 363.3 24.148

Maximum system load 2003 (MWh/h) Denmark – West 3780 Denmark – East 2665 Finland 14,040 Norway 19,984 Sweden 26,400 Source: Nordel 2003 Annual Report.

Operators (TSOs) and allowing consumers to choose their supplier have been integral parts of reform. With a long tradition of Nordic co-operation, and with development of the jointly owned power exchange, Nord Pool, the Nordic market is now de facto fully integrated, at least at the wholesale level.3 The Nordic market consists of over 91,000 MW of installed capacity, roughly half of which is hydroelectric, notably in Norway. Statistical details are provided in Table 5.1. Figure 5.1 shows the installed capacity in various countries. By way of background, the Nord Pool’s physical market is a day-ahead spot market where both producers and consumers/distributors submit bids for purchase and sale of electricity for each hour of the coming 24-hour period. These bids are then matched and separate hourly prices are calculated for the next day (midnight to midnight). Capacity constraints are reported by the four TSOs in the Nordic region: Energinet.dk in Denmark (merged from Eltra in West Denmark and Elkraft in East Denmark), Fingrid in Finland, Svenska Kraftnätt in Sweden and Statnett in Norway. Firstly a “system price” is calculated disregarding capacity constraints. This price is used as a reference for cash settlement in the financial derivative market. Then transmission capacities are taken into account. If there are no bottlenecks, then the price for all areas is the same as the system price. However, if there is a bottleneck, then some or all areas may have different prices. Norway is divided into two or more areas (set dynamically by Statnett, the Norwegian TSO), while Sweden, Finland, West and East Denmark are fixed areas. (Note that East and West Denmark have no direct 3

For a detailed map of the Nordic countries showing location and types of plants as a well as a layout of the transmission network, see http://www.nordel.org/Content/Default.asp?PageID⫽125& LanguageCode⫽EN

147

The Nordic electricity market: robust by design? 100

Denmark Finland Sweden Norway

90 80 70

GW

60 50 40 30 20

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

1978

1976

1974

1972

0

1970

10

Fig. 5.1. Installed capacity in the Nordic region. Source: Nordel.

physical connection between them.) In this way, players do not pay explicitly for capacity but bids in the market are used to define the need for various area prices. In 2003 about 25% of hours had a common price across all areas. Often there will be small differences between area prices, and only occasionally will the differences be large. Note also that as the demand side participates there is no demand forecasting. One of the important features of the Nordic market is the evolution of a robust power exchange. As shown in Figure 5.2, Nord Pool’s spot market saw over 40% of Nordic consumption traded in 2004. The remaining physical trade occurs through bilateral agreements or other arrangements. There is also a thriving derivatives market, where futures, forwards and options are traded actively by over 300 participants from both inside and outside the Nordic region. In 2004, nearly 1800 TWh were traded both on and off the exchange. Figure 5.2 shows trading volumes for Nord Pool’s physical market (Elspot, the day-ahead spot market and Elbas, an adjustment market), derivative volumes traded over the exchange’s Financial Market, as well as that which was traded “over the counter” (OTC) or bilaterally but cleared at Nord Pool. Volumes peaked in 2001 and 2002, falling somewhat in the following years due to a severe drought yielding high prices, and the withdrawal of Enron and other major international players from the market. Financial trading volumes have been between 4 and 8 times physical production in the Nordic region. In summary, the Nordic electricity market now constitutes a well-functioning market with the following key design features: ● ● ●

Competition in generation and retailing. Regulation of transmission and distribution. National TSOs responsible for system operation and for running real-time markets (differing from country to country).

148

Electricity Market Reform 3500 3000

TWh

2500

Cleared OTC volumes Financial market traded and cleared at Nord Pool Physical market: Day-ahead spot and adjustment markets

2000 1500 1000 500 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Fig. 5.2. Nord Pool market volumes. Source: Nord Pool ASA, Nord Pool Spot AS.

● ● ●

Point-of-connection transmission tariffs. Zonal pricing system of wholesale electricity. Free choice of supplier.

Table 5.2 lists the main events of the Nordic electricity market. In this chapter, a particular focus is put on how the Nordic market was able to withstand a serious shortage of capacity resulting from an unprecedented drought during the 2002–2003 period. This resilience stands in contrast to a number of other markets where less significant perturbations resulted in major mayhem.4 In the second half of 2002, inflow to hydro reservoirs was only 54% of the average of the preceding 20-year period (Bye et al., 2003b). As a result, reservoir fillings were at a record low at the beginning of the low-inflow/high-demand Winter season. Foreseeing tighter market conditions, producers began restricting supply in late Autumn and prices started to rise. The (daily average) spot price peaked at 850 NOK/MWh in January 2003, 2–3 times the normal level. High spot prices feed through to consumers, who in some cases faced increases in electricity bills of 50% or more.5 There was speculation that high prices were the result of abuse of market power, as well as a lack of investment in both generation and transmission in earlier years, and that rationing on a massive scale would be required. As it turned out, no such drastic measures were warranted, as responses from consumers and thermal-power 4

For further discussion of the features of the Nordic market, consult Bergman et al. (1999) as well as Nord Pool. 5 Note that, since many Nordic consumers rely on electricity for most domestic energy needs, including heating, electricity bills tend to make up a considerable share of household budgets. For a typical Norwegian household, annual electricity consumption is around 20 MWh (compared with an average of 3.6 MWh in Britain), while the annual bill would amount to around NOK 14,000 (approximately euro 1700) at a price of 250 NOK/MWh.

The Nordic electricity market: robust by design?

149

Table 5.2. Major milestones in the evolution of the Nordic market. Event

Year

Comments

Energy Act, Norway

1990

Power Exchange, Norway

1993

The Nordic Power Exchange

1994

Regulatory reform, Sweden

1996

Norwegian–Swedish Power Exchange

1996

Regulatory reform, Finland

1997

Further market integration

1998


1999

Regulatory reform, Denmark

1999


2000

Reorganization of Nord Pool

2002

Serious hydro shortage

2002–2003

TGCs

2003

Integration of TSOs

2005

ETS

2005

Nuclear power moratorium in Sweden

2005

Provides legal framework for restructuring and regulation Statnett Market AS established as an independent company The Nordic Power Exchange financial market grows through product development. The first Market Council of the Nordic Power Exchange established Provides legal framework for restructuring and regulation The world’s first multinational exchange for trade in power contracts. The power exchange is renamed Nord Pool ASA Provides legal framework for restructuring and regulation Finland joins the Nordic power exchange market area. EL-EX power exchange in Helsinki to represent Nord Pool Elbas established as a separate balance adjustment market in Finland and Sweden. Western Denmark joins Nord Pool Provides legal framework for restructuring and regulation Eastern Denmark joins Nord Pool as a separate price area Spot-market activities organized in Nord Pool Spot AS owned by all of the Nordic TSOs and by Nord Pool ASA Sharply rising electricity prices with significant demand side response. No intervention from authorities. As the first Nordic country Sweden introduces a TGCs market Energinet.dk in Denmark established by merging from Eltra in West Denmark and Elkraft in East Denmark EU Emissions Trading starts January 1. Norway joins A referendum in Sweden in 1980 resulted in a decision to close down all 11 nuclear reactors in Sweden by 2010. Barsebäck’s second reactor closed down in Summer.

producers balanced the market. Even though prices remained high during most of 2003, market conditions gradually normalized. Some saw the events of 2002–2003 as a warning sign, or indeed as outright proof that the electricity market is flawed. Others consider its performance through this period as evidence that the market has reached maturity and is robust enough to withstand even quite extreme shocks. We tend to lean toward the latter view. Nevertheless, the supply

150


shock brought to the surface a number of potential weaknesses that warrant careful analysis and which may eventually lead to further improvements in the regulatory framework as well as in other market institutions. After describing the events of 2002–2003 in some detail, we analyze two issues that have attracted considerable attention in the aftermath of these events. The first issue concerns the operation of the retail market; in particular, whether there is sufficient competition on the market, and whether current contractual arrangements are adequate for consumer needs. The second issue concerns generation and transmission capacity; in particular, whether sufficient investment is forthcoming and reasonable levels of supply security and adequacy may be maintained. We conclude this chapter with a discussion of how the implementation of the Kyoto protocol may affect the Nordic electricity market. In particular, we address the effects of the new Emissions Trading Scheme (ETS) of the European Union (EU) and the effects of the emerging Tradable Green Certificates (TGC) system of the Nordic countries. These environmental markets are now making a significant impact on the Nordic electricity market. Hence, in order to understand the functioning of the Nordic electricity market it is essential to recognize the importance of these new environmental markets.

5.2. Coping with a Supply Shock6 The development of the electricity market during the Winter of 2002–2003 was spectacular, with prices reaching unprecedented levels and a constant threat, according to some observers, of rationing on a massive scale. We concentrate our attention on events in the Norwegian segment of the market, where effects were at their most extreme. Figure 5.3 shows the Nord Pool (system) spot price (daily average) and the level of Norwegian hydro stocks7 from the beginning of 2002 to the end of the Summer of 2003. (The average exchange rate for 2002 was 7.5 NOK/euro.) The figure also shows median hydro stocks over the period 1990–2000. From a low level during the Summer of 2002, the spot price rose gradually during the early Autumn. This is normal and reflects the fact that limitedstorage capacity makes it impossible to transfer sufficient water into the high-demand/ low-inflow Winter season to equate prices over the year. However, toward the end of the Autumn the spot price rose steeply and continued to rise well into the Winter, when it peaked at around 850 NOK/kWh. The spot price then fell during the late Winter and Spring, but remained relatively high during most of 2003. The price development reflects the development of hydro stocks. Stocks fell from high levels in the Summer of 2002 to record low levels in the following Winter. The most obvious reason for this unusual development was the extremely dry hydrological conditions with an almost total stop in inflows to reservoir during the normally wet weeks of the late Autumn.8 As shown in Figure 5.4, which compares weekly inflow to Norwegian hydro reservoirs during 2002 with yearly averages over the period 1970–1999, the year 2002 actually started out as rather wet. Until the Summer, inflow was consistently above the historical average; 6

This section draws extensively on Bye et al. (2003a); see also Bye (2003), Bye and Bergh (2003) and Bye et al. (2003b). 7 The Nordic hydro generation capacity is almost entirely located in Norway and Sweden. The Swedish hydro stocks followed a parallel development to the Norwegian. 8 Inflow to reservoirs is at its minimum in the Winter season when most precipitation is in the form of snow, and it reaches its maximum in the late Spring and early Summer when the snow melts. Autumn is usually rather wet, with almost all precipitation in the form of rain, and hence inflow is relatively high in this period also.

151


100 90 80 70 60 50 40 30 20 10 0

800

Percent

600 500 400 300

NOK/MWh

700

200

Spot price

Hydro stocks

24-09-03

13-08-03

02-07-03

21-05-03

09-04-03

26-02-03

15-01-03

04-12-02

23-10-02

11-09-02

31-07-02

19-06-02

08-05-02

27-03-02

13-02-02

02-01-02

100 0

Median stocks

Fig. 5.3. Spot price and Norwegian hydro stocks, actual 2002–2003 and median 1990–2000. Source: Statistics Norway and Norwegian Water and Energy Authorities. 9000

Year 2002 Yearly average, 1970–1999

8000 7000

GWL

6000 5000 4000 3000 2000 1000 0 1

5

9

13

17

21

25

29

33

37

41

45

49

Fig. 5.4. Weekly inflow (GWh) to Norwegian hydro reservoirs. Source: Norwegian Water and Energy Authority.

indeed, in the 24 first weeks of 2002 inflow was 14 TWh, or nearly 20%, above average. However, in early Autumn inflow fell below normal levels and from October onwards it more or less dried up completely; during weeks 38–48 inflow was 9.3 TWh below average. It has been argued that the fall in hydro stocks could have been avoided if generators had restricted supply at an earlier stage. However, with the very high levels of stocks in the early Autumn there was apparently a real risk that – with a wet Autumn – reservoirs would have become so full that water would have been spilled and wasted. Figure 5.5 shows actual total hydro stocks, as well as projected levels as seen from week 30. With maximum inflow reservoirs would have become completely full. Given differing levels of stocks, inflow patterns and constraints on output, the risk that any individual

152


Percent

100 90 80 70 60 50 40 30 20 10 0

Actual Projected minimum Projected maximum Projected mean 1

5

9

13 17 21 25 29 33 37 41 45 49 Week

Fig. 5.5. Norwegian hydro stocks 2002, actual and projected from week 30. Source: Statistics Norway and Norwegian Water and Energy Authorities.

reservoir would be filled up was likely to be higher than suggested by these figures. As it turned out, such a wet outcome did not occur; instead, inflow and hydro stocks were in fact close to the lowest projected level. This evidence would seem to be consistent with a view that generators acted rationally upon information available at the time. On the other hand, the evidence is also consistent with the view that hydro generators were overly anxious to tap their reservoirs, possibly with the intention of pushing up prices, which, at the time, looked to remain at modest levels during the Winter. In any case, the mere suspicion that such anti-competitive practices were pursued may well have been influential in determining the tough stance that the Norwegian Government took on attempts by the dominant (and state-owned) generator Statkraft to buy up more of its smaller Norwegian rivals. Another possible explanation for the price spike, popular among a number of commentators, was that a lack of investment over a long period of time had eventually lead to undercapacity and a consequent imbalance between supply and demand. It is certainly true that investment levels had been low for a number of years, but whether this was a sign of market imperfection is questionable. We discuss the investment issue in some detail below. Here we only point to the fact that even though prices reached record levels during 2002–2003, they had been well below levels that would make new investment profitable for most of the preceding 10–15 years. Forward contract prices were also low. For instance, in 1999 prices in forward contracts covering the year 2002 were around 150 NOK/MWh, well below the 200 NOK/MWh estimate at the time of unit costs for new gas-fired plants. Forward prices gradually increased in subsequent years, although in 2002 prices covering the year 2006 were still not higher than 180 NOK/MWh. There consequently seems to be little support for the view that generators had not seized on profitable investment opportunities. A criticism along similar lines concerned investment in transmission capacity.9 Some commentators argued that lack of investment in transmission capacity, both within and between 9

A completely opposite, but nevertheless quite popular view was that high prices in Norway were caused by excessive interconnection with neighboring countries, leading to a drain on hydro resources and consequent shortage prices. While there was certainly some truth in this view, it seemed to overlook the fact that in the face of a severe, negative supply shock prices would reach even higher levels without access to imports. We do not discuss the possible (protectionist) implications of this view.

153

The Nordic electricity market: robust by design? 700 Households Service industries Manufacturing industries

600 NOK/MWh

500 400 300 200 100

2004:1

2003:4

2003:3

2003:2

2003:1

2002:4

2002:3

2002:2

2002:1

2001:4

2001:3

2001:2

2001:1

2000:4

2000:3

2000:2

2000:1

1999:4

1999:3

1999:2

1999:1

0

Fig. 5.6. End-user prices (excluding taxes and network tariffs), quarterly observations, 1999–2004. Source: Statistics Norway.

the Nordic countries, had allowed the development of areas where severe shortages were bound to arise. Again, it is true that there had been relatively little investment in transmission capacity over a number of years, and that the 2002 supply shock leads to long periods of time in which the market was effectively segmented. In particular, import capacity to Norway, and export capacities from the thermal-dominated systems in Denmark and Finland were constrained during much of the Winter of 2002–2003. Nevertheless, it is not clear that this was a sign of deficiencies in the transmission network. On the one hand, the Nordic countries are highly interconnected, and most of the time the market is fully integrated at a single market-wide price. On the other hand, bottlenecks are becoming more frequent, and, in combination with increasing levels of market concentration, there is a worry that the result may be imperfect competition and inefficient market outcomes. We return to this latter issue below. We also return to the issue of transmission investment in connection with our discussion of supply security and adequacy. The high spot prices feed through to end-user prices. Figure 5.6 shows quarterly observations of the energy element in average retail prices from early 1999 to early 2004 for households, services industries and manufacturing industries, respectively. Retail prices shot up at the end of 2002 and soon reached unprecedented levels. Prices paid by households tended to increase more than those paid by industrial consumers. The difference seems to be explained by the different composition of contracts in the various segments of the market. Most household consumers have the so-called “variable-price contracts”, according to which retailers can change the price with a few weeks notice. As of the first quarter of 2003, 85% of household consumers had such contracts; another 7% were on spot-price contracts (with the retail price directly linked to the Nord Pool spot price) and only 8% on fixed-price contracts (see Fig. 5.7). This is different from industrial consumers, especially in the manufacturing industries, who tend to rely more on long-term, fixed-price contracts. In the first quarter of 2003, 55% of consumers in the manufacturing industries and 22% of consumers in service industries had fixed-price contracts. The corresponding figures for spot-price contracts were 35% and 53%, and for variable-price contracts 10% and 24%. Consequently, industrial consumers were less exposed to price increases than were households. There was a general move from variable-price contracts to fixed-price contracts in the wake of the events of the Winter of 2003. This trend seems now to be reversed (maybe because the

154 100 90 80 70 60 50 40 30 20 10 0

2002:1 2002:2 2002:3 2002:4 2003:1 2003:2 2003:3 2003:4 2004:1 2002:1 2002:2 2002:3 2002:4 2003:1 2003:2 2003:3 2003:4 2004:1 2002:1 2002:2 2002:3 2002:4 2003:1 2003:2 2003:3 2003:4 2004:1

Percent


Households

Service industries

Variable price

Spot price

Manufacturing industries Fixed price

Fig. 5.7. Contract shares, quarterly observations. Source: Statistics Norway.

memory of the price spike is starting to fade?). Interestingly, there is little interest among household consumers, as opposed to industrial consumers, for the so-called “spot-price” contracts, in which the retail price is linked to (an average of) the Nord Pool Elspot price. There is ample evidence that spot-price contracts perform consistently better than variable-price contracts in the longer term; in particular, it would seem that competition between suppliers of variable-price contracts is not always entirely effective. The reason households have not embraced spot-price contracts may be (an erroneous) view that these contracts, being linked to the highly variable spot price, are somehow more risky than variable-price contracts. Increases in end-user prices had a considerable impact on demand. Roughly speaking, demand may be seen as consisting of three segments: the very flexible boiler segment (approximately 5% of the total), the heavily contracted power-intensive industry (approximately 30%) and the rest (approximately 65%). Demand from the boiler segment, which can easily switch between oil and electricity, fell sharply when prices started to rise in October 2002 and remained low during the Winter; all in all, electricity consumption by boilers over the period November 2002 to May 2003 was around one-third of that of the corresponding period in 2001–2002. In the energy-intensive industries, some plants stopped production, but the overall response was relatively small, probably less than 5% (Bye et al., 2003a).10 In the remaining segment, households and other industry, temperature-adjusted demand fell by 7% over the November–May period compared to the year before; given an average increase in end-user prices of 30%, this corresponds to a price elasticity of 0.23. The experience in Norway may be contrasted with that of the other Nordic countries. Although wholesale prices moved more or less in parallel, retail prices were much less affected in these countries. This would seem to be explained by the fact that retail markets differ, particularly in the availability and composition of contracts, but also in market structure and the extent of competition. In Denmark and Finland, where fixed-price contracts 10

The response seems small in comparison to estimated reservation prices for the power-intensive industry. The industry has the option of moth-balling plant and selling power, obtained of preferential terms determined by the Norwegian Parliament, in the spot market (Bye and Larsson, 2003). The lack of response may be due to uncertainty about price developments, long stop and start-up times and the risk of undermining popular and political support for the industry.


155

dominate, domestic consumers were much less exposed to price increases than in Norway. In Sweden, there is a greater variety of contract types, although the incidence of long-term, fixed-price contracts is higher than in Norway. Moreover, as we discuss below, there seems to be less competition among Swedish than among Norwegian retailers. Also, in Sweden retail prices reacted much less than in Norway. As a result, the demand response was much less in these countries than in Norway. 5.3. Market Integration and Retail Competition The events of 2002–2003 cast light on a number of potential problems, including concentration and scope for market power, contractual coverage and exposure of consumers to price risks, investment and its impact on supply security, as well as bottlenecks in the transmission network and segmentation of the market. Below we focus on the retail market (this section) and issues concerning supply security and adequacy (next section). We also discuss how the introduction of new, market-based instruments for regulation of environmental emissions may impact on the performance of the electricity industry (penultimate section). The establishment of Nord Pool and the elimination of border tariffs between the Nordic countries were key elements in a strategy aiming at an integrated Nordic market for electricity. The success of this strategy may be measured by the degree of wholesale and retail price equalization between the different “price areas”.11 Obviously, an uneconomically large transmission capacity would be required if transmission constraints were to be eliminated, enabling wholesale prices to be equalized across all areas at all times. However, significant and persistent deviations between area prices would imply that the Nordic market consists, in effect, of a set of national or regional electricity markets. As we show immediately below, the wholesale market appears to be strongly integrated, with prices in different areas diverging for shorter periods only. However, as mentioned above, retail market prices reacted very differently across the Nordic countries to the 2002–2003 increase in wholesale prices: while they shot up in Norway, the reaction was much more subdued in Sweden, and in Denmark and Finland retail prices hardly changed at all. There are also considerable differences in the level of retail prices, even when one corrects for differences in taxes and network tariffs. Some of these differences can be explained by differences in regulatory regimes. We concentrate our attention on Norway and Sweden, where regulations are similar, but where retail markets nevertheless seem to perform quite differently.

5.4. Wholesale Prices Table 5.3 displays the annual averages of the Elspot system12 and area prices over the period 1996–2003. The figures indicate that deviations between system and area prices have been

11

Whenever interconnector capacity constrains power flows, the Nord Pool market is divided into two or more “price areas”. Sweden is always treated as a single price area, and the same applies to Finland. This is because congestion in the national transmission systems is managed by means of the so-called counter-trade in these countries. In Denmark, the eastern and western parts of the country are physically separated and hence there are always two price areas – East and West. In Norway, segmentation of the market is part of the handling of transmission constraints and the country may be divided into two to five price areas, depending on the demand–supply configuration. 12 The “system price” is calculated under the assumption that there are no transmission constraints. Actual trade is carried out at the system price only when transmission constraints are not binding.

156


Table 5.3. Elspot system and area prices 1996–2003, annual averages, NOK/MWh.

System Norway, Oslo Norway, Tromsø Sweden Finland Denmark, West Denmark, East

1996

1997

1998

1999

2000

2001

2002

2003

253.6 256.7 251.2 250.6 – – –

135.0 137.5 133.0 135.0 – – –

116.4 115.7 116.2 114.3 116.3 – –

112.1 109.2 119.5 113.1 113.7 113.7 –

103.4 97.7 100.7 115.5 120.7 120.6 –

186.5 186.0 188.6 184.2 184.0 191.2 189.7

201.0 198.5 200.2 206.3 203.8 190.7 213.7

297.5 301.7 295.7 292.8 277.9 268.3 291.7

Source: Nord Pool. Table 5.4. Number of hours with complete Elspot price equalization 1997–2002.13

Number of hours Share of time (%)

1997

1998

1999

2000

2001

2002

5201 59.4

3825 43.7

3788 43.2

1703 19.4

4487 51.2

3076 35.1

Source: www.seef.nu.

quite small. Except for the years 2000 and 2003 – when the supply of hydropower, especially in Norway, significantly deviated from normal levels – the Nordic electricity market appears to be reasonably close to being a “single market”. However, small discrepancies between annual averages may hide short-term variations in different directions and is thus only a very crude indicator of the degree of price equalization. As can be seen from Table 5.4, the number of hours in which the entire market has been integrated (i.e. when the system price has been exactly equal to all area prices) is below 60%. The corresponding figures for Finland–Sweden are in the range 75–100%, for Norway (Oslo)–Sweden 70–85% (except in 2000) and for Finland–Norway (Oslo)–Sweden 60–85% (except in 2000). Moreover, the share of the time when Sweden has been a “price island” is in the range 0–5%. Thus, in terms of wholesale price equalization the Nordic electricity market would seem to be reasonably well integrated. 5.5. Retail Competition There is complete market opening (i.e. full retail competition) in all the Nordic countries. In some of the countries, such as Sweden, a household consumer may even buy electricity from suppliers in any Nordic country. Given this, the pre-tax retail prices should not differ very much between the four Nordic countries. However, there are obstacles to transactions between suppliers in one country and households and other small customers in other countries. For instance, in order to be able to supply electricity to a Swedish customer located in Stockholm a non-Swedish supplier needs to buy electricity in the Stockholm price area. Moreover, as a buyer in the Stockholm price area the supplier needs to have a contract with a so-called “balance responsible party” (as well as in the home country).

13

Presentation by Dr. Niklas Strand, Swedish Competition Authority, at the Swedish Association for Energy Economics.


157

As a result of such obstacles, only a sub-set of all retailers in the Nordic countries is actually competing on the Swedish market. In addition, only a sub-set of all Swedish retailers competes outside the geographical area in which they are located. Corresponding situations prevail in the other Nordic countries. Consequently, retail electricity prices need not necessarily be equalized across national borders. In order to shed some light on this issue, we have compared retail prices in Norway and Sweden, the two countries that have operated a common wholesale market since 1996. Retail prices differ between households for several reasons. One reason is related to nonlinearity of price schedules and annual consumption patterns of households. Thus prices paid by households living in a single-family houses with electric heating, typically consuming around 20 MWh/year, are lower than the prices paid by households using electricity only for lighting and electrical appliances, consuming as little as 2 MWh/year.14 Another reason for retail price differences is that customers can choose between fixed-price contracts (with different duration) and variable-price contracts, and that the prices charged for these contracts may deviate in the course of year. In both Norway and Sweden, the “default contract” – the type of contract that applies for customers who have not actively chosen to change supplier or signed a new contract with the “old” supplier – is a variable-price contract; that is, a contract that allows the supplier to adjust the price (after notifying the customer) and so, in effect, pass on cost increases to customers. A third reason for retail price differences is that individual retailers may adopt different market strategies and offer different combinations of prices and services. In Tables 5.5 and 5.6, annual averages household prices over the period 1997–2003 are displayed for, respectively, Norway and Sweden. The numbers reflect averages of prices offered by all retailers to households with an annual consumption equal to 20 MWh. It is immediately clear that Norwegian and Swedish retail prices differ significantly. No “law of one price” is visible, and the retail prices have been significantly higher in Sweden during most of this period. Thus, with regard to the household market, there seems to be two national rather than one integrated Norwegian–Swedish electricity market. Annual variations of the Norwegian retail prices are quite significant, but also relatively well correlated with variations in Elspot prices (compare Tables 5.3 and 5.5). Thus retail prices fell during the “wet” period 1997–2000, increased in 2001 when precipitation was “normal”, and skyrocketed in 2003 when Nord Pool prices were extremely high. In the case of Sweden, however, the correlation between Elspot and retail prices was rather weak during the period 1996–1999. Instead, average retail prices remained high in 1998 and fell only marginally in 1999, while a quite significant reduction took place in 2000 and 2001. The most obvious explanation for these differences between the development of retail prices in Norway and Sweden are related to switching15 costs at the household level. In Norway, a system of profiling was adopted from the outset and customers could switch to another supplier at no cost; more specifically, there were no requirement to install interval or real-time meters, nor any charges levied. The option of changing supplier at no charge has led to a sizeable swing of customers away from the “old” local suppliers to

14

To some extent this pattern is surprising. On the one hand, it is likely that economies of scale may motivate some quantity discounts. On the other hand, it is obvious that households with electric heating consume electricity primarily during the Winter period when peak generation capacity is used and spot prices are high. As spot prices are sometimes very high during the Winter season, one would expect that the cost of offering the customer insurance against Winter “price spikes”, which is what the retailer does, would be rather high.

158


Table 5.5. Retail prices, net of taxes, for 20-MWh household consumer according to contract type in Norway 1997–2003, NOK/MWh.

Normal Spot 1 year fixed Average

1997

1998

1999

2000

2001

2002

2003

n.a. n.a. n.a. 210

160 n.a. n.a. 162

151 131 152 152

141 123 144 141

210 193 189 206

205 193 195 203

454 323 287 414

Source: Statistics Norway.

Table 5.6. Retail prices, net of taxes, for 20 MWh household consumer according to contracts in Sweden 1997–2003, NOK/MWh.

Normal 1 year fixed 2 year fixed 3 year fixed

1997

1998

1999

2000

2001

2002

2003

259

251

244

218 178 177 182

225 181 184 186

296 256 253 252

447 397 351 324

Source: Statistics Sweden.

“new” suppliers that offer electricity at lower prices. In other words, there were no significant switching costs protecting the “old” suppliers from competition. In Sweden, on the other hand, costly real-time metering and reporting were required for consumers wanting to change supplier. These regulations were in effect until November 1999, and as a result few households changed supplier or renegotiated their contract with the “old” supplier. After November 1999, a system of profiling has been in place. Consequently, there was a significant reduction of switching costs in Sweden at the end of 1999. The figures in Table 5.6 suggest that retail competition was rather modest as long as switching costs were high. At the same time, the figures indicate that the reduction of switching costs opened up a significant competitive pressure on retail prices. Nevertheless, even with a regulatory regime more conducive to competition, retail prices continue to be higher in Sweden than in Norway, in spite of the fact that retailers procure electricity at the same well-integrated wholesale market. A natural hypothesis is that this is a result of market power.

5.6. Market Structure Traditionally, the major generating companies in Sweden have had rather small shares of the retail market. Thus, although Vattenfall for a very long time has been the single biggest

15

It should be noted that the very high per capita electricity consumption for heating, lighting and household utensils of Norwegian and Swedish households make it profitable to switch electricity supplier when prices differ. This may, however, not be the case in other European countries (e.g. the UK) where electricity is mainly used for lighting and household utensils.


159

retailing company, its share of the retail market was only around 15% in the middle of the 1990s. However, in the last few years the major generating companies – Vattenfall, Sydkraft and Fortum (formerly Birka) – have bought majority or minority shares in a number of small- and medium-sized retailing companies. In most cases, sellers have been towns and municipalities. Moreover, some of the independent retailing companies that entered the market in 1999, such as the Norwegian oil and gas company Statoil, have since left the market. As a result of these developments the number of retailing companies has been reduced, and the “big three” have become dominant players on the retail market. For instance, if we include retailing companies in which generators own minority shares, Vattenfall is currently serving around 30% of all Swedish customers, while the corresponding number for the “big three” is around 70%. Similar numbers apply to shares of volumes of electricity delivered to final consumers. In Norway, power generation has traditionally been much less concentrated than in Sweden. Except for the state-owned company Statkraft, accounting for some 30% of total Norwegian power generation, generators are small, with market shares of 5–6% or lower. As Statkraft and the second largest company, Norsk Hydro, almost exclusively serve industry and other businesses on long-term contracts, the retail and household market in Norway has been much less concentrated than in Sweden. In recent years, however, significant changes in the Norwegian power sector have taken place. In particular, many companies in local-government ownership have been turned into limited-liability companies, often as part of a process leading up to the sale of ownership interests. Also, larger regional power companies have been established, partly by acquisition and partly through mergers. Furthermore, Statkraft has acquired stakes in several Norwegian power companies. Foreign companies have also acquired some ownership interests in Norwegian companies (notably in grid management and operations and in power retailing). In spite of these developments, however, the Norwegian market, both at the wholesale and retail level, remains less concentrated than its Swedish counterpart. 5.7. Market Power and Price Discrimination So far, the retailing segment of the Swedish electricity supply industry has not been very profitable, and several entrants to the market have had to leave after having suffered significant losses. On the whole, it seems that the costs of the retailing business have been severely underestimated; in particular, it seems that the exposure to price and quantity risks have been more costly than expected.16 But in 2002 the established retailers were able to implement an increase in trade margins without attracting new entrants to the market. The question then is why “the big three” have grown while several independent retailers have left the market. A possible explanation may be that integrated generation-retailing companies are more efficient than independent retailers. In Sweden, legal separation between retailing and distribution is required. This provision has paved the way for significant integration 16

Retailers can hedge Elspot system price risks at Eltermin, and the liquidity of these instruments (futures, forwards and options) traded at Eltermin is high. However, retailers in Sweden have to buy electricity at the relevant Swedish area price, and opportunities for hedging idiosyncratic area price risks are not very well developed. Moreover, retailers can only hedge price risks for a fixed number of MW per hour, while their customers do not have to commit to a certain quantity. Thus retailers are exposed to the risk of having to buy “extra” electricity at spot-market price during hours when their customers have an unexpectedly high level of consumption.

160


between generation and retailing. It seems that the integrated companies have a competitive advantage in relation to independent retailers stemming from the lack of efficient markets for hedging against area price and quantity risks. Thus, while all retailers may suffer from a combination of unexpectedly high area prices and consumption levels, the extra costs in the retailing business become extra revenues in the generation business for the integrated generation-retailing companies. If these hypotheses could be verified they clearly point at an unexpected effect of the legal separation requirement in Sweden. The intention was to stimulate retail competition by preventing cross-subsidization between distribution and retailing. This objective may have been attained, but recent developments suggest that allowing vertical integration between generation and retailing also made it possible to exercise market power in the retail market. There also seems to be an element of price discrimination in Swedish retail prices. As seen in Table 5.6, there is a considerable spread between prices in the so-called “normal contract” and prices in fixed-prices contracts or between prices paid by customers with default contracts and customers who have actively switched to a new contract; in particular, prices in default contracts are higher than those in fixed-price contracts. This suggests that Swedish retailers are able to price-discriminate against the default-contract customers. After all, by refraining from actively choosing another type of contract these customers have demonstrated that they are not very price-sensitive and it must be tempting for suppliers to utilize this information. 5.8. Supply Security and Adequacy A notable consequence of regulatory reform in the Nordic countries has been the almost complete halt to investment in generation and, to a lesser extent, in transmission and distribution. At the outset this should be considered a positive feature as this has implied a reduction of an inefficient over-capacity in generation inherited from the past. Figure 5.8. shows installed generation capacity in the Nordic market since 1980. Over the 10-year period preceding the first regulatory initiative, from 1980 to 1990 (the year before the new Energy Act took effect in Norway), generation capacity grew by 30%. Over the next 10-year period, from 1990 to 2000, installed capacity grew by a mere 3%. Indeed, in 2003 installed capacity was more or less the same as in 1996, the year when regulatory reform was introduced in Sweden: a fall during 1998–1999 was only reversed by subsequent increases in recent years.17 The stagnation in capacity growth cannot be explained by development of demand. Admittedly, demand did not grow at the pace experienced in the early 1980s and before, but gross consumption continued to grow at a more or less constant rate of 1–1.5% per year. As a consequence, over-capacity inherited from the pre-reform era has gradually being reduced, if not entirely eliminated. Comparing generation capacity and maximum system load, we find a similar picture, although the trend is perhaps not as pronounced. Nevertheless, the capacity margin, defined as the excess of installed generation capacity over maximum load, reached it lowest level in 2001. The development of generation capacity must be seen in relation to stricter regulatory policies, arising mostly from environmental concerns. Although a considerable amount of undeveloped hydro capacity still remains, it is unlikely that many more new hydro sites will be developed. Nuclear power, traditionally important in both Finland and Sweden, has long been viewed with great skepticism (although the Finnish Parliament recently 17

This movement seems to be explained largely by plants that were first mothballed and then re-opened.

161


100 90

400

70

350

60

300

50

Installed capacity Gross consumption Maximum system load

40

GWh

GW

80

250 200

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

1982

1980

30

Fig. 5.8. Capacity, consumption and system load. Source: Nordel.

approved the building of a new nuclear power plant). Increasing concerns about air pollution has lead to strict regulations, not only on coal- and oil-fired power plants, but also gasfired plants. Notwithstanding the importance of environmental regulation, it would seem that regulatory reform – with the abolition of monopoly rights, integration of markets and development of market places – has been the most important factor in determining generator investment behavior. Market-based competition not only reduced the prices, but also turned the focus of market participants toward profitability. Indeed, in an industry traditionally committed to a public service ethos, regulatory reform legitimized a more “capitalist attitude”. The greater emphasis on profits lead to company restructuring and mergers, as well as to increased efficiency (e.g. employment in the Norwegian electricity industry fell from almost 20,000 in 1993 to below 13,000 in 2002). Specifically, electricity supply companies increasingly required returns on investment in line with those obtained in other industries. Figure 5.9 shows return on capital in the Norwegian electricity industry since 1990, the year before deregulation.18 Over this period, rate of return has averaged 5.5%, only half that achieved in the Norwegian manufacturing industry. Due to the existence of a “resource rent” in a hydro-dominated electricity industry, we would expect the average rate of return to exceed that on a marginal plant. Consequently, it seems relatively safe to conclude that new investment in generation capacity could not have achieved reasonable levels of profitability over this period. Nevertheless, even though low levels of investment seem to have been a rational response to prevailing market conditions, the question remains whether investment will be forthcoming as the earlier over-capacity is eroded and market conditions become tighter. In other words are there reasons to believe that market imperfections will inhibit investment and undermine the future performance of the industry? To answer this question, it is important to distinguish between two related, but nevertheless entirely different, concepts: supply security or balancing consumption and generation

18

Rate of return is measured according to National Accounts as operating surplus relative to the value of capital employed.

162

Electricity Market Reform 18 Electricity industry

16

Manufacturing industries

14 Percent

12 10 8 6 4 2 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Fig. 5.9. Return on capital, 1990–2002. Source: Statistics Norway.

on a continuous basis within existing capacity limits; and supply adequacy or ensuring optimal capacity investment by balancing willingness to pay for new capacity against its cost. In other words, while the supply-security issue is short term and mainly concerns system operation, the supply-adequacy issue is long term and concerns the evolution of capacity in relation to consumption. In interpreting the balance of consumption and generation one must be aware of transmission constraints that limit the range of generators that can meet demand in a given location. Below, we discuss these issues in some detail.

5.9. Supply Security: Balancing Demand and Supply As is well known, the problem of continuously balancing consumption and available generation arises from specific features of electricity markets, including the need for electrical equilibrium at all times, unexpected variations in demand and supply, limited possibilities for establishing and transmitting adequate price signals to market participants on a continuous basis, and limited short-run response by market participants to price signals. The gain from increasing supply security is associated with a reduction in the costs of rationing. A rationing event occurs when, at prevailing prices, the desired demand and/or supply of market participants cannot be satisfied and hence their decisions have to be constrained. For example, if demand exceeds supply at prevailing prices, either additional supply (if available) has to be ordered onto the system, or the consumption of one or more consumers has to be forced down. Rationing may be voluntary or involuntary. One example of a voluntary rationing agreement is contracts for operational reserves, by which the system operator obtains the right to call on additional supplies when needed. Another example is long-term load-shedding contracts between consumers and their suppliers (or between consumers and the system operator), by which consumers, upon conditions that have been agreed in the contracts, can be called on to reduce their load. Involuntary rationing typically occurs by zonal interruptions of power supplies. Supply security cannot be ensured by capacity investment alone, although a larger capacity may reduce supply-security problems. Optimal utilization of existing generation capacity, whatever its level, involves setting prices that allow for the highest possible degree of


163

capacity utilization while at the same time securing sufficient reserve margins. Having more capacity available essentially means that prices will remain at lower levels so that more demand is encouraged and a high level of capacity utilization is ensured. Conversely, when capacity is limited prices will increase so as to reduce demand. Provided generation capacity is optimally used, on longer time scales demand for capacity will follow available capacity and system reserves will not be directly linked to total capacity. Obviously, in the Nordic system, with a large incidence of hydropower, there will be sustained periods of time in which the energy balance is tight. In such situations, a continued balance between demand and supply requires that prices rise, as happened during the Winter of 2002–2003. Note that the price rise would have been less in that event if more of demand had been exposed to the actual cost of electricity: the reliance on fixed-price contracts in large segments of the market exacerbated the supply deficit and pushed wholesale prices higher than they otherwise would have been. On the other hand, it is conceivable that one may in the future experience an even more severe reduction in inflows, with even higher prices as a result. Nevertheless, if, by adjusting price levels, demand can be scaled to be (on average) aligned with existing capacity, the supply-security problem is not associated with the level of demand (relative to existing capacity) per se but rather with variations in demand (and capacity availability). Given this, it must be noted that, in a hydro system, even if the energy balance is tight there is generally a large amount of (power) capacity available. Since hydro plants are typically constructed so as to be able to handle peak inflow levels, and adjustment of output is extremely cheap, the ability to deal with short-term variations in the system is not necessarily impaired in dry periods. To sum up, it is not at all clear that the Nordic market is particularly vulnerable to supply insecurity, even if market conditions were to become tighter in the future. Indeed, the technology mix, and the mere size of the market, would seem to allow for a very high degree of supply security. Moreover, regulations are in place, which should provide Nordic system operators with the tools they need to balance the system. Take the Norwegian TSO Statnett as an example. Firstly, rights and responsibilities have been clearly set out in specific regulations; in particular, Statnett is responsible for system balance and has the right to make the necessary contractual arrangements with market participants to achieve this task. Secondly, marketbased institutions have been set up to achieve balancing in a cost-effective manner. Most importantly, Statnett runs a (near-to) real-time balancing markets in which balancing services are sourced on a short-term basis. Furthermore, at regular intervals Statnett procure strategic reserves, partly in the form of contracts for interruptible demand. All in all, these instruments should be sufficient to guarantee that balancing is achieved at reasonable costs. Studies have indicated that the efficiency of system operations would be further enhanced by tighter co-operation between of system operators (Bjørndal and Jørnsten, 2001; see also von der Fehr et al., 2002). Developments in this direction have recently taken place, with integration of the national balancing markets (Nordel, 2003). However, these efforts are probably not sufficient, and further gains may be had from optimizing the system as a whole, including lower overall reserve margins and increased transmission capacity. 5.10. Supply Adequacy: Optimizing Capacity Investment The supply-adequacy problem essentially consists of two elements, namely ensuring an optimal level of overall generation capacity and an optimal mix of different generation technologies.

164


An optimal level of overall capacity is characterized by equality between willingness to pay for new capacity and the cost of such capacity. In other words, a situation of underinvestment in generation capacity would be characterized by investment not forthcoming even though the willingness to pay for the associated increase in output is more than sufficient to cover its cost. Similarly, a situation of over-investment would be characterized by the cost of marginal capacity units exceeding consumer willingness to pay for the associated output, a situation well known from the history of the Nordic electricity industry. An optimal mix of generation technologies is characterized by the minimization of costs of satisfying a given consumption profile. Cost-efficient operation requires a mix of technologies with different variable to fixed cost ratios. At one extreme, low-variable/high-fixed costs technologies, such as hydro, nuclear and conventional thermal, operate continuously as base-load units; at the other extreme, high-variable/low-fixed costs technologies, such as small gas- or oil-fired units, are used for demand peaks only. An optimal capacity mix balances the gains from reducing variable operating costs by having more base-load units available against the higher-fixed costs of such units. At the moment, there would seem to be two major concerns regarding supply adequacy in the Nordic market: firstly, whether investment incentives are sufficient to allow overall capacity to expand at a reasonable pace and secondly, how the generation park will be affected by the introduction of new environmental regulation initiatives. Here we focus on the first issue of general investment incentives and leave the latter issue for the next section. In the hydro-dominated Nordic market, wholesale prices swing from year to year, depending upon hydrological conditions. However, these fluctuations in prices tend to average out. Indeed, judging from prices in forward contracts, which are traded up to 3 years ahead on Nord Pool, expected prices tend to be quite stable, evolving slowly in reaction to changes in underlying fundamentals.19 As at late 2004, contracts for 2 and 3 years ahead are trading at around 250 NOK/MWh (31 euros/MWh), a level that would make investment in conventional gas-fired plant approximately break even. It would seem that these prices are in fact sufficient to attract investor interest. The Finnish Government and the EU Commission recently approved a new 1.6-GW nuclear power plant, expected to be in operation from 2009. In Norway, Statoil recently unveiled plans to build a 860-MW gas-fired plant and a 280-MW co-generation unit at its industrial plants at Tjeldbergodden and Mongstad (landing points for North-Sea gas). Gas-fired generation has been a source of controversy in Norway, leading to the downfall of at least one government, and it remains to be seen whether the Statoil projects will be approved. However, whatever the outcome, it would seem that generation investment in the Nordic market is not so much a question of commercial, as of political, will.20 Transmission investment has been fairly limited for a number of years. There seems to be two main issues: firstly, how regulation of the individual TSO affects investment incentives and secondly, the ability of national TSOs to solve co-ordination problems associated with investment in interconnector capacity. In the case of the Norwegian TSO, Statnett, there appears to be no lack of will to invest. Given that costs of new investment will in effect be 19

Hence, due to the well-functioning market for forward contracts and the fact that the Nordic countries are integrated within and beyond, seasonal and annual swings of precipitation in a hydrodominated country like Norway (close to 100%) do not result in similar swings of prices and threats of rationing as in countries like Brazil and Colombia (see other chapters in this book). 20 Uncertainty surrounding the future of the Swedish nuclear capacity is another important example of how the political process interacts with investment incentives. Investment in renewables, such as wind power and generation based on burning of biomass, is critically dependent upon government support.


165

passed on to network users, Statnett has shown considerable willingness to undertake new projects. Were it not for the more restrictive view taken by Norwegian regulatory authorities, more transmission capacity would indeed have been built.21 The regulatory regimes differ somewhat between the Nordic countries, which may explain why other TSOs have been more reluctant to invest, particularly in new interconnector capacity. However, Nordel, the association of Nordic TSOs, has taken several initiatives both to resolve problems associated with transit (Nordel, 2001), as well as producing plans for co-ordinated expansion of the Nordic transmission network (cf. the “Nordic Grid Master Plan” described in Nordel, 2003). These initiatives would seem to go a long way in resolving transmission problems, but, again, it would seem that regulatory and political will, rather than commercial will, is going to be decisive. 5.11. Emissions Trading In the following we address the effects of the new ETS of the EU and the effects of the emerging TGC system of the Nordic countries. These environmental markets are now making a significant impact on the Nordic electricity market. Among the Nordic countries, Denmark, Finland and Sweden will, as EU Member States, have to implement the ETS,22 while Norway, not being a member state, is not obliged to do so. However, Norway has recently redesigned its former national (more ambitious) ETS proposal to better fit the EU Directive, and Norway seeks in this way to co-operate and participate in the EU arrangement on an equal footing with the other Nordic countries. A possible outcome of this process is that Norway adopts the EU Directive and thus commits to the rules and regulations of the ETS. Table 5.7 gives numbers for CO2 emission in electricity and heat supply. The levels of CO2 emissions are very different. Denmark, with its relatively small electricity industry, primarily based on thermal power, has the highest level of emissions, whereas Norway, with a relatively large electricity industry, almost exclusively based on hydropower, has very little emissions. The simple idea of the ETS is that the permit price will function as a cost increment of using a CO2 emitting resource, with the increment being in proportion to how much that resource emits per unit used. As a result, input substitution is expected to take place in electricity generation, away from coal and gas power toward hydro, wind and nuclear power. Hence it is to be expected that, in the Nordic countries, the ETS will result in an increase in Table 5.7. CO2 emissions in public electricity and heat supply in 1998 (million tons). Denmark 27.8

Finland

Norway

Sweden

18.2

0.2

7.9

Source: IEA (2000). 21

Statnett recently failed to get approval for a sub-sea link to the UK. It would seem that the business case for this project was indeed weak (Aune, 2003). Statnett is however going to undertake considerable investments in the Norwegian network over the coming years (see www.Statnett.no), and the Dutch regulator has now cleared a 700 MW cable to Norway (NorNed). 22 The ETS and its effects on generation are described extensively in the chapter on Germany in this book.

166


the relative importance of electricity based on “clean” power, accompanied by increasing electricity prices and a relative increase in the export of electricity from countries endowed with clean natural power resources, such as Norway. This is, indeed, what earlier studies of the effects of a common Nordic market for emission permits show; see Amundsen et al. (1999), Unger and Alm (2000) and Hauch (2003). It should be noted that these studies show that the introduction of a common Nordic electricity market has in itself lead to considerable reductions in CO2 emissions. The reason for this is that increased levels of electricity trade lead to substitution of low-cost and emission-free hydro power for electricity generated by high-cost emission-intensive sources.23 Nevertheless, introducing a common Nordic market for emission trade will provide an additional gain. The reason is that equalization of both marginal generation costs (resulting from trade in electricity) and marginal abatement costs (resulting from trade in permits) is more efficient than equalization of either of these alone.

5.12. The Nordic Market for Green Certificates In recent years, a Nordic market for TGCs has been under development in order to facilitate the use of the so-called green electricity, that is electricity generated by wind-power plants, new small hydro-power plants and power plants using bio-fuels. Sweden introduced its TGC system on May 1, 2003 and Norway will follow suit later. It is the intention that Norway and Sweden will start trading TGCs. The Nordic TGC system was to a large extent designed in Denmark and plans existed for its introduction already in 2001. Although the legal foundation for implementation was in place in Denmark, the TGC system has been put on hold, and no decision has been made as to when it will be implemented. Finland has no immediate plans of introducing a TGC system but does participate in the European Renewable Energy Certificate System (RECS), as do the other Nordic countries.24 Sellers of TGCs are producers of electricity using renewable sources. They are issued with a number of certificates corresponding to the amount of electricity they feed into the electricity network. Buyers of TGCs are consumers/distribution companies that are required by the government to hold a certain percentage of certificates corresponding to their total consumption/end-use deliveries of electricity. The percentage requirement functions as a check on total electricity consumption, as the total number of certificates available is determined by the total capacity of renewable technologies. The TGCs are thus seen as permits for consuming electricity. The system implies that producers using renewable energy sources receive both the wholesale price and a certificate for each kWh fed into the electricity network. In a

23

This result is, however, sensitive to the relative cost of electricity generation for the various technologies employed and to future market development. For instance, enlarging the power market to include Northern Germany may well result in larger imports of low-cost, carbon-intensive coal power and lead to more CO2 emission. In such a setting, the value of introducing ETS will be significant. The introduction of ETS may lead to an increase of end-user prices and reduced demand for electricity. Hence, even though ETS will stimulate electricity exports from countries rich in clean power resources, reduced demand for electricity may still dampen electricity transmission between countries. 24 The European RECS facilitates many support schemes for green energy, rather than being a support scheme itself. It is not restricted by national boundaries. RECS provides a mechanism for presenting production of a megawatt-hour of renewable energy by a unique certificate which can be transferred from owner to owner before being used as proof of generation, or exchanged for financial support (http://www.treckin.com)


167

competitive equilibrium, therefore, marginal generation cost of green electricity is equal to the marginal generation cost of non-green electricity plus the subsidy element (i.e. the TGC price). Hence, for the producer, the marginal cost of providing a kWh of green electricity is the same as that of non-green electricity, wherefore both technologies will be involved in power generation at the margin. In this way, the TGC system is supposed to stimulate new investments in green electricity generation. Analytically, the TGC system implies that the end-user price of electricity becomes equal to the wholesale price plus a percentage of the certificate price (assuming zero distribution costs), determined by the holding requirement. In a competitive electricity market, end-user price is also equal to a linear combination of the marginal generation cost of green electricity and the marginal cost of non-green electricity, with the percentage requirement as the combination weight. Hence a change of the percentage requirement will affect the relevant marginal generation cost. However, as the system is founded on a percentage basis, it is not necessarily true that an increase of the percentage requirement leads to more green generation capacity, though it will lead to less non-green generation capacity (Amundsen and Mortensen, 2001, 2002).25 5.13. Joint Effects and Compatibility of ETS and TGC Both the ETS system and the TGC system will affect CO2 emissions from electricity and heat generation. In this sense, these are two broad market-based measures (on top of other measures like emission standards, direct subsidies for renewables, etc.) to obtain the goal of reducing CO2 emissions (other considerations, such as supply security and infant industry protection also motivated the introduction of the TGC system). In order to provide an indication of the possible effects of these measures, we refer to some results from a simulation study of the Nordic electricity market focusing on Sweden, which, as mentioned, has already introduced a TGC market (Bergman and Radetzki, 2003). The introduction of the TGC system in Sweden (based on a 7% requirement) will lead to certificate prices at the stipulated upper price bound and an increase in the production of green electricity by 10 TWh. However, net export to the other Nordic countries will increase by 5.2 TWh. This somewhat surprising result is explained by the high price of certificates and the resulting low net cost of generating green electricity in Sweden. Hence the introduction of a TGC system in Sweden significantly affects investment decisions in the electricity industry. However, since in these simulations Sweden is assumed to be the only country applying a TGC system, the effect on the common Nordic electricity wholesale price is rather small. With a joint Nordic ETS system in place (or, equivalently, a joint CO2 tax), Bergman and Radetzki calculate that electricity consumption in Sweden would be reduced by an amount corresponding to the growth in demand over the period 2001–2010 which would otherwise have occurred (i.e. without such a scheme). Hence, CO2 emissions from the electricity sector are reduced at the expense of electricity consumption. However, as the cost increase is even higher in Denmark and Finland, the competitive position of the Swedish electricity industry improves and Sweden begins to export electricity to Denmark and Finland.26 25

Potentially, the TGC system involves some serious problems. For instance, certificate prices may be extremely volatile if wind power constitutes a large part of the renewable technologies. Also, problems of market power (i.e. gaming on the electricity and Green Certificate markets) may be severe and lead to a collapse of the system (Amundsen and Nese, 2004). 26 Broadly speaking, these conclusions seem to be in line with other model simulation studies of the joint effects of ETS and TGC systems in the Nordic countries; see Hindsberger et al. (2003) and Unger and Ahlgren (2003).

168


As for Sweden, the introduction of the TGC system in Norway is targeted at stimulating power generation from wind and bio-fuel sources. Norway is already well endowed with environmentally friendly hydro-power resources, but the construction of new large hydropower plants has more or less come to a halt, due to the lack of sites for large-scale developments. This is in part due to politically imposed environmental constraints and conflicting interests of land use. The emerging alternative is large-scale gas-power plants. The introduction of the ETS system will reduce the profitability of planned gas-power plants (unless they are not exempted from the ETS system). However, with the present ambition of the ETS system it is not quite obvious that the resulting increase in the wholesale price will be sufficient to stimulate the required investment in electricity plants based on wind power and bio-fuel. Thus, in order to realize these plans a sizable stimulus, such as a TGC market (or plain subsidies for that matter), is called for. Although the two measures work toward the same end (i.e. of reducing CO2 emissions) they are not quite compatible. For instance, under the ETS system, stricter emission constraints will lead to increasing permit prices, increasing generation costs for non-green power and thus increasing wholesale prices of electricity. Remuneration to generators of green electricity (i.e. the wholesale price of electricity plus the value of a TGC) will however decrease. The reason for this somewhat paradoxical result is the particular construction of the TGC system, whereby an increase of the wholesale price by 1 cent results in a reduction of the TGC price by several cents (depending on the size of the percentage requirement). The increase in the wholesale price, following a higher permit price, leads to a corresponding reduction of the margin between the end-user price and the wholesale price. This margin is equal to the TGC price multiplied by the percentage requirement. Hence if this margin is reduced by 1 cent, and the percentage requirement is 20%, the TGC price will be reduced 1/0.2 ⫽ 5 cents. The remuneration to a green producer (the sum of the wholesale price and the TGC price) is therefore reduced. The equilibrium effect of stricter emissions constraints is a reduction of both nongreen and green electricity generation, such that the percentage requirement is still satisfied (for further explanation, see Amundsen and Mortensen, 2001). Investigating this problem in a numerical model, comprising both an ETS system and TGC system for the Nordic countries, Unger and Ahlgren (2003) identify a negative effect of stricter CO2 constraints on green electricity generation capacity. They conclude that this effect is probably not very large in the longer run, as the effect of constraining CO2 emissions further will have only a small effect on electricity wholesale price. 5.14. Conclusion Since the deregulation process started in Norway in 1990, the Nordic electricity market has expanded steadily both in terms of countries and regions included and in terms of contracts traded. For the first many years the market developed rather smoothly without any significant calamities to be dealt with and the market, thus, started to resemble as success. This impression was reinforced after the critical dry Winter season of 2002–2003 when consumers, used to a sustained period of low prices, became confronted with sudden and sharp price increases. In general it seems fair to say that the market withstood this test of robustness and handled the supply shock rather well: prices adjusted rapidly and both demand and supply responded. Drastic measures such as rationing, foreseen by some and feared by many, were never warranted. As such, it would seem that the market has reached maturity. Nevertheless, this event also made it clear that potential problems still exist. As the overcapacity of the pre-reform era is vanishing, the market is becoming increasingly tight. One must consequently be prepared, not only for higher, but also for more variable prices.


169

Tighter market conditions also means that system operation will become more challenging, particularly ensuring the availability of sufficient reserves of generation capacity and, most importantly, interruptible demand so as to achieve system balance at all times. Nordic system operators seem on the whole to be well prepared, although further co-operation (integration) of system operations could potentially enhance efficiency. Tighter market conditions and increasing prices signal a need for capacity additions. So far, there is little evidence that generators do not react to such signals; indeed, at the time of writing capacity additions are either being planned or are already under way. However, uncertainty surrounding the political will to accept growth in electricity generation and consumption – exemplified with the lack of clarity about new measures for regulating environmental pollution – may undermine the willingness to invest. From a supply-adequacy point of view, it is important that such political and regulatory uncertainty be kept to a minimum. Further investment in transmission capacity is also likely to be warranted, especially in light of increasing concentration of the Nordic electricity industry. A low level of concentration may have been the most important factor underlying the success of the regulatory reform so far, and it would be a pity if mergers, horizontally and vertically, were allowed to undermine the performance of the industry. There is reason to believe that the difference in end-user prices between Norway and Sweden may be explained by a lack of competition in the Swedish retail market resulting from a combination of vertical integration between generation and retail and high levels of concentration. Similarly, tighter market conditions and more frequent occurrence of bottlenecks in the transmission system are signs of increasing segmentation of the market that may lead to higher prices also at the wholesale level. Only a combination of adequate investment incentives and strict competition policy can ensure the continued success of the Nordic electricity market. In conclusion, it seems fair to say that the electricity market reform in the Nordic countries has been relatively successful in comparison with other electricity markets, for example the California market that collapsed when exposed to severe shocks in 2000–2001. There seems to be four main factors explaining this (see Amundsen, 2005; Bergman, 2005; Amundsen and Bergman, 2006). Firstly, the market design of the Nordic market is simple but sound and to a large extent made possible by the large share of hydropower. Secondly, dilution of market power, attained by the integration of the four national markets into a single Nordic market, has been rather successful. Thirdly, there has been a strong political support for a market-based electricity supply system without intervention in the market mechanisms in stressful situations. Fourthly, the Nordic power industry seems to have a strong voluntary informal commitment to public service. Clearly, only the second and third of the above-mentioned factors are “transferable” as suggestions to other countries whishing to reform their electricity markets, while the first and the fourth to a large extent are country specific. Hence, the experiences of the Nordic electricity market suggest that a market for electricity works well if there are no price regulations and constraints on the development of financial markets and that there is continued political support for a market-based electricity supply system also when electricity is scarce and prices are high. References Amundsen, E.S. (2005). Smooth adjustment vs. melt-down: experiences from a Nordic supply shock and comparison with the California case. Paper Prepared for the EPRI International Conference Global Electri-city Industry Restructuring: In Search of Alternative Pathways, San Francisco, California, May 11–12, 2005.

170


Amundsen, E.S. and Bergman, L. (2006). Why has the Nordic electricity market worked so well? Utilities Policy (forthcoming). Amundsen, E.S. and Mortensen, J.B. (2001). The Danish Green Certificate System: some simple analytical results. Energy Economics, 23, 489–509, with Erratum in Energy Economics, 24, 523–524. Amundsen, E.S. and Nese, G. (2004). Market power in interactive environmental and energy markets: The case of Green Certificates. SNF-Report 10/2004. Samfunns- og næringslivsforskning AS, Bergen. Amundsen, E.S., Nesse, A. and Tjøtta, S. (1999). Deregulation of the Nordic power market and environmental policy. Energy Economics, 21, 417–434. Aune, F.R. (2003). Fremskrivning for kraftmarkedet til 2020 – virkninger av utenlandskabler og fremskyndet gasskraftutbygging. Reports No. 2003/11. Statistics Norway. Bergman, L. (2005). Why has the Nordic electricity marked worked so well? Paper Prepared for the EPRI International Conference Global Electricity Industry Restructuring: In Search of Alternative Pathways, San Francisco, California, May 11–12, 2005. Bergman, L. and Radetzki, M. (2003). The Continue Project: Global Climate Policy and Implications for the Energy Sector in a Small Open Economy: The Case of Sweden. Multi-Science Publishing Company Ltd, Stockholm. Bergman, L., Brunekreeft, G., Doyle,C., von der Fehr, N.-H.M., Newbery, D.M., Pollitt, M. and Regibeau, P. (1999). A European Market for Electricity? Monitoring European Dergulation 2. Centre for Economic Policy Research, London/Stockholm. Bjørndal, M. and Jørnsten, K. (2001). Koordinering av nordiske systemoperatører i kraftmarkedet – gevinster ved bedret kapasitetsutnyttelse og mer fleksibel prisområdeinndeling. SNF-Report No. 29/01. Samfunns- og næringslivsforskning AS, Bergen. Bye, T. (2003). A Nordic energy market under stress. Economic Surveys, 4, 26–37. Statistics Norway. Bye, T. and Bergh, P.M. (2003). Utviklingen i energiforbruket i Norge 2002–2003. Reports No. 2003/19. Statistics Norway. Bye, T. and Larsson (2003). Lønnsomhet ved tilbakesalg av kraft fra kraftintensiv industri i et anstrengt kraftmarked. Økonomisk Forum, Vol. 57, 26–29. Bye, T., von der Fehr, N.-H.M., Riis, C. and Sørgard, L. (2003a). Kraft og makt – en analyse av konkurranseforholdene i kraftmarkedet. Report of an Expert Group Appointed by the Norwegian Ministry of Labour and Administration. Bye, T., Hansen, P.V. and Aune, F.R. (2003b). Utviklingen i energimarkedet i Norden 2002–3. Reports No. 2003/21. Statistics Norway. Hauch, J. (2003). Electricity trade and CO2 emission reductions in the Nordic countries. Energy Economics, 25, 509–526. Hindsberger, M., Nybroe, M., Ravn, H. and Schmidt, R. (2003). Co-existence of electricity, TEP and TGC markets in the Baltic Sea Region. Energy Policy, 31, 85–96. IEA (2000). CO2 emissions from fuel combustion 1971–1998. IEA Statistics, ISBN 92-64-08506-8. Nordel (2001). The transit solution in the Nordi electricity power system. Feature article in 2001 Annual Report. Nordel (2003). The future infrastructure of the Nordic countries. Feature article in 2003 Annual report. Treckin (2004). The one-stop information network for tradable renewable certificates, http://www. treckin.com Unger, T. and Ahlgren, E.O. (2003). Impacts of a common green certificate market on electricity and CO2 emission markets in the Nordic countries. In T. Unger (ed.), Common Energy and Climate Strategies for the Nordic Countries – A Model Analysis. Ph.D. dissertation, University of Gothenburg, Gothenburg, Sweden. Unger, T. and Alm, L. (2000). Electricity and emission-permits trade as a means of curbing CO2 emissions in the Nordic countries. Integrated Assessment, 1, 229–240. von der Fehr, N.-H.M., Amundsen, E.S. and Bergman, L. (2005). The Nordic Market. Signs of stress? The Energy Journal, 61–88, European Electricity Liberalization Special Issue. von der Fehr, N.-H.M., Hagen, K.P. and Hope, E. (2002). Nettregulering. SNF-Report No. 29/01. Samfunns- og næringslivsforskning AS, Bergen.

PART III Evolving Markets


Chapter 6 The Electricity Industry in Australia: Problems Along the Way to a National Electricity Market ALAN MORAN1 Institute of Public Affairs, Melbourne, Australia

Summary2 In Australia prior to 1994 virtually all electricity was supplied through vertically integrated state monopolies. A decade later, the integrated monopolies had been disaggregated into different businesses with the competitive aspects of supply (generation and retailing) reconstituted into dozens of independent firms, many of them privately owned and the rest “corporatized” and operating at arms length from their government owners. Monopoly aspects of supply are regulated by agencies independent from the jurisdictional governments. The key milestones have been as follows: Sequence of events: ● ●

●

●

● ●

●

Industry Commission Report into electricity, 1991. Institute of Public Affairs/Tasman Institute report on Victorian electricity corporatization and privatization, 1991. National Grid Management Council’s National Electricity Market (NEM) Paper Trial, 1993/1994. Victorian Electricity Market was commenced on 1994. This comprised six major generation businesses, a transmission business and five distributor/retailers. Distribution regulated by independent body. Market gradually opened to competition in 1995–2001. Some price caps remain on household supply. Victorian electricity and gas privatizations 1995–1999. Competition Principles Agreement, 1995 (A$4.2 billion of Commonwealth funds were set aside for the period of 2005/2006 to implement electricity reform). New South Wales (NSW) Electricity Market, 1996. Competitive arrangements were established that were similar to Victoria’s. Market opened to competition in 1996–2002 with some price controls on households remaining.

1

Helpful comments were received from many people including Perry Sioshansi, Paul Simshouser (Braemar Power), Darren Barlow (Ergon Energy) and Ben Skinner (TRUenergy). 2 Prices here are quoted in $A or cents, which refer to Australian cents. The Australian dollar is worth around 75 US cents.

173

174

● ●

●

● ●

●


Agreement on National Electricity Code, 1996. Start of NEM, 1998 with the National Electricity Code Manager and the Australian Competition and Consumer Commission (ACCC) setting rule changes at the national level and ACCC setting transmission prices. Queensland Electricity Market, 1998. Market opened to competition in 1998–2004 except for households. South Australia privatization, 2001. Full retail competition phased in 1998–2003. Australian Energy Regulator (AER) (price setting) and Australian Energy Market Commission (AEMC) (Rule changes) commenced operation in 2005. Western Australia de-aggregation of supply but with the generation left within one business, 2005.

Outcomes of these developments have been reductions in prices, especially for commercial users, which were previously subject to Ramsey-type price gouging.3 The reformed system has delivered increases in capacity in line with market needs and vast improvements in productivity and reliability across the industry. Government interventions have, however, not been eliminated and continue to threaten the on-going orderly development of the market. Among these potential market distortions are the ramifications stemming from state governments’ ownership of over half of the industry. Although all government businesses are corporatized and operate under company law, government ownership brings corporate inflexibilities and sometimes means political interference in key commercial decisions. More generally, electricity remains an industry with a high political profile. Federal and state governments see electricity supply and pricing as providing them a somewhat unique legitimacy to control. At the very least, this brings distractions to the industry’s entrepreneurship and there is ample risk of more serious consequences to the industry’s efficiency.

6.1. The Market Breakdown and Supply Profile 6.1.1. The market profile Australian electricity demand is about 200,000 GWh/annum, which is transmitted along some 850,000 km of lines of which some 26,000 km is along circuits rated at 220 KV or more. There are some nine million customers with demand being divided almost equally between households and industry. Figure 6.1 illustrates this. The geographic spread of generation facilities and transmission and distribution lines is illustrated below.

6.1.2. The fuel profile of Australian electricity generation Australia is fundamentally a coal-based electricity system. Coal represents some 85% of electricity supply, roughly one third of which is Victorian and South Australian brown coal.

3

Under the state owned vertically integrated monopolies, large, footloose users negotiated cost-based prices (below cost in the case of some aluminum smelting contracts). Household supply benefited from a cross-subsidy paid for in higher prices to most business customers. Since the market has been operating almost all business customers and many household customers have seen real price reductions.

175

The Electricity Industry in Australia Electricity consumption by sector 1.1%

1.5%

23.8% 46.9% 26.7%

Industry Residential Transport

Commercial Agriculture

Fig. 6.1. Electricity consumption. Source: NEMMCO.

Hydro capacity is relatively small despite the large land area and is nearly fully developed. Tasmania has further minor potential but green activism will prevent any substantial new development. The share of hydro within total supply has, therefore, been falling to its current level of about 7%. Gas has shown a modest increase growing from less than 2% of generation 15 years ago to 7–8% at present. Much of this is in the least heavily populated states of South Australia, Western Australia and the Northern Territory where coal is more expensive. Elsewhere gas mainly fills a peaking role due to its lower capital but higher fuel cost. Figure 6.2 illustrates the fuel sources of electricity. Australia’s eight States and Territories vary in size, with New South Wales and Victoria being the most populous and the two Territories, the city-state Australian Capital Territory and the vast Northern Territory each having populations of a few hundred thousand. Most people live along some 4000 km of coastline from Adelaide in South Australia to Cairns in North Queensland. This has resulted in a long, skinny interconnected grid: the NEM. The

176


island of Tasmania is in the process of connecting to the NEM via a 600 MW undersea cable. The only other significant population center, South West Western Australia, has its own grid, which is not economic to connect to the NEM. The same applies for smaller isolated population centers such as the Northern Territory. Each state has only one pricing node4 that sets price for all customers and generators within that state. Generally there is little transmission congestion within the states, and a moderate amount between the states – with “interconnectors” constraining ⬃5% of the time. At these times, the regional prices diverge, sometimes by thousands of dollars per MWh, but very rarely for more than a few hours. Queensland and Victoria are major net exporters of electricity via the NEM (to South Australia and New South Wales respectively) and New South Wales is also an exporter because it hosts the jointly owned Snowy Mountains Hydro-Electric facility. Tasmania is hydro based and Basslink will allow Tasmania to export its hydro at times of high mainland prices and import baseload low-priced mainland-generated electricity at other times. Figure 6.3 illustrates the relative size of the NEM States in terms of energy sent out.

Generation by fuel source 250

GWh

200 150

Gas Coal Hydro

100 50 0 1991

1993

1995

1997

1999

2001

2003

Fig. 6.2. Source: ESAA.

Total energy sent out 2003–2004 2.5% 6.5%

28%

27%

36%

QLD NSW TAS

VIC SA

Fig. 6.3. Source: NEMMCO.

4

NSW has a second pricing node in the snowy mountains scheme, however there are no significant customers at this node.

The Electricity Industry in Australia

177

6.2. Reforming the State Owned Integrated Utilities 6.2.1. The process of reform The development of a market for electricity got underway in the early 1990s and had three precursors: ●

●

●

the recognition that other countries were achieving considerably greater efficiencies than Australia in electricity supply;5,6 National Competition Policy (NCP) involving a general review of the operations of “essential facilities” (which were, in the main, owned by governments) and a requirement that they be opened to non-affiliates on reasonable terms; and the consequences of poor financial circumstances in the States of Victoria and South Australia resulting in new governments which sold its energy assets partly in pursuit of a privatization agenda and in part to reduce debt.

The State of Victoria initiated the reform process. By the early 1990s the then Labor Government had commenced a process of reform particularly focusing on labor shedding in the monopoly supplier, the State Electricity Commission of Victoria (SECV). A Liberal (Conservative) Government, which won office in 1992, set about a much more aggressive reform and privatization process. That Government had a strong philosophical belief in the beneficial effects that capital markets could bring to a business as well as being attracted to the idea of transferring risk to private equity. However, the Government was equally determined to get the structure of the market right first. In this, lessons were learned from the outcome of the structural weaknesses in the UK wholesale generation market, where the two dominant suppliers were able to operate so that prices remained high (See Newbery). Great care was therefore taken to establish competition in supply with this being given a higher priority than maximizing sale proceeds. 7500 MW of plant was privatized as seven generating companies with transitional prohibitions on re-aggregation. For downstream supply, five distribution/retail businesses were created and transmission assets were vested in a single business. An independent regulatory framework, the Office of the Regulator General (ORR) later renamed the Essential Services Commission (ESC) was put into place to set prices on the monopoly network assets. This body also gathered performance data that would prove invaluable in defusing claims that privatization had led to higher prices and lower reliability. The program for reform was a challenging task. Not only did it involve the need to establish a regulatory framework, in which Australia had no experience, but also because the reform process faced considerable hostility from the unions, the Labor opposition and much of the media. However by the time of the Liberal Government’s fall in 1999, the electricity industry had been privatized and assets transferred to seven generation businesses, five distributor/retailers plus a transmission business. There has since been some merger rationalization as well as some new entry. The sale process earned A$23 billion, far more than the $9–10 billion that was widely expected. The gas industry’s downstream assets were also privatized; gas production was always privately owned. 5

Project Victoria: A Rebuilding Strategy for Electricity in Victoria, Tasman Institute/Institute of Public Affairs, 1991. 6 Industry Commission, Energy Generation and Distribution Report, No. 11, 1991, http://www.pc.gov.au/ic/ inquiry/11energy/finalreport/index.html

178


The fully privatized Victorian electricity and gas industry has three main retailers (CLP Power’s TRUenergy, Origin Energy, and AGL), four distribution businesses (CKI-Powercor, Alinta, AGL, and SPI), a single transmission business (SPI) and five major generators (International Power, CLP Power’s TRUenergy, Loy Yang Power, Ecogen, and Southern Hydro). There have also been significant new entrants into the competitive generation and retailing businesses. South Australia, which is only one quarter Victoria’s size in terms of customers, privatized its industry (in 1999 and 2000). Its previous government monopoly supplier had been the ETSA Corporation. As in other jurisdictions, ETSA’s disaggregation preceded the industry’s full privatization. There is now one distributor, the CKI owned ETSA Utilities and one transmission company, ElectraNet. Only one retailer was created at privatization, which AGL bought, however there has been significant new entry. Privatized generation is owned by International Power, NRG, and TRUenergy. AGL and Origin built significant new generation capacity early this decade. In Queensland, the Queensland Electricity Commission (QEC) was a vertically integrated organization responsible for virtually all electricity supply and was disaggregated in several steps into separate distributor/retailers, a single generation business and an SPI. In 1997 a further progressive disaggregation took place which has resulted in the present industry profile comprising government ownership of two main distributor/retailers (Energex and Ergon), a transmission business (Powerlink) and four generation businesses (Enertrade, CS Energy, Stanwell, and Tarong Energy). There is also private ownership of generation including Intergen and NRG. The New South Wales Government’s prior ownership of the integrated Electricity Commission has undergone several iterations of disaggregation but the separate businesses remain under government ownership. Individual suppliers include four distributor/retailers (Energy Australia, Integral, Country Energy, and Australian Inland) a transmission business (Transgrid) and three coal-based generation businesses (Delta Electricity, Macquarie Generation, and Eraring Energy). There are also two smaller private generators and the Snowy Hydro jointly owned with the Victorian and Commonwealth Governments. The Commonwealth, NSW and victorian governments, which jointly own Snowy, announced in early 2006 that they are to privatize it. The other jurisdictions have also disaggregated their formerly integrated industry but Tasmania (which will be linked to the mainland grid in 2006) and Western Australia have retained government ownership. The major ESI businesses and their ownership are shown in Table 6.1. Overall Australia’s ESI remains mainly under government ownership as illustrated in Figure 6.4. 6.2.2. The national setting for electricity market reform The Victorian disaggregation and privatization took place as a result of financial pressures and pro-competition/privatization views that were especially prevalent in that state. In the case of other states, the reforms followed on a program of NCP, agreed by the state and Commonwealth Governments in 1996. This was given expression in a new provision of the Trade Practices Act, Part IIIA, an important provision of which required essential facilities to be opened to competition, clear structural separation for the competitive and monopoly parts of integrated businesses and, where practicable, non-natural suppliers were to be divided into rival businesses. The competition policy reforms called for natural monopolies to be opened to all users on terms that were fair and reasonable. Provisions were made to ensure that regulatory

179

The Electricity Industry in Australia Table 6.1. Major Australian electricity supply businesses. Ownership

State of operations

Major generators Macquarie Delta Snowy Eraring TRUenergy CS energy Loy Yang Intergen Tarong NRG flinders International power Stanwell Origin Ecogen NRG Gladstone Hydro Tasmania Southern hydro Enertrade

NSW government NSW government NSW, Vic federal governments NSW government CLP Qld government/Intergen AGL, Tokyo electric and others International consortium Qld government NRG International power Qld government Australian private International consortium NRG, Comalco Tas government Meridian energy Qld Government

NSW NSW NSW, Vic NSW Vic, SA Qld Vic Qld Qld SA SA, Vic Qld Vic, SA, Qld Vic Qld Tas Vic Qld

Major transmission SPI Transgrid Powerlink Electranet

Singapore power NSW government Qld government public private consortium

Vic NSW Qld SA

Major distribution AGL Alinta Aurora Citipower/Powercor/ETSA Country Energex Ergon energyAustralia Integral energy SPI

Australian private Australian private Tas government CKI NSW government Qld government Qld government NSW government NSW government Singapore power

Vic, SA Vic, WA Tas Vic, SA NSW Qld Qld NSW NSW Vic

Major retailers AGL Aurora energyAustralia Integral energy Country Energex Ergon Origin TRUenergy

Australian private Tas government NSW government NSW government NSW government Qld government Qld government Australian private CLP

Vic, SA Tas NSW NSW NSW Qld Qld SA, Vic Vic SA

180


Ownership of NEM assets

36%

64%

43%

Generation

57%

Transmission Public

50%

50%

Distribution

55%

45%

Retail

Private

Fig. 6.4. Source: ESAA

arrangements were put in place to determine where such access was required and, in the event of disputes, the prices at which the access was to be made available. Far and away the most important monopolies over essential services were those of governments themselves – in addition to electricity and gas they included ports, airports, rail, and telecommunications. Indeed, the government owned or controlled businesses were the only ones that in practice had the integration and monopoly characteristics that required the regulatory control.7 6.2.3. Market governance The national market was devised jointly by the state and Commonwealth governments and refined and approved by the national regulator, the ACCC. It originally envisaged two bodies: the National Electricity Code Administrator (NECA) and the National Electricity Market Management Company (NEMMCO) being responsible respectively for the market rules and the market scheduling and planning matters. In the event, the confluence of the electricity market developments and NCP meant that the national regulatory body, the ACCC had to have a role in approving market rule changes because of the inherent monopolistic “collusion” that the NEM entails. The ACCC also was given the role in setting prices for the transmission lines and for gas transmission and the market code for gas. In addition, state regulators were put in place to set prices for the distribution lines so that in all there were some 12 bodies involved in economic regulation of the industry. Also, there were Ministerial Councils that set agenda issues and generally sought to influence the market. Fortunately, inconsistencies in decisions from such a plethora of regulatory bodies was not as serious as it might have been since there were strong liaison relationships forged through an informal “Energy Regulators’ Forum”. Even so, the process of National Electricity Code change was proving somewhat unwieldy with each proposal being examined first by NECA and then by the ACCC. New arrangements were agreed in 2005 that introduce a rationalized control of the industry so that the AEMC is responsible for market rules for both gas and electricity and the AER is responsible for policing the rules and for network price setting at all levels of supply. NEMMCO’s functions (as Market/System Operator) remain as those of the ACCC, which also has common staff with the AEMC and AER, with regard to control of mergers and monopolies. It is, of

7

Exceptions included gas where AGL had a monopoly in NSW and, arguably, BHP steel, the integrated plant of which had been found by the High Court, in the Queensland Wire case, to be required to be opened to competition. (See W. Pengilly www.deakin.edu.au/buslaw/aef/publications/workingpapers/ swp2004_181.pdf)


181

course, uncertain that the changed arrangements will prove more workable especially as there are statutory requirements for these bodies to follow the policies established by the collective state and federal governments. The AEMC is responsible for rule making and market development. The rule-making role does not involve initiating changes to the Rules other than where the change involves correcting minor errors or where the change is of a non-material nature. Rather, the role involves managing the rule change process, and consulting and deciding on rule changes proposed by others. In regard to its market development function, the AEMC conducts reviews at the request of the Ministerial Council on Energy or at its own volition on the operation and effectiveness of the Rules or any matter relating to them. In doing this, the AEMC relies on the assistance and cooperation of industry relationships and interested parties in its decision making. 6.2.4. Structural developments At the onset of both the Victorian and the later National Market rules, the provision implementing structural separation of generation, transmission and distribution/retailing contained no specific long-term measures to prevent re-aggregation. This was because there were no firm views as to the most productive structure of the industry, only that the previous state owned integrated monopolies were not optimal. Although retailing and distribution were sold as combined units, they were to be “ring fenced” to prevent the distribution business favoring its affiliate. The ring fencing has generally proved satisfactory. Not only have incumbent affiliates not been favored but new retailers have entered the market and all five of the original Victorian host distribution business/retailers now have separate companies handling the two activities.8 This reflects the very different types of business involved. Retailing involves strong marketing and risk management skills in assembling and promoting packages of supply, while distribution is much more concerned with maintaining and reducing costs of an established business. Two of the three major retailers, Origin and TRUenergy, no longer are affiliated with a distributor. Their disaggregation was driven by a search for improved shareholder value whereby different types of business can appeal to different types of investor. There has also been a trend towards an unexpected form of re-aggregation. This has not been to restore monopolistic supply, which in any event would be combated by general anti-trust laws administered by the ACCC. Instead, all the privatized retailers and many of those remaining in government hands have moved to acquire some generation of their own. Also, Snowy Hydro and several privatized generators have created retail arms. These developments are a function of the need that generators and retailers see to manage risk. Risk management strategies that firms have employed, cover a gamut from different forms of contracting through ownership of supply. Seeking some control over their supply (for retailers) or customers (for generators), is a risk management strategy that follows from the wholesale price shifts to which electricity is now subject. Early concerns that retailer and generator amalgamations would lead to monopoly power and market inefficiencies have tended to abate as a result of evidence that competition is bringing lower prices and retail churn. The generator–retailer re-aggregation that has occurred has not been anti-competitive. The fact is that electricity retailers like other firms 8

Mergers have reduced the original five host distributors/retailers to three retailers and four distributors. Only one of the retailers retains common ownership with its “host” distribution business.

182


that assemble supplies sourced from affiliated and non-affiliated businesses are forced by market circumstances to ensure that the affiliates are not favored. The risk of high-price occurrences are too great for electricity retailers to gamble on fulfilling most of their needs in-house and the costs of alienating other sources by unprofessional behavior far outweigh any short-term benefits possible through collusion with in-house suppliers.

6.2.5. Retail price control and competitive churn Retail was a part of the electricity industry envisaged as being contestable and requiring no more regulation than is found in other retail activities. Given its relatively small share of aggregate revenue, retail was also envisaged as being only a minor actor within electricity supply. However, retailing’s importance has been re-assessed. In a competitive situation retail, as the interface with the consumer, drives efficiencies by signaling demand shifts and eroding cross-subsidies. For businesses, full retail competition has been extended to all but the smallest customers and in the four major eastern states probably half the commercial load has shifted from its original retailer. Governments have been more cautious about deregulating household supply. In NSW, South Australia and Victoria, retail competition at the household level has been accompanied by maximum prices that make it less attractive for retailers to poach customers. Nonetheless there has been a quite considerable churn rate – some 16% in NSW, 42% in South Australia and 44% in Victoria after 3 years of open competition.9,10 In Queensland, Tasmania and Western Australia households are captive to their host retailers which are all owned by the respective state government. In Queensland, household rates are fixed and managed so that Brisbane subsidises the provincial areas. The Brisbane-based supplier is levied a surcharge which is passed to the supplier of provincial areas. The crosssubsidy would be placed under considerable pressure with full retail competition (See Haas in this volume) and the electoral implications of this present the main barrier to the Queensland Government allowing full retail competition. In September 2005, the Queensland Government announced that full retail competition, albeit with a regulated price cap, would be introduced in 2007. Table 6.2 shows the timetable for retail competition by state and user tranche. 6.2.6. Other forms of retail regulation On top of price safety nets, the Labor state governments have all imposed their social and green policy objectives via retail regulations for domestic customers. This has resulted in a considerable mish-mash of compliance requirements for retailers selling to small customers and reduced the potential for competition. From the social policy side we have seen prohibitions on pre-payment meters and latepayment fees supposedly in the interest of protecting the poor. 9

For Vic and NSW, http://www.nemmco.com.au/data/ret_transfer_data.htm. For S.A. http://www.saiir. sa.gov.au/site/page.cfm?u⫽4&c⫽1435 10 These figures refer to small customers, some 95% of which are households. They also include some double counting where customers have changed retailers more than once. An alternative measure is of customers who are no longer with their original retailer. For Vic as at September 2005, this is 32.5% and for NSW 11%; these latter figures exclude customers who have new supply contracts with their original retailer.

183

The Electricity Industry in Australia Table 6.2. Timetable for retail competition. Date for eligibility

Site thresholds

Estimated number of customers

Percentage of total energy

New South Wales

October 1996 April 1997 July 1997 July 1998 January 2001 July 2001 January 2002

⭓40 GWh ⭓4 GWh ⭓750 MWh ⭓160 MWh ⭓100 MWh ⭓40 MWh All sites

47 660 3560 10,860 19,000 53,000 3,000,000

14 29 40 47 49 53 100

Victoria

November 1994 July 1995 July 1996 July 1998 January 2001 January 2002

⭓5 MW ⭓1 MW ⭓750 MWh ⭓160 MWh ⭓40 MWh All sites

47 380 1900 6900 N/A 2,100,000

23 29 41 49 N/A 100

Queensland

March 1998 January 1999 January 2000 1 July 2004

⭓40 GWh ⭓4 GWh ⭓200 MWh ⭓100 MWh

80 540 8110 7890

16.0 15.0 15.0 8.0

South Australia

December 1998 July 1999 January 2000 January 2003

⭓4 GWh ⭓750 MWh ⭓160 MWh All sites

160 760 3360 720,000

30.0 40.0 50.0 100.0

Western Australia

July 1997 July 1998 January 2000 July 2001 January 2003 January 2005

10 MW 5 MW 1 MW 230 KW 34 KW 50 MW

N/A N/A 120 450 2550 10,000

N/A N/A N/A N/A N/A N/A

Tasmania

July 2006 July 2007 July 2008 July 2009 July 2010

20 GWh 4 GWh 750 MWh 150 MWh All customers

10 54 295 1030 230,000

N/A N/A N/A N/A N/A

From the green side we have seen irksome rules like requiring minimum area allocations on retailers’ bills for graphs on usage and implied carbon release. More significant are the requirements on retailers to source different percentages of variously defined green power within their aggregate supply. This is addressed in Section 6.7. 6.2.7. Control of transmission Each State has one transmission owner but the planning models are inconsistent. In Queensland, NSW and Tasmania, the transmission owner itself is responsible for planning. In Victoria, a government agency, VENCorp, is responsible for planning the network and SPI owns and operates the assets, whilst South Australia has a hybrid model. SPI in Victoria is unique in that it acquired assets formerly owned by TXU so that it now also owns part of Victoria’s distribution network.

184


A centrally planned provision of transmission was the basic model adopted from the outset. However, it was also recognized that transmission and new generation are alternatives. If transmission is provided free or at regulated prices this may discourage a more rational and lower cost development of new generation. The trade-off between nearby and remote generation (via transmission) is uniquely critical for Australia, where distances between load centers and therefore the cost of transmission are very large, and fossil fuel sources are relatively inexpensive and quite widespread. This led to provision being made for entrepreneurial interconnects in the National Electricity Law. And Transenergie, a subsidiary of Hydro Quebec, built two of these entrepreneurial links where transmission links were less than robust. Transenergie sought to finance these links by selling generators’ access rights to markets and by itself arbitraging price differentials. This merchant transmission gave rise to issues concerning the circumstances under which a regulated augmentation of links should be permitted. A long series of hearings on a regulated link between South Australia and New South Wales resulted in stalemate, with the NSW government transmission business (Transgrid) apparently abandoning its proposal, possibly because NSW is not envisaged to have a generation surplus in future years.11 In the event, the merchant links in Australia could not compete against the links receiving a regulated return and have applied for and been given regulated status.12 The danger is that links which are financed by a compulsory charge on the customer, might lead to incentives to site generation in places that are distant from major markets. If someone else is paying for transmission, the rational generation business will be indifferent to its costs thus distorting the efficient trade-off transmission costs and generation costs. It is argued that the externalities are too great to allow profitable merchant transmission since the benefits of lower prices (actually arbitraged prices) accrue to all and not only to those paying for the asset. However, this is not markedly different from the situation concerning a new generation facility, which will tend to suppress the price of all delivered electricity in its interconnected region. Few would argue that by analogy all generation should therefore be government owned or subsidized even though many argue for a form of general overhead support in the form of capacity payments. The fact is that supply across the economy is seldom unaccompanied by some externalities. Associated with the claim that transmission would be inadequately provided in the absent of it being made subject to regulated support, is the contention that a transmission 11

This has brought a voluminous level of studies. Those in Australia include the skeptical like Mountain, B. and Swier, G. Entrepreneurial interconnectors and transmission planning in Australia, The Electricity Journal, March 2003. London Economics in its work for the ACCC (Review of Australian Transmission Pricing, 1999), also concluded that entrepreneurial links could not cover their fixed costs. This skepticism is also seen in the work of Joskow and Tirole (e.g. Merchant Transmission Investment, CMI Working Paper, 24 The Cambridge-MIT Institute, 2003). The Australian 2002 Paper on Independent Review of Energy Market Directions (www.energymarket review.org) saw a possible role. Littlechild has been more supportive both in studies in Australia and Argentina (e.g. Littlechild, S. (2004) “Regulated and merchant interconnectors in Australia: SNI and Murraylink revisited.” Applied Economics Department and The Cambridge-MIT Institute, Cambridge University, Cambridge Working Papers in Economics CWPE No. 0410 and CMI Working Paper 37; and Stephen C. Littlechild and Carlos J. Skerk Regulation of transmission expansion in Argentina CMI, Working Paper 61, University of Cambridge, Department of Applied Economics, 15 November 2004). Further complications to the issue also result from the flowgate and nodal pricing debate. 12 This tends to confirm a dimension of Hogan’s “slippery slope” hypotheses whereby due to its ability to compel payment government ownership tends to drive out private ownership. Hogan W.W., Transmission Market Design; http://ksghome.harvard.edu/⬃whogan/trans_mkt_design_040403.pdf


185

line has market power and its prices should be regulated. However, for the most part, transmission inter-ties or inter-connects offer no more market power than that of a significant generator portfolio. Inter-ties in Australia can account for some 35% of supply (Victoria to South Australia) but normally provide much less than this. Their market power is confined to influence over those wishing to export and such firms are normally capable of writing contracts to cover and vulnerabilities they foresee. Issues on how best to allow expansion of transmission, especially in terms of the regional linkages has been subject to heated debate in Australia. An uneasy compromise is presently in place for transmission under which regulated links will be permitted as long as a net market benefit is judged by the regulator to be the outcome and as long as the proposed link is the best of a range of feasible alternatives. This, however, remains dissimilar from the decision making structure that is seen in the generation sector or in markets more generally since it may incorporate some to the network benefit externalities which an comparable investment in a new generator would not capture. The competing solutions that generation and transmission often offer mean disputes about the merits of a new transmission solution are likely to remain. These may be exacerbated since the Queensland Government has encouraged new government owned capacity to be built in that State, driving down prices below those in NSW and is seeking to augment transmission links. Other states regard this as facilitating dumping and are opposing to having expanded capacity financed as a regulated link since most of the costs fall directly on consumers. These considerations have been further complicated by the growth of subsidized wind generation. Wind power is always likely to be relatively dispersed and remote and, in addition to production subsidies, its sponsors have already extracted concessions from some governments that smear its transmission costs. Over a thousand MW are planned in South Australia where conventional capacity is only 3000 MW. 6.2.8. Network price setting The ACCC is responsible for setting prices for electricity transmission lines (as well as those of gas). Local State regulators at the present time are responsible for setting distribution prices. The price setting process has assumed a vast complexity as the regulator and the businesses each hire accountancy and economics advice to determine the appropriate prices. Regulated businesses are never likely to express satisfaction with the determinations of a regulator, but for the most part over recent years the outcomes have been more predictable and less contentious. There remains the risk that price cuts can induce sub-optimal investment. In Victoria, following a price reduction on distribution businesses averaging 15% in 2001 further real price cuts averaging 14–26% are proposed for 2006 when the businesses claim that all the fat was cut out in the first price re-set. In the case of gas transmission emerging competition is generally to be seen across the country. In spite of this, the ACCC has been seeking to maintain its regulatory powers and its decisions have been at variance with those of the market itself. Both the Minister for Energy,13 who has review powers over certain ACCC decisions, and the Productivity Commission,14 the advice of which has been sought by the government, have suggested 13

Macfarlane I (Minister for Industry, Tourism and Resources) (2003), Applications for Revocation of Coverage of Certain Portions of the Moomba to Sydney Pipeline System: Statement of Reasons, www.industry.gov.au/ content/itrinternet/cmscontent.cfm?objectid⫽F14AF2D8-0F2F-FF46-E371BA8885EA4888& indexPages⫽/content/whatsnew.cfm&CFID⫽30998&CFTOKEN⫽24724739. 14 10 August (2004). Review of the gas access regime, productivity commission, report No. 31.

186


that the ACCC’s powers be curtailed somewhat where, as is evidenced with gas, competition is itself providing adequate market disciplines. The danger of a regulator setting transmission prices too low is that this will result in inadequate new investment. Setting prices too low on existing network facilities (transmission and distribution), irrespective of any general provisions that are in place to maintain quality, will also result in inadequate maintenance. A claimed outcome of mandated price cuts by the Queensland regulator was a series of blackouts in 2004. On the other hand is the risk of setting regulated transmission returns too high. Where these are provided by compulsory charges there is likely to be over-building with the previously discussed adverse impacts on alternative, more efficient solutions like additional local generation capacity. This has been foreseen in the NEM Code which has provisions for rationalizing redundant facilities. These however are difficult to activate.

6.3. Operations of the National Market 6.3.1. The spot and contract markets The Australian National Market is underpinned by a “gross pool” system under which all major suppliers must bid. There is no capacity payment system and the market for energy is linked and simultaneously cleared to eight separate markets for reactive power and other “ancillary services”. Although virtually all power must be bid into the pool (a major exception being wind power, which must be taken as it comes) few customers or suppliers would find it prudent to rely solely on pool prices. In effect the market is now a multiplicity of bilateral contracts between generators and retailers, usually in the form of contracts for difference, strongly underpinned by the “gross pool”. Essentially the NEM is a one-way market with the demand emerging from whatever end-users require and with supply being offered by generators at different price levels. Through the pool, all supply is paid the same price, that of the highest bid supply that is dispatched. In the short term, generators are relatively indifferent to the price of that part of their supply for which they have contracts (probably 90% plus of the market) and will usually bid close to their marginal costs for this part of the load.15 Compared to a price norm of about $A30/MWh, at present prices can rise to as much as $A10,000/MWh which makes retailers especially keen to be fully contracted or to manage spot exposure through ownership of peaking capacity. A complex National Electricity Code controls the rules under which generation and transmission are placed on the market. Generators bid in a maximum of 10 price bands and although there are restraints on price re-bidding, there are none on shifting energy between different price bands meaning, in effect, that prices can be changed at any time prior to 30 seconds of dispatch. Each generator has controls built into the dispatch algorithm covering ramp rates and minimum loads. Loads can also bid to be offloaded though few do. In some cases this is due to lack of “smart” metering though it is likely that only those users, for example smelters, who have 15

For a discussion on whether the generators exercise market power see, Moran, “Is there market power in Australian electricity generation?” ACCC Regulatory Conference, July 2004, http://www.ipa. org.au/ files/ammarketpower.pdf. For an analysis of rational bidding strategies see Peter Cramton, Competitive bidding behavior in uniform-price markets. Hawaii International Conference on System Sciences, January 2004.


187

energy as a very large share of their aggregate costs would consider it worthwhile to make savings by shutting down and re-starting. It is an assumption of those calling for mandatory roll-outs of smart meters that this will markedly influence household behavior. It is, however, likely that any such influence would be in the form of contracts that suppliers arrange with households for automatic partial disconnect for short periods. How prevalent these will be is open to conjecture. As an insurance against inadequate capacity, there is a (supposedly temporary) provision for a “Reserve Trader”. This entails NEMMCO seeking additional supply offers (or demand side offers) in the event of it determining that a shortfall will occur in forward supply. Such a shortfall would be determined from forecasts of maximum capacity 2 years ahead that generators are required to provide, forecasts that are married to NEMMCO’s own demand forecasts. Under current standards, which are a combination of N-1 and unserved energy standards, NEMMCO is required to ensure 850 MW of reserve is carried across the entire NEM – including during the periods of extreme demand – to provide the required level of supply reliability. Though the Reserve Trader provisions have in fact been used, they have an intrinsic deficiency in that they simply call forth supplies that are available in any event. No government body would contemplate building additional capacity solely for reserve purposes. Reserve Trader capacity, therefore simply adds to costs unless the market manager has superior prescience to the various market participants. This is because, absent the government intervention, to the degree that supply shortfalls emerge, prices will be driven up and suppliers have every incentive to make plant available that was formerly mothballed, due for scrapping or being restrained to economize on maintenance. Moreover, a Reserve Trader policy carries seeds that can distort the entire market. If the Reserve Trader is in place and is offering higher prices than those anticipated by market participants, a rational generator will withdraw capacity thus exacerbating the apparent future shortfall. If such actions were to snowball, there would be a progressive increase in the apparent shortfall and an increasing need for the authorities to contract outside of the normal market, thus undermining it.

6.3.2. Market integration: access to transmission Although Australia does not have a nodal pricing system, there are procedures in place that in principal, lead to new regions being created when areas are islanded with significant price separation for more than 40 hours/year. However, political pressure has prevented new regions being created, notably in Queensland where the entire state remains a region in spite of considerable transmission weaknesses between the populous south east of the state and the northern parts. Loss factors for transmission and capacity constraints mean that there may be significant differences in the spot price for any trading interval across NEM regions, though market separation rarely occurs for more than an hour or so. Once a region is created, inter-regional settlements are auctioned. These settlements represent the difference between the value of electricity in the region where it is generated and its value if sold in another region. The settlement residue that accumulates is made available to the market by the conduct of an auction. Holders of this auctioned revenue value have a form of hedge that contributes to facilitating inter-regional trade by providing market participants a risk management mechanism as protection against high prices. The risk is however not eliminated since if the link constrains, revenues are not able to benefit fully from high prices in one region because of volume shortfalls.

188


6.3.3. Issues concerning generator market power Market supplies in Australia have on the whole been sufficiently disaggregated to prevent monopolistic power to be exercised other than on a transitory basis. In the interconnected NEM of six regions, the largest generation business has only 14% of the market and two other businesses each have over 10%. Concentration is somewhat greater in individual regions and these sometimes have constrained interconnects. Three businesses have 10% or more of the national market’s capacity (one of which, Snowy, is a hydro generator with the maximum annual output equal to about one quarter of capacity). A further 11 businesses each have 2–8% of the capacity market while a host of smaller suppliers collectively account for about 10% of the market (Fig. 6.5). In some states the capacity is more concentrated. Thus, NSW has three baseload generators (all government owned) controlling 70% of capacity and it is claimed that they have used market power on occasion to ensure that the mandatory insurance system for retail prices created by government regulation (and discussed in Section 6.7) does not accumulate income. NSW market shares are shown in Figure 6.6. In addition to this level of market supply, there are two transmission links from Queensland, theoretically rated at a combined 880 MW.

St IP an w e O ll pt im Ec a og en Fl in de rs O th er

D el ta Sn ar o in w g y en Ya er llo gy ur n (C C S LP) en e Lo rgy y ya n Ta g ro ng N R G Er

M

ac q

ua rie

%

Major generators’ market share 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0


Production capacity: NSW


nt

ry

L ED

ou C

nk

e

ba ed

L AG

g

ta

ar in Er

D

el

y w Sn o

Si th

R

M

ac q ge ua n ri

e

5000 4000 3000 2000 1000 0

189


In Victoria and South Australia, which have strong transmission capacity links with each other, the market supply includes that which can be transferred through Snowy either from the hydro facility itself or from NSW power suppliers. As from early 2006 a further transmission connection from Tasmania will operate with a capacity of 300 MW into Victoria and 630 MW into Tasmania. Even though CLP has over 30% of the capacity, there are several other powerful suppliers and there are few concerns about abuse of market power (Fig. 6.7). The fact that there is so much flexibility in bidding under the Australian system has contributed to unease in certain quarters that price oscillations have been due to supplier market manipulation. Of course, such price oscillations have always been present – they are inevitable because of the wide and sometimes rapid variations of demand (and occasionally supply). They were masked in the past because the monopoly provider called in higher merit order plant to cover peaks or sudden needs without this actually being specifically priced. The sharp oscillations in price now visible have led to claims that the generators are unfairly “gaming” the market, driving up prices to take advantage of opportunities in which there is a monopoly. A lengthy series of investigations into the structure of the bidding rules was underway between 1999 and 2002. Similar debates were held in the UK and other countries. In the event there was recognition that temporary market power is common to many industries and that attempts to combat it by fixing or constraining prices could exacerbate the underlying conditions that create it. Thus an ability to earn very high prices from being able to react quickly to opportunities or provide capacity in an area that can become islanded and subject to sudden price surges/supply shortfalls tends to encourage desirable investment behavior. Requiring a window of several hours between bid and dispatch would frustrate this. By the same token, preventing firms from reacting quickly to cover their contracts in the event of an unexpected outage is likely to result in over-cautious holding back on capacity and higher costs. The abandonment of proposals to place restraints on re-bidding close to dispatch was influenced by observations that most of the very late changes to price in the market had been price reductions. These reflected decisions by marginal suppliers to bring plant on line in response to market opportunities that were unfolding (due to a demand surge, network constraint, etc.). Following the various investigations into market power, changes to the Code were made to recognize the possibility that a supplier could bid erratically and benefit from creating great instability to the market.16 Hence rules were tightened to require some explanation of Market supply in Victoria/South Australia 5000 4000 3000 2000 1000 s O

th

er

n er

he

ge

rn Sy n

ut So

N

R

G

y ow Sn

ng Ya

IP

y Lo

Ya (C llou LP rn )

0

Fig. 6.7. Source: NEMMCO. 16

This led to minor modifications to the guidelines on re-bidding volumes close to dispatch. See http://www.aer.gov.au/content/index.phtml/itemId/659216

190


re-bidding behavior. These changes were more of the nature of increased insurance against aberrant bidding behavior rather than affecting bidding patterns themselves. 6.3.4. Price outcomes in the wholesale market The resiling from major changes to the market rules was made more acceptable in view of the very low prices that had on average prevailed in the market. Compared to a notional wholesale price pre-market of around $A38/MWh in Victoria and $A40/MWh in NSW, competition has maintained average prices at low levels, below $A35 for most of the period. Higher prices prevailed early in the Queensland and South Australian regions as they were being bedded down. Table 6.3 illustrates the price developments in each of the regions. In addition, the forward price has tended to be stable, though edging up beyond the $A40 level by the end of the current decade, indicating a need, though not a pressing one, for new plant since at such prices new coal-based plant investment is profitable in Queensland, NSW, and Victoria. A synthetic forward price is published with the following indicative price for baseload energy (Fig. 6.8). Table 6.3. Average price ($A/MWh). Year

NSW

QLD

SA

SNOWY

1998–1999 1999–2000 2000–2001 2001–2002 2002–2003 2003–2004 2004–2005 2005–2006 (2 months)

33.13 28.27 37.69 34.76 32.91 32.37 39.33 26.68

51.65 44.11 41.33 35.34 37.79 28.18 28.96 19.48

156.02 59.27 56.39 31.61 30.11 34.86 36.07 32.36

32.34 27.96 37.06 31.59 29.83 30.80 34.05 27.33

TAS

VIC

190.38 110.63

36.33 26.35 44.57 30.97 27.56 25.38 27.62 28.61

Source: NEMMCO.

Regional quarterly base futures prices 80

$/MWh

60 40

Queensland

New South Wales

Fig. 6.8. Forward baseload price. Source: NECA.

Victoria

Q4 09

Q3 09

Q2 09

Q1 09

Q4 08

Q3 08

Q2 08

Q1 08

Q4 07

Q3 07

Q2 07

Q1 07

Q4 06

Q3 06

Q2 06

0

Q1 06

20

South Australia

191

The Electricity Industry in Australia 6.4. Performance of the Reformed Electricity Market 6.4.1. Generation

The outcome of Australia’s reforms has been considerable improvements in productivity. In the case of power stations, increases in labor productivity since 1990 have ranged from a fivefold improvement in Victoria to a 50% improvement in Queensland, where it is generally acknowledged that power stations were operated more efficiently at the outset. Although there has been some new construction, the stock of generating capacity has not markedly changed over the period. Figure 6.9 illustrates the improved productivity.

Generator labor productivity (GWh/employee)

60 50 40

1990/1991 1996/1997 1999/2000 2000/2001 2002/2003

30 20 10 0

New South Wales

Victoria Queensland

South Australia

Tasmania

Western Australia (Western Power)

Power stations’ availability to run 100 95 90

1990/1991 1996/1997 1999/2000 2000/2001 2001/2002 2002/2003

85 80 75 70

New South Wales

Victoria

(b) Fig. 6.9. Source: ESAA.

Queensland

South Australia

Tasmania


192


In spite of these vast improvements in labor productivity, Australia’s power stations have also shown much greater reliability. As illustrated in Figure 6.9, the “Availability to Run” has improved in all of the States but especially in NSW and Victoria. This in itself has raised the de facto capacity of the industry. The comparisons indicate a relatively better performance on the part of those states that have privatized their businesses. In the privatized Victorian and South Australian systems, labor productivity has respectively increased fourfold and threefold. NSW with its government owned generation business and Queensland, under predominantly government ownership, have seen more modest increases – twofold in the case of NSW. In terms of labor productivity and plant availability, the privatized Victorian system now surpasses the performance of the state owned NSW system. This is in spite of Victoria having the disadvantage of relying predominantly on brown coal, which requires greater processing before being burned than black coal, which is the fuel source of the NSW system. Part, but by no means all of the relative Victorian improvement is due to greater use of contractors in the privatized firms. Moreover, to the extent that NSW uses fewer contractors this is likely to be a reflection of its shareholder’s preferences for union labor and would contribute, of itself, to lower levels of efficiency. There is also some indication of improved capital productivity in the privatized businesses. In particular, privatization brought a new lease of life to International Power’s 1600 MW Hazelwood brown coal generator in Victoria. The station had been previously scheduled to close in 2005 but has had its capacity increased and now likely to operate for another 20 years.

6.4.2. Distribution As with generation, distribution, which accounts for over 40% of final costs, has shown strong productivity improvements. Figure 6.10 shows that customers per employee in

Distribution businesses: customers per employee 900 800 700

1994/1995 2002/2003

600 500 400 300 200 100 0 New South Wales


Victoria

Queensland South Australia

Tasmania



193

Victoria and NSW have almost doubled and all states, bar Queensland, have shown impressive gains. Again Victoria shows impressive labor productivity (the Western Australian system applies only to the interconnected, relatively urbanized system). In terms of outages the system performance has been mixed – in most areas outages are heavily influenced by occasional severe storms that do not occur with any degree of regularity. Figure 6.11 below indicates that outages have generally remained low and shown little trend, except in Victoria where they have been reduced and become comparable with outages in other states.

6.4.3. Consumer price outcomes Price level comparisons are distorted by the previous, and for households to a major degree on-going, regulation of maximum prices. Household retail price controls have suppressed prices in all states, though in recent years price controls have been eased or been allowed to rise in response to regulated increases permitted in line charges. This is especially the case in South Australia, where electricity costs are intrinsically higher than in other states and where the load profile is more skewed towards the summer peak than in other states. Household consumers’ prices in NSW and Queensland remain under relatively tight government control and are below those that would prevail in a commercial market. Over the past 7 years real prices for residential consumers have risen by as little as 9% in NSW and as much as 52% in South Australia. Those in Victoria have risen by 15%. As expected from a system with access to low-cost energy inputs, Australian prices remain among the lowest in the world. Business customers have been freed from price controls for several years. In the past, the business customers subsidized household customers but once the market became contestable this was no longer possible. As a result, real prices for business customers have fallen by over 23% in the three eastern states; they have risen about 2% in South Australia. Figure 6.12 below illustrates the trends.

Average outage duration (minutes off supply per customer) 350 300

1995/1996 1996/1997 1997/1998 1998/1999 1999/2000 2000/2001 2001/2002 2002/2003

250 200 150 100 50 0 New South Wales

Victoria


Queensland

South Australia

Tasmania


194


1996/1997 1996/1998 1996/1999 1999/2000 2000/2001 2001/2002 2002/2003 2003/2004 re NS si W de nt ia l N SW bu si lar ne g ss e re Vic si to de ria nt ia l Vi bu c l si arg ne e ss Q ue re e si ns de la nt nd ia l Q bu ld l si arg ne e ss S. re Au si st de ra nt lia ia l SA bu la si rg ne e ss

Cents/KWh

Real electricity prices (2001/2002 dollars) 20 18 16 14 12 10 8 6 4 2 0


6.5. Issues Concerning Public and Private Ownership The post 1990 period in Australia has seen dramatic improvements in the productivity of the ESI. Initially these strides were made under public ownership with reform of the overstaffed integrated utilities and their incorporation under company law. A further productivity improvement took place with the privatization of the Victorian industry from 1997. State owned businesses generally adopted many of the labor saving and production enhancing measures of their private sector counterparts. However the fact that the Australian industry remains mostly under public ownership carries several potentially damaging consequences. The first of these is the intrinsically greater incentives to save on costs that are present in privately owned businesses. Part of this may be due to government as the shareholder and appointer of the company boards, is more reluctant than a private company to shed labor; allied to this is close links that the Australian State Governments owners of the electricity businesses have with trade unions. This means they are reluctant to allow non-union labor and keen to ensure that union rights and privileges are maintained. In the main this will reduce the capacity of the managements to manage. The evidence available in Figures 6.9–6.11 indicates that the efficiency levels of the private sector businesses exceed those of the public sector firms. Public ownership also impedes firms from re-arranging their assets. All of the private sector businesses post-privatization have undergone several structural changes as parts of them have been spun off, other parts have been augmented and some have undertaken major new investments. Private sector businesses, in search of operational economies and improved shareholder value have re-arranged asset ownership so that activities are better grouped together. In the case of the distributor/retailers, this has led most of the private sector to have the two functions separated and housed in differently owned firms, whilst none of the public sector businesses have taken such steps. The public sector businesses need to approach their government shareholder for approval of any new capital investment. These decisions face the familiar issues of government decision taking. In New South Wales for example, the State Government has adopted an anticoal philosophy on greenhouse grounds (in June 2005 the “ecologically friendly” then State


195

Premier even opened up the debate on nuclear generation, a debate that had been dormant due to opposition by green groups and because cheap coal offers a more economic baseload solution). By contrast to the NSW Government’s restraint on new investment, it is claimed that the Queensland State Government is accepting a lower than commercial rate of return in order to encourage the building of new power stations. It is certainly the case that most new capacity has been built in Queensland and about 60% of this has been government financed but Queensland also has the fastest growing load and has the cheapest coal. In this respect, data for Queensland (Table 6.3) shows spot prices were one third less than in NSW in the early part of 2005/2006. This may offer some corroboration of the over-build of generation in the State; however, the spot market data does not appear to be reflected in higher contract prices (Fig. 6.7). Whether or not Queensland new generation investment has been fully justifiable on profit grounds, public ownership adds a non-commercial dimension into the industry which diminishes the predictability of private firms’ competitive environment. Other things being equal, this brings additional business risk, excessive conservatism and higher prices/lower reliability.

6.6. Government Market Interventions 6.6.1. Mandatory insurance schemes Public ownership leaves governments with greater political vulnerability in the event of poor decision making. Hence decisions by both the NSW and the Queensland Governments to implement a form of mandatory insurance for the supplies to household retail customers. This, called the Electricity Tariff Equalization Fund (ETEF) in NSW, tends to blunten the market forces through reducing apparent retail risk by having the government assume much of it. One outcome is price suppression, especially for peaks, and a muted demand signal for new investment. ETEF operates by placing a ceiling and floor on wholesale prices as they impact on the household part of the aggregate load. When prices are high the generators receive only the stipulated price and reserves are accumulated in the fund; these are released when spot prices are below the floor set for the fund. As well as creating an insurance risk for the state, it is likely that such measures also impede the market for various financial instruments, the depth of which in NSW lags considerably behind Victoria. Although having similar features, the Queensland scheme, called the Long-term energy procurement or (LEP), has been designed with the aim of increasing liquidity and market depth by encouraging the incumbent retailers to actively seek contract cover for the franchise load. That is, the structure of the LEP is such that the incumbent retailers are exposed to energy price and volume risk and have incentives to manage their position through contracting, with compensation provided for efficient purchasing against a benchmark.

6.6.2. Retail price control Retail prices to business consumers have been deregulated in all states but all retain some controls over household prices. In the case of Victoria and South Australia, these controls are not considered to be significantly in excess of underlying market prices, and in both states there has been significant customer churn. This is also true to some degree of NSW and to a lesser extent in Queensland.

196


For electricity, retailers need to balance their demand and supplies and act as the agent of the final consumer. They pick up the tab when the price spirals out of control and this gives them a great incentive to ensure they measure their sales correctly and contract for supplies appropriately. They perform the same essential function in the energy market as elsewhere – they look to demand, seek to attract customers who they can profitably supply and package supply to meet their customers’ needs. The outcome is signals that drive efficient market activity. These signals include the prices that attract the right form of new supply (peak, off peak, etc.) They also develop prices that choke off or encourage increased demand. Of course, all this is made more difficult in the ESI. The absence of half hourly metering at the domestic level and the price cap are among the market realities that prevent this from operating with full effect. Even so, the residual retail price controls dampen the economic signals that businesses need to determine the optimum time to invest or undertake other strategic decisions. 6.6.3. Greenhouse gas abatement schemes Australian coal is inexpensive and located conveniently to major electricity loads. However, its economics change markedly in a Kyoto constrained electric power future, whether the measures in place to reduce carbon dioxide emissions are a form of cap and trade regulation, a carbon tax or even a more arbitrary set of regulations that have similar effects. Australian measures ostensibly aimed at reducing emissions of carbon dioxide appear in many forms and are inconsistent from one State to another. They impact most heavily on coal, Australia’s lowest cost energy source and include: ● ● ● ●

●

the federal government’s mandatory renewable energy target (MRET), the Queensland’s 13% gas target, the NSW’s Greenhouse Gas Abatement Certificate (NGAC) scheme, subsidies to wind and other exotic renewable sources offered through the Australian Greenhouse Office and state governments (the latter in the form of regulatory measures that reduce connection costs to wind generation), and schemes that mandate minimum energy savings on appliances. First applied to fridges and freezers and targeted at energy conservation, these regulatory requirements have been re-badged as greenhouse measures and extended to include houses as well as other appliances.

The MRET scheme’s focus is on renewable energy and requires retailers to acquire and annually surrender a progressively increased number of Renewable Energy Certificates (RECs). These essentially require usage of novel energy sources like wind, though some existing and expanded hydro capacity has managed to redefine itself to be eligible. By 2010, 9500 GWh (around 4.5% of demand) will be required under the MRET scheme. For the three schemes combined over 30,000 GWh is estimated to be covered amounting to over 13% of total demand.17 The Queensland scheme seeks to substitute gas for coal-based electricity inputs. The NSW scheme seeks to introduce a penalty on carbon dioxide graduated in line with the emissions per unit of energy of each electricity generation source. 17

Alan Moran. November (2004) Economic and Environmental Potential of Energy Efficiency Regulations: Submission to the Productivity Commission Inquiry into Energy Efficiency, at http://www.ipa.org.au/files/ Energy33.pdf

197


The default penalty costs of the three regulatory measures provide a cap on the costs they are likely to entail. These costs entail a premium over the costs of conventional electricity to retailers. By 2010, when the schemes are at full maturity, the fall-back penalty rates for the Commonwealth, NSW and Queensland schemes respectively are $A40, $A14.3 and $A13.1/MWh. These rates provide the (maximum) subsidies to the non-carbon or low-carbon emitting fuels. The Commonwealth’s RECs during 2005 were trading 20% below the maximum rate. Table 6.4 below summarizes the more readily identified costs. A greenhouse trading regime has considerable support in Australia, with most state governments urging its national adoption. Figure 6.13 below estimates the costs of electricity with and without the sort of additional charges implicit if Australia adopted the EU cap and trade scheme and other schemes are left in place. Based on recent developments, we have relatively good information on the costs of conventional generation for the eastern seaboard of Australia. Nuclear costs have been rigorously Table 6.4. 2010 Costs of greenhouse gas support measures.

$AM

MRET

NSW NGAC

Qld 13% gas

Commonwealth subsidies

State subsidies

380

222

68

124 (2006/2007)

32 (2004/2005)

Source: Budget documents.

Comparative costs of power $90 $80 $70 $60 Cost per MWh Cost per MWh post rights

$50 $40 $30 $20 $10

Si uc m le ar N uc Sc ul m lea ly od r ul SA ar IC he G liu T m

d

N

in

G en

W N

uc

N

le ar

at

ur a

n

lg as

co al

ld Q ow Br

al co

co

Bl

ck Bl a

ac k

al

N

SW

$-

Assuming CO2 rights trade at $41/tonne Fig. 6.13. Australian power generation costs with EU tradable rights prices. Sources: Australian costs are based on recent estimates. Nuclear power costs are based on University of Chicago, The Economic Future of Nuclear Power. Carbon emissions per gigajoule of energy is derived from http://www.greenhouse.gov.au/workbook/pubs/workbook.pdf

198


evaluated in a recent University of Chicago report that brought together three contemporary estimates of the costs of nuclear generation (excluding the disposal costs). The models compared are the Shanghai automotive industry corporation (SAIC) industry model, the Scully Capital financial model, an Energy Information Administration (EIA) model and GenSim, which is based on the EIA approach.18 Wind power costs are also relatively well known. The price cited below ($A75/MWh) represents the costs at prime sites. These costs exclude any additional transmission charges that may be required for the more remotely located facilities. They also do not take into account the need for conventional power back up which is necessary once wind, with its unpredictable and intermittent nature, becomes a significant component of the aggregate supply. At present, coal-based generation costs is estimated to vary from about $A32/MWh in Queensland,19 around $A40 in NSW and $A38 in Victoria. Gas is estimated at around $A45/MWh based on a cost of $A4/Gigajoule, a cost that may rise if greenhouse measures raise the demand for gas. Nuclear costs exclude waste storage, the estimated costs of which vary. One estimate by the Uranium Information Centre puts the decommissioning costs as adding 5% to the price and waste disposal a further 10%.20 Applying the EU price of euros 24/tonne of carbon dioxide, cost to current power sources would result in nuclear power based on some cost estimates becoming marginally cheaper than coal and natural gas generation. Government greenhouse mitigation policies and laws have also been used by advocacy groups (themselves often government financed) to tie up new proposals in the courts. The highest profile case, which is discussed below, has been the NSW State Government using greenhouse policies to maneuver itself out of a (high-priced) contract for a new generator designed to use waste coal. Other cases have included the prevarication of the Victorian Government in granting approval for a major power station to have its life extended. Snowy Hydro illustrates an outcome of subsidies that cannot or are not fully defined to meet their stated goals. Snowy has been given an annual baseline level of generation above which it earns RECs that have a default value of $A40/MWh. Because it does not receive penalties for underperforming, it operates its system to generate strongly in 1 year and build up its reserves for the next year. This allowed Snowy to earn about $A67 million in essentially phantom RECs in 2003. Not only did this not contribute in net terms to the government policy of reducing greenhouse gas emissions, it actually increased emission levels. This is because 20% of Snowy’s RECs may have been created through pumping water uphill for reuse. Though the energy used in pumping is netted out from the REC creation, shifting water from an underperforming year to a hard generating year creates a credit. In pumping water uphill, Snowy uses almost twice as much (coal-derived) energy as it produces in subsequent generation. 6.6.4. Sovereign risk One further facet that is not considered in the comparative cost data is the sovereign risk involved in building fossil fuel power stations, especially involving coal. The activities of 18

See http://www.anl.gov/Special_Reports/NuclEconAug04.pdf Some private generators suggest this price, which is cited as being available from new government generators, is sub-economic price because it incorporates a de facto subsidy as a result of the Queensland government accepting lower levels of return than the private sector would require. 20 Source Uranium Information Centre http://www.uic.com.au/nip08.htm 19


199

the NSW Government in using environmental pretexts to renege on coal-based contracts and the additional costs the Victorian Governments have required of the Hazelwood Power station upgrade proposals are likely to require a risk premium for coal powered electricity. The NSW Government has created obstacles and uncertainties in the way of new electricity generation. These make it inevitable that investors will require a risk premium before committing funds, thus increasing the price of electricity in the state and the risk of shortages. One example of this was the government’s treatment of a private sector investment undertaken by the US firm National Power. This stemmed from a commitment – in the event an unwise commitment – by energy Australia, the biggest retailer in the NSW for two power stations, Redbank 1 and 2. Soon after the deal was struck, the price in the market halved and remains 30% below the Redbank contract price. Some estimates put the contract loss at $A750 million. To renege on the deal for the second power station, the NSW Government set up an inquiry into it. Various Government funded green groups offered opposition to the project on grounds of its greenhouse gas emissions and the Government refused its development approval, thus avoiding an onerous contract. In fact opposition to the development on environmental grounds is ironical since the project uses waste coal which could otherwise pollute the Hunter River. Indeed, in 2001 Redbank 1 won the Institution of Engineers Award for Environmental Excellence. The Government’s performance on this matter must add to the risk premium required of private sector developers of power stations and will probably require some enforceable undertakings before any private funding is extended to coal-based generation in the state. In addition, private sector investors, especially in coal-based generation, would hardly be re-assured by the statements the NSW Government makes in a Green Paper issued in January 2005.21 This expresses considerable hostility to new coal power, canvasses greenhouse taxes and a policy (p. 24) by mid century ranging from stabilization of emission at current levels to reducing them by 40%. In terms of a business-as-usual growth in energy demand at 2%/annum, this range of outcomes would amount to a highly ambitious reduction of between 60% and 75% in emission levels. In claiming in concert with this, that the Government will let the market decide which technologies should be developed, the Green Paper is giving expression to grand sounding laissez-faire principles that clothe a highly intrusive policy approach. The Green Paper, due to proceed to a White Paper stage by June 2005, had not so progressed by October 2005. In the interim, the State Premier, Mr. Carr who was highly supportive of greenhouse gas reducing initiatives, unexpectedly retired.

6.7. Outcome of Australian Electricity Market Arrangements 6.7.1. Summary of market distortions Aside from sovereign risk associated with government de facto expropriations as occurred in the NSW Redbank case, distortions that could lead to serious market damage include: ●

21

The NSW mandatory insurance system or ETEF, provides a weaker incentive for retailers to ensure that they are forecasting market demand accurately. ETEF means the

NSW Government, Energy Directions Green Paper. http://www.deus.nsw.gov.au/new/NSW% 20Energy%20Directions%20Statement%20-%20702 KB.pdf

200

●

●

●

●

●

●


government has eliminated the risks to retailers of failing to forecast the household load accurately. This may bring mistakes caused by unexpected demand shifts. Retail price caps being kept below market levels. This is an area where governments in Victoria and SA have managed to control their propensities to intervene and are allowing prices to shift market levels. NSW however retains very low allowable retail margins, which seriously restrict competition. The risk that regulators will offer inadequate incentives for expansions and optimal maintenance. Major price reductions have been insisted upon by several state regulators. Regulated businesses are always likely to profess dissatisfaction at such outcomes but a risk remains that price cuts can deter investment. Such requirements may have been a feature of the fragility of distribution networks, especially in Queensland where the regulator demanded a 17% cost saving of the largest distributor, Energex, which was heavily criticized following power outages in 2005. Interventions favoring subsidized and uneconomic generation can suppress demand, which means reduced new investment especially in the sort of energy intensive industries that Australia is well placed to win. The various schemes like MRET and NGAC add costs to industry and in the case of NSW, mean that some 23% of electricity is now slated to be subsidized; this probably rules the state out of consideration for major new energy intensive industry. Less draconian measures are in place in other states – Queensland has its 13% gas requirement and Victoria was less than firm in controlling its state financed green groups who campaigned to prevent the Hazelwood expansion; nor has the state made a wise choice in its appointment of a relatively activist Presiding Judge to the Land and Environment Court whose legal interpretations prolonged the case and added expenses. Some private sector generation businesses claim that the new capacity building by the Queensland government is not based on commercial principles but are being subsidized indirectly by a government intent on using its cheap coal as an industry development tool. Though subsidized plant adds to capacity in the first instance, each new tranche of it considerably reduces the incentive of commercial parties to seek out opportunities to build plant in line with market requirements. Subsidized plant puts us on the slippery slide to total government ownership or control of the industry. Much the same risk has in the past been offered by subsidized transmission. If a power station is stranded by low-cost power being brought in from elsewhere it suffers lower than expected returns. If this is due to it being stranded as a result of government regulations that effectively subsidize costs, damage is done to the market’s automatic ability to supply demand. Finally, there remains the risk of other interventions. Australia has not moved to create trading barriers in response to fears about market power being exercised in re-bidding. Such measures would create major uncertainties by constraining generators’ abilities to respond to sudden emergency issues, would have gummed up the bidding system and created costs and uncertainties.

6.7.2. Outcomes in terms of new capacity and prices Notwithstanding the adverse effects of government intervention in Australia, a successful outcome has been observed. Prices are lower than expected, reliability has been high, and although the most market exposed sector, generation, has seen very low returns, new investment has been forthcoming. And the investment that has been made broadly corresponds

201


to that which most experts expected: increased peaking capacity in Victoria and South Australia and more baseload to meet the faster growing Queensland demand. This demonstrates a resilience in markets. As long as the various participants in the market are free to contract with each other and as long as there is no significant monopoly over supply, interventions may not seriously distort the market and lead to its failure. Aside from wind power, which is totally dependent on subsidies, significant new power facilities built over the past 5 years are as shown in Table 6.5. Figure 6.14 illustrates how demand and supply have been fairly well synchronized over the past 6 or 7 years. The message is that the real dangers to the supply industry in both gas and electricity in Australia are those stemming not from too little government but from too much. The

Table 6.5. New capacity 2000–2005.

Redbank Bairnsdale ValleyPower Somerton Laverton Loy Yang Oakey Millmerran Swanbank E Tarong N Kogan creek Hallett Pelican point Ladbroke Quarantine

State

Capacity (MW)

Type

Ownership

NSW Vic Vic Vic Vic Vic Qld Qld Qld Qld Qld SA SA SA SA

150 92 300 160 312 236 282 852 360 450 750 220 320 80 100

Coal Gas Gas Gas Gas Coal Gas Coal Gas Coal Coal Gas Gas Gas Gas

Private Private Private Private Government Private Private Government/Private Government Government Government Private Private Private Private

Source: ESAA.

NEM generation and peak demand 45000 40000 35000 30000 NEM generation

25000 20000

Summer peak

15000 10000 5000 0 1998

1999


2000

2001

2002

2003

2004

202


industry has expanded and maintained low costs in the 6 years it has been operating. It is however fragile and government actions could seriously harm investor confidence and lead to ever increasing interventions to ensure investment keeps pace with demand. Such measures would gradually erode the massive lifts in productivity that has been observed over the past decade or so.

6.8. Glossary of Acronyms ACCC AEMC AER ESAA ESC ESI ETEF LEP MRET NCP NECA NEM NEMMCO NGAC ORR PC QEC REC SECV

Australian Competition and Consumer Commission Australian Energy Market Commission Australian Energy Regulator Electricity Supply Association of Australia (Victorian) Essential Services Commission Electricity Supply Industry Electricity tariff equalization fund (Queensland) Long-term energy procurement Mandatory renewable energy target National competition policy National electricity code administrator National electricity market National electricity market management company (NSW) Greenhouse Gas Abatement Certificate (Victorian) Office of the Regulator General Productivity Commission Queensland Electricity Commission Renewable Energy Certificate State Electricity Commission of Victoria

Chapter 7 Restructuring the New Zealand Electricity Sector 1984–2005 GEOFF BERTRAM School of Economics and Finance, Victoria University of Wellington, Wellington, New Zealand

Summary This chapter describes New Zealand’s failure over two decades of reform to establish a viable industry self-governance framework, and the parallel failure to achieve restraint on monopoly profits by means of light-handed regulation. Starting from a classic publicly owned monopoly of generation, transmission, distribution, and retailing, New Zealand corporatized all levels of the supply chain, separated lines businesses from generation and retail, removed retail franchises, and broke up the monopoly generator into five companies, two of them privately owned. These measures were insufficient to achieve competitive outcomes in the absence of hands-on regulation. Generators integrated vertically by takeover of retailers, and the resulting retail oligopoly erected an effective barrier to entry by withholding affiliated generators’ capacity from the very thin market for hedge contacts. Grid pricing and contract provisions foreclosed demand-side innovation and distributed generation. Distribution lines businesses ramped up mark-ups from 30% to 70% without any regulatory restraint, and were allowed to revalue their assets to underwrite the new high margins. Faced with failure of the original design, the Government in 2003 established a new industry regulator and invested in new state-owned thermal generation to plug the country’s yawning gap in reserve capacity.

7.1. Background New Zealand is a country of 4-million people spread across an area the size of Italy or the UK. From south to north the country is over 2000 kilometers in length, with the two main islands separated by the 30-kilometer-wide Cook Strait. The largest city (and major electricity load center) is Auckland, with 1.3 million inhabitants. Electrification began in the late 19th century, when local authorities and private entrepreneurs constructed small generation facilities to serve local markets.1 Following the First World War the Government embarked on the construction of a set of large state-owned 1

A detailed history of the New Zealand electricity industry is Martin (1998). See also Rennie (1988), Jackson (1988, 1990).

203

204


hydroelectric plants on major rivers, linked by a transmission grid from which power was taken off by local-government distribution and retail companies (Electrical Supply Authorities, ESAs), each with a territorial monopoly franchise. ESAs supplied a bundled service, comprising low-voltage distribution networks, the retailing of electricity to final customers, and supply and servicing of household electrical appliances. In the 1950s, when major new investments in generation plant struggled to keep pace with demand growth and blackouts were a common occurrence, most households were placed on ripple control to switch off water heaters at times of peak demand.2 The state-owned generation and transmission system built up from the 1920s displaced most locally owned generating plant, and standardized the countrywide retail supply voltage at 220/240 volts at a frequency of 50 MHz (matching the UK settings). For the next halfcentury, electricity generation and transmission remained a state-owned monopoly, while distribution and retail remained franchised, publicly owned, local monopolies. Regulation in this setting was redundant, since both central and local government were democratically accountable, and operated the electricity supply system with social, rather than commercial, goals. Prices were set to achieve break even, in cash flow terms, over the long run. Financial disclosure, in terms of the cash flow model used for much of the public sector, was comprehensive, with detailed accounts for all levels of the system published annually.3 Asset values were recorded in historic-cost terms without adjustment for inflation, and were also lowered by the common practice of expensing day-to-day small-scale acquisition of capital equipment. Initially the two main islands had separate electricity grids, but there was an obvious mismatch between the abundant hydro resources of the South Island and the concentration of load in the North Island, particularly in Auckland. In 1965 a high-voltage direct current (HVDC) cable across Cook Strait connected the two systems together, allowing power from large hydroelectric developments in the South Island – particularly Benmore (540 MW) and Roxburgh (320 MW), on the Waitaki and Clutha Rivers, respectively – to be sent north. Thereafter the entire national generation and transmission system developed as a single integrated whole. The North Island accounts for around two-thirds of national demand but only one-third of generating capacity; the South Island has two-thirds of generation capacity but only one-third of demand.4 New Zealand’s annual electricity consumption is currently around 36,000 GWh, supplied from a system with 8500 MW of installed capacity. The 50% capacity utilization ratio reflects

2

Ironically, this almost universal penetration of simple demand-management technology in the period of public-sector monopoly has been allowed to slide away in the era of “market reforms” since 1987, as large commercially oriented firms on the supply side have welcomed demand-driven price spikes which they could take directly to their bottom lines. 3 The Minister in charge of the New Zealand Electricity Department (NZED) tabled a full annual report in Parliament each year. All ESA financial and operational data was published annually from the early 1960s under the cumbersome title Annual Statistics in Relation to Electric Power Development and Operation for the Year Ended 31 March. The latter publication rapidly reduced its coverage in the early 1990s and was discontinued in 1994. Its successor, the company-by-company regulatory information disclosure from 1994 on, was both less informative and inconsistent from company to company, which means that public monitoring of performance has been more difficult after the reforms than before. 4 The mismatch between the two islands would have been greater still had it not been for the establishment in the 1960s of the large Comalco aluminium smelter at Bluff in the far south, which by itself comprises about 17% of national demand and provides the principal market for the Manapouri hydro scheme, the country’s largest with capacity of 710 MW (upgraded from 585 MW in 2002).

205

Restructuring of the New Zealand Electricity sector 1984–2005 Table 7.1. Trends 1965–2004.

1965 1970 1975 1980 1985 1990 1995 2000 2004

Total installed generating capacity (MW) 2336 3683 4784 5860 6988 7067 7910 8845 8515

Peak load (MW)

Total consumption (GWh)

2048 2690 3391 3677 4642 5122 5240 5830 6090

8189 11,069 16,272 19,040 23,994 27,309 29,925 32,735 35,795

Total sales revenue ($m)

Average final price (c/kWh)

Real average price, c/kWh at March 2004 prices

90.0 143.3 196.4 681.5 1190.4 2144.2 2490.2 2888.2 4014.5

1.10 1.29 1.21 3.58 4.96 7.85 8.32 8.82 11.22

14.56 13.79 8.41 12.58 9.58 10.64 10.23 10.36 11.85

Sources: Installed capacity from Annual Electricity Statistics and Energy Data File for years shown. Consumption, revenue, and prices from Energy Data File January 2005, p. 126 Table G.12, p. 134 Table I.1, and p. 135 Table I.2. Real average price 1965–1975 derived using CPI.

the fact that two-thirds of supply comes from hydro generators which are designed to run at a low load factor, combined with the existence at the margin of some high-cost thermal generating capacity which is operated for only part of the year. System-wide capacity utilization has risen steadily over recent decades, reaching 40% in the mid-1980s and approaching 50% in the mid-2000s. Table 7.1 sets out key statistics of capacity, consumption, revenue, and final price from 1965 to 2004. This period includes the last two decades of the old system, the “reform” years from 1986 to 1998, and recent experience with the restructured system. Figure 7.1 shows installed capacity and peak load since 1964. Capacity growth has proceeded in a stop-start fashion, attributable partly to the lumpiness of generation projects, partly to swings in policy, and partly to commercial decisions since corporatization. In the mid-1960s the momentum of the hydro construction program was at last outstripping demand growth after a decade of stress in the 1950s. System peak load in the mid-1960s was around 90% of installed capacity, but with hydro capacity expanding 8.5% per year until the mid-1970s, the ratio was brought down to below 70% by the late 1970s, and has remained around that level for the subsequent three decades. Peak load growth, which caused concern among power planners in the 1970s and 1980s, slowed down from the late 1980s; the central problem since 1990 has been maintaining supply in dry years. Figure 7.1 shows also a slackening in the pace of new construction following deregulation in the early 1990s, and the impact of the periodic decommissioning by the new owners of commercially unattractive dry-year-reserve thermal plant, which has left the system increasingly exposed to climatic fluctuations. The map of the main high-tension transmission grid in Figure 7.25 shows the location of the two main bottlenecks in the transmission system: the HVDC link from Benmore to Haywards, and the central North Island between Haywards and Otahuhu. For the purpose of understanding the basic economics of the network, the nodal spot prices at these three key measurement points suffice to put a price on the two key transmission constraints, which cause market segmentation into three main regions at times of stress (Videbeck, 2004). 5

For a detailed map of the entire grid showing all nodes, see http://www.transpower.co.nz/?id⫽4631

206


10,000 9000

Meremere (coal) closed

Marsden A (oil) closed

Otahuhu B and Taranaki CC (gas) commissioned

Clyde (hydro) commissioned

Stratford, Otahuhu A, and Whirinaki (gas/diesel) decommissioned Manapouri (hydro) expanded

8000 7000

MW

6000 5000 4000 3000 Capacity Peak load

2000 1000 1960

1965 1970

1975 1980

1985

1990 1995

2000 2005 2010

Fig. 7.1. Generating capacity and peak load, 1964–2004. Source: Data compiled from Annual Statistics in Relation to Electric Power Operation in New Zealand 1965–1993, and from Energy Data File for years 1995–2004.

Auckland Central North Island constraint

Otahuhu Huntly

New Plymouth Cook Strait constraint

Bunnythorpe Haywards substation

Major generation and substations Main load centres HVDC link

Wellington

Manapouri

Christchurch

Main high-voltage AC transmission lines Key constraints

Benmore Dunedin Comalco smelter

Fig. 7.2. Major transmission lines, showing location of the two main constraints.

Restructuring of the New Zealand Electricity sector 1984–2005

207

7.2. Supply/Demand Balance Three key features of the electricity supply industry (ESI) in New Zealand have to be borne in mind when considering options for restructuring: ● ●

●

Most generation (60–70%) is from renewable sources (hydro and geothermal). The hydro lakes are located mostly in steeply sloping river valleys and provide storage capacity for only a few weeks, which means that unusually dry climatic conditions quickly translate into reduced supply. Similarly, unusually high inflows of water must be utilized within quite a short time horizon, or else be spilled to waste. New Zealand is a stand-alone closed market, with no means of importing or exporting electricity. A supply shortage, therefore, results directly in demand rationing and/or price spikes, while excess potential supply can be neither stored beyond a short period, nor sold in external markets.

Prior to the restructuring, which began in the mid-1980s, New Zealand’s generation plants were operated on the basis of control procedures that equated the shadow value of stored water to the short-run marginal cost of thermal generation. So long as river flows were adequate, hydro plant could be operated as baseload, with thermal peaking plant utilized in periods when demand exceeded the supply available from optimal utilization of water. The usual roles of hydro and thermal generation were thus reversed. However, hydro also performed (and still performs) the very short-run task of frequency control, via the Maraetai II generating station on the Waikato River.6 Until the early 1990s the state-owned monopoly generator and grid operator, the New Zealand Electricity Division (NZED, later the Electricity Corporation of New Zealand, ECNZ), carried out this optimization exercise internally, and scheduled its various generation facilities to optimize the utilization of water by attaching a shadow value to hydro generation to reflect both foregone opportunities to utilize water in later periods, and planners’ judgments regarding future hydrological conditions. If lakes were full and high inflows were expected, hydro plant would be operated at capacity. If lake levels were low and a dry year was anticipated, water would be held back and more thermal plant brought online to fill the resulting gap in supply. Unchallenged control of a balanced portfolio of generation options enabled NZED to reap economies of scope as well as scale, because of its ability to internalize spillover externalities amongst various generation technologies. In particular, the explicit balancing of hydro and thermal generation options to maximize year-round operating efficiency of the system as a whole was the key to the ability of NZED to provide a very high level of security, and quality, of supply across the entire country, even in the face of climatic variability (mainly uncertainty about rainfall and, hence, river flows). NZED’s explicitly forward-looking scheduling and planning procedures took advantage of this heterogeneity of its generation assets to supply wholesale power at an average-cost price (the bulk supply tariff, BST), with operating surpluses from hydro generation used to cross-subsidize the high-operating-cost thermal firming plant. From 1957 on, the BST

6

The two generating stations attached to the Maraetai dam have a total capacity of 360 MW, well in excess of the capacity needed to utilize run-of-the-river flow. The second station (excess capacity) installed in 1971 was designed to provide frequency control for the national grid, and has metering and control equipment to detect and offset load fluctuations. See http://www.mightyriverpower.co.nz/ Generation/AboutUs/HydroStations/Maraetai/Default.aspx

208


included a levy on consumers to fund new investment in generation and transmission as well as covering operating costs of the system. This cash-in-advance approach meant that whenever a major new round of investment was undertaken, the BST would be raised to provide the necessary funds in advance. Consumers were immediately conscious of the resulting rise in retail charges, which meant that electricity investment was always politically sensitive. The managers of the system were motivated both by the quest for engineering efficiency, and by this political sensitivity, since NZED and its controlling Minister would carry the political blame for any supply outages. There were strong incentives to invest ahead of demand,7 keeping a substantial safety margin in both generation and transmission; but there was a countervailing possibility of political backlash if excessive investment programs drove up the BST, and hence the price to consumers, unduly. In the late 1970s and early 1980s the system’s planners maintained a wide margin of excess capacity and embarked on a major round of large hydro construction, which exposed NZED to criticism that it was over-investing relative to a socially optimal benchmark. Such criticism was particularly acute in the mid-1980s as it became apparent that the momentum of the ongoing hydro construction program had carried NZED into a series of large hydro projects (Tongariro, Rangipo, and Clyde) whose unit costs were orders of magnitude higher than the BST. As it became generally accepted that the long-run marginal cost (LRMC) of new generation had risen sharply relative to the average cost of supply, a noisy debate ensued between advocates of immediate increases in the wholesale electricity price to signal future costs, and supporters of continuing with the long-established average-cost pricing approach. This pricing debate is discussed further below.

7.3. Restructuring the Sector 7.3.1. First steps There was a sea change in New Zealand economic policy in the mid-1980s, as neoliberal economic doctrines (largely copied from the UK) were adopted by key ministers in the Labor Government elected in mid-1984, resulting in radical changes to all state-owned operations, including electricity. Initially the aim was to ensure that state-owned monopolies increased their profitability by raising their prices to contribute to reducing the government’s budget deficit (Ministry of Energy Financial Objectives and Pricing Review Team, 1984). A second goal, initially also motivated by revenue maximization rather than structural reform, was to raise the economic efficiency of state-owned operations by converting them into profit-oriented commercial corporate organizations. Linked to this was a desire to curb what were perceived by the New Zealand Treasury at the time as excessive investments in new capacity, which officials regarded as a drain on scarce government resources. In 1986 the Labor Government announced its decision to reform publicly owned trading activities, including the generation and transmission sectors of the electricity industry,8 and a State-Owned Enterprises Act was passed to govern the process of corporatization.

7

As Chapter 1 notes, many countries have had difficulty with investment incentives in the new restructured environment. 8 For a detailed official history of the reforms summarized here, see Chronology of the New Zealand Electricity Reform, at http://www.med.govt.nz/ers/electric/chronology/index.html


209

In April 1987, the NZED was converted into the ECNZ and a private-sector entrepreneur was recruited to head the new board. The following year the operation of the transmission grid was transferred to a new ECNZ subsidiary, Transpower Ltd, as a first step toward separation of generation from transmission. The expectation of key policy-makers was that the generation assets of ECNZ would in due course be privatized, while the grid would be separated off under a governance arrangement that would restrain its exercise of market power. In December 1987 the Government set up an Electricity Task Force to advise on the new industry structure and regulatory requirements. The Task Force reported in September 1989, with three key recommendations: establishment of a competitive generation market, separation of the Transpower grid from the ECNZ, and introduction of competition at retail level. Box 7.1 lists the detailed recommendations.

Box 7.1 1989 Task Force Recommendations Generation ●

●

●

Generation entry barriers should be minimized and a regulatory rule against price discrimination by ECNZ be explored. Large-scale break up of the generation system is not favored but it is recommended that further study of the costs and benefits of spinning off one or two competitive generating companies be undertaken. Subject to satisfaction on competitive pressures in the generating sector, ECNZ should be privatized.

Transmission ● ● ●

The ownership of transmission assets should be separated from the generator. Distributors and generators should form a club to own the transmission grid. The regulatory framework for transmission performance monitoring should provide recourse to and reliance on intervention provisions in the Commerce Act 1986.

Distribution ●

●

●

●

Removal of franchise areas for the supply authority monopoly distribution and retailing of electricity, this to be combined with the removal of the obligation to supply. Tariffs to consumers should show transmission and distribution costs separately from energy costs. Supply authorities should be corporatized and subsequently privatized for listing on the share market. No regulation of retail energy prices, and regulation of distribution line charges should be “light handed”.

Source: Report of the Electricity Task Force, 1989.

210

Electricity Market Reform 100%

Share of generation Generation

ECNZ Transmission

Distribution ESAs Retailing

Consumers

Residential, commercial, agricultural, small and medium industrial consumers

Large directsupply industrial consumers

Fig. 7.3. Electricity industry structure 1990.

The last of these recommendations, namely no price regulation, and adoption of a light-handed approach to regulation in general, was wholeheartedly adopted by the Government. New Zealand’s early decision not to set up an industry regulator, and to rely solely on general competition law (the Commerce Act 1986) to protect the competitive process and the interests of consumers, distinguished subsequent experience sharply from that in Australia where a specialist regulator was established. Until recently, Germany (see Chapter 8) was the only other OECD country to embark on electricity restructuring without a specialist regulator. Both New Zealand and Germany have now established such regulators.

7.3.2. Initial structure Prior to restructuring, there were two tiers in the electricity sector: the NZED, a government department controlling all large generation and the high-tension transmission grid; and a large number of ESAs running low-voltage distribution networks bundled with retail energy sales and appliance sales and service. A limited number of large industrial customers took supply direct from the grid; all other final purchasers were customers of local franchise-monopoly ESAs. NZED delivered wholesale electricity (bundled generation and transmission) to distributors at a bundled price (the BST). The pre-reform structure is shown in Figure 7.3. Distributors set prices to recover their costs, with price discrimination in favor of domestic consumers (low priced) relative to commercial customers (high priced) and industrial customers (in between). This price discrimination may have been Ramsey efficient,9 but 9

Residential electricity demand is probably more elastic than commercial, because of households’ ability to switch to alternative fuels such as gas, coal, and wood.


211

was portrayed by reformers as being due solely to politically motivated cross-subsidies in favor of residential users (Jackson, 1990).10 Although it was a dominant monopoly, the NZED prior to the mid-1980s exercised its market power only in pursuit of a politically set target of covering costs and collecting a margin sufficient to fund new investment projects. Similarly, ESAs had secure monopoly franchises in their territories but their boards were accountable to consumers via regular elections, which had the effect of maintaining continual pressure on management to maintain high standards of supply and to seek only small profit margins. The restructuring timetable over the two decades from 1984 is summarized in Table 7.2. 7.3.3. Generation and transmission restructuring Change began with corporatization of the NZED in 1987 to form ECNZ. In 1994 generation was fully separated from transmission, leaving ECNZ with generation while the transmission grid company Transpower became an independent state-owned enterprise charged with operating the grid and scheduling the dispatch of generators. Thereafter, in a series of steps from 1996 to 1999, the larger ECNZ generation assets were split up among four successor companies: Contact Energy, Meridian Energy, Mighty River Power, and Genesis; while the smaller ECNZ stations (plus a number of other generation plants formerly owned by supply authorities) were privatized by sale to Trustpower, Todd Energy, and two smaller operations owned by Natural Gas Corporation (NGC) and Tuaropaki Power. Contact Energy was privatized by a share float in March 1999; the other three large successor companies remain state owned.11 By 2004 these were the eight generator class members of the New Zealand Wholesale Electricity Market.12 The evolving market shares of the main generators, as measured by capacity, are shown in Table 7.3. Two generators, Contact and Meridian, between them now account for 57% of installed capacity, with the remaining 43% distributed among the other six players. An important consequence of the break-up of the ECNZ generation portfolio was that some complementarities among different types of generation in the formerly integrated system were lost. Of the successor companies, Genesis was heavy on thermal plant and light on hydro; Meridian and Mighty River initially had only hydro and wind generation, with no thermal;13 Trustpower’s portfolio of small plants comprises entirely hydro and wind. The only operator to inherit a diversified generation portfolio was Contact Energy, the first firm to be split off from ECNZ and privatized. Contact’s ownership of large North Island thermal (at New Plymouth, Otahuhu, and Stratford), large geothermal plant at Wairakei and Ohaaki,

10

A feature of reform rhetoric in the early 1990s was the alleged need to eliminate “cross-subsidies” by lowering commercial tariffs and raising domestic ones. No evidence of the relative demand elasticities of these groups was ever publicly advanced to demonstrate that the prevailing price relativities were not Ramsey efficient. The elimination of retail price differentials in the 1990s was driven more by commercial-sector political lobbying than by economic analysis. 11 There is no evidence to date that the state-owned companies have performed any differently from the private ones. 12 http://www.nzelectricity.co.nz/C2bMarket.htm 13 Mighty River subsequently took over a 125 MW gas cogeneration plant at Southdown, and was vested with ownership of the mothballed (never commissioned) Marsden B station, which it is now planning to convert to coal.

212


Table 7.2. Major milestones in the New Zealand reform process. Event

Date

Comments

Pricing review

1984

ECNZ established

1987

Electricity Task Force Partial grid separation

1987 1988

Task Force Report

1989

Ministry of Energy abolished ESA corporatization announced Transpower Establishment Board set up Transpower Establishment Board report

1989

Officials sought revenue gains from increasing electricity prices. Corporatization of the state-owned generation and transmission system. Task force set up to design restructuring program. Transpower set up as ECNZ subsidiary to be grid and system operator. Recommendations: privatize generation and distribution, separate the grid as a club, end distribution franchises, adopt light-handed regulation. Removed Government’s in-house specialist resource, hence lowered policy and analytical firepower available to Ministers. ESA boards converted to trustees, commercial directors appointed. To implement Task Force recommendations re grid restructuring.

1990 1990 1991

Adopted the novel optimized deprival value methodology to value assets at separation from ECNZ; stuck with club ownership proposal. Energy Companies Act 1992 Distribution companies (ESAs) to be corporatized. Parliamentary Select 1992 Rejected ECNZ case for wholesale price increases; recommended Committee report adoption of progressive (increasing block) pricing of power. on pricing Echoed by private sector “Hydro New Zealand” proposal (Terry et al., 1992). Winter supply crisis 1992 May–July drought caused blackouts; ECNZ water allocation criticized. Committee of Inquiry 1992 Investigated the winter crisis, recommended greater security margins. WEMS report 1992 Private-sector proposals for generation restructuring and pricing. WEMDG set up 1993 To advance WEMS agenda for competitive pricing and by Government wholesale market. Electricity Market Co 1993 New company established to manage and monitor a wholesale market. Retail franchises 1993–1994 First small consumers, then large consumers open to retail removed competition. Full grid separation 1994 Transpower becomes a State Owned Enterprise SOE; club proposal abandoned. Disclosure regulations 1994 Information disclosure becomes mandatory for all lines businesses; accounting separation of retail and lines activities. WEMDG report 1994 Recommended competitive pool and spot market, separate grid, long-term tradable wholesale contracts, restrictions on ECNZ market power. Generation split up 1995 ECNZ to be split in two, small hydro to be privatized. Contact Energy 1996 Separate SOE generator set up with 25% of ECNZ’s generation assets. MARIA established 1996 Industry arrangements to resolve competitive reconciliation issues at retail level. Wholesale market 1996 Pool, spot price, wholesale market come into being.


213


Date

Comments

Auckland CBD event

1997

Line/energy separation ECNZ split announced

1998

Contact Energy privatization announced ECNZ split carried through MACQS agreement Ministerial Inquiry

1998

Distribution company’s line into central Auckland city fails. Recurrent blackouts, emergency new line built by Transpower. Deferred maintenance probably a contributory factor to the breakdown. All ESAs forced to divest either their retail or their lines businesses. ECNZ to be split into three state-owned generators at April 1999. Shares floated in March 1999; cornerstone 40% to Edison Mission. Now four major generators plus privatized small hydro.

1999 2000

Governance Committee

2000

Electricity Industry Bill

2001

Winter supply crisis

2001

On Energy bankruptcy

2001

Hydro spill reporting Market bids and offers disclosure Light-handed regulation fails Another dry-year looms

2002 2002

Targeted regulation

2003

Electricity Commission

2003

New regulatory framework for grid investment and pricing New market arrangements Whirinaki opens

2004

1998

1999

2002 2003

2003 2004

Electricity Governance Rules

2004

Core grid defined

2005

Industry self-governing arrangement for grid security. Reported on regulatory issues; gave lines businesses a clean bill of health. Electricity Governance Establishment Project to create a unified self-governing framework. Made provision for direct regulation of lines businesses and Government imposition of governance arrangements if industry failed to self regulate. July–September shortage due to low lake levels. Blackouts averted by voluntary savings achieved by publicity campaign. Last independent retailer driven out, all retailers now vertically integrated with generators. Hydro generators must report any spillage to waste. Full detailed information to be published with a 4-week delay. Commerce Commission retrospectively legitimizes lines businesses’ asset revaluations. March–June predictions of a dry winter, and Contact’s withdrawal of some thermal capacity, led to major spot-price spike in April. Commerce Commission moves toward regulation of lines businesses. Industry regulator set up to organize governance, oversee supply security, build and contract for reserve thermal, regulate prices. Electricity Commission to coordinate new investments in grid and generation. Electricity Commission takes over the running of the sector under new rules and regulations. New state-owned reserve generator to underpin security of supply. New governance framework decreed by Electricity Commission after industry participants fail to reach agreement. Commission identifies a subset of grid assets which must meet very high reliability standards to avoid “cascade failure”.

214


Table 7.3. Generator shares of capacity, 1994–2004. 1994 Firm ECNZ Contact Genesis Mighty River Meridian Trustpower Others Total

1998

2004

Capacity (MW)

Percent of total

Capacity (MW)

Percent of total

7391

95.9

5361 2046

66.2 25.2

696 8103

8.6 100.0

317 7708

4.1 100.0

Capacity (MW)

Percent of total

2448 1541 1266 2539 452 474 8719

28.1 17.7 14.5 29.1 5.2 5.4 100

Sources: 1994 from Electricity Enterprise Statistics 1994, pp. 24–25. 1998 from ECNZ Annual Report 1997, p. 31; Contact Energy from Prospectus dated March 31, 1999, p. 21. 2004 from Ministry of Economic Development Energy Data File January 2005, pp. 116–119.

and two of the largest South Island dams on the Clutha River, has endowed it with greater ability than its competitors to schedule its generating plant strategically.14 The wholesale electricity spot market, set up in 1996 and run by the Marketplace Company (M-Co), is based on the interaction of supply and demand.15 The final price is equal to the last offer price necessary to meet demand, in a single-price auction where all generators receive the same final price regardless of their bid prices. A constraint-adjusted spot price is then set for every half-hour at approximately 250 “nodes” on the national grid.16 In theory, each nodal price is optimized to achieve the lowest overall cost to the country as a whole, given the offers into the pool by generators.17 There has been a sharp contrast between the adoption of complex and sophisticated pricing mechanisms on the supply side of the wholesale market and the almost complete absence of scope for economic incentives to operate on the demand side. The system operator treats demand as completely price inelastic, and there is no mechanism by which either electricity saving by consumers or small-scale distributed generation can participate in the wholesale market from the demand side. In the dry-year crises of 1992 and 2001 the Government resorted to mass publicity campaigns urging voluntary savings by consumers, but at no stage have economic rewards been offered for conservation effort.18 The New Zealand electricity reforms have been notable for the absence of initiatives such as real-time retail pricing to reward conservation effort by consumers, and opportunities for small-scale distributed generators to enter the market.19 14

A detailed history of Contact Energy in New Zealand, from an avowedly critical point of view, is at http://www.converge.org.nz/watchdog/08/06.htm 15 NZEM Pricing, www.nzelectricity.co.nz 16 NZEM Pricing, www.nzelectricity.co.nz 17 NZEM Pricing, www.nzelectricity.co.nz 18 An exception may be the Comalco aluminim smelter, whose contract with Meridian Energy is confidential but is rumored to include a provision for interruptibility. 19 There is a strong contrast between New Zealand’s effective foreclosure of small distributed generation and Tasmania’s well-established policy of purchasing power from individual consumers who have installed photovoltaic equipment on their properties; see http://www.auroraenergy.com.au/askaurora/ solarpower.html#Anchor-You-33869


215

The absence of initiatives toward providing consumers with time-of-use metering and pricing, or net-metering arrangements for consumers with small-scale generation of their own, has contributed to the inexorable growth of demand and stands in sharp contrast to the finetuned and complex system of pricing signals on the supply side. Perhaps most striking has been the adoption of a detailed nodal pricing system for the delivery of power off the grid. 7.3.4. Nodal pricing A feature of the New Zealand reforms has been wholehearted adoption of the concept of detailed nodal pricing (Hogan, 1992, 1999; Ring et al., 1993a, b), with the result that there are no fewer than 250 separate nodal prices posted across a grid with only 480 entry and exit points. Much of this detail seems redundant to effective functioning of the market, and on balance has probably impacted negatively on market efficiency.20 Until recently there have been only two important bottlenecks in the New Zealand grid (see Fig. 7.2): the inter-island HVDC link, and the central North Island. (In the near future the latter constraint will shift northward to the transmission lines between Huntly and Auckland, once a planned new large thermal generator at Huntly is commissioned.21) The three key nodes in the system are Benmore (at the southern end of the HVDC link), Haywards (at the northern end of the HVDC link), and Otahuhu, in Auckland (north of the mid-North Island bottleneck). Figure 7.4 shows that the spot prices at these three key nodes move quite closely together, although from time to time one or other of the two transmission constraints binds, causing regional prices to diverge. These divergences, however, are of second-order significance relative to the overall volatility of the wholesale spot price. Price divergences at the other 247 nodes are generally insignificant. From time to time, the three principal nodal prices become separated due to grid constraints. During October 2000, for example, when the mid-North Island constraint was tight, the Otahuhu nodal price was roughly double the Haywards price, while Haywards and Benmore tracked closely together. Similarly, in January 2003, the Haywards price of 3.58 cents (c)/kWh became 5.03 c/kWh at Otahuhu, a difference of 41% from south to north of the North Island.22 An example of the HVDC constraint binding occurred in January 2002 when the Benmore price of 1.61 c/kWh was nearly doubled to 2.98 c/kWh at Haywards.23 Again in December 2002, the Benmore price of 3.65 c/kWh became 4.94 c/kWh at Haywards, and 6.12 c/kWh at Otahuhu.24

20

It could be argued that the design and implementation of the detailed nodal pricing arrangement has been driven primarily by engineers and consultants for whom the issue has been both lucrative and technically interesting. 21 Inspection of Figure 7.2 shows that major generation at or north of Huntly will be downstream of the central North Island constraint and will thereby relieve it. However, expanded transmission capacity will then be required between the new generator and the Auckland market. The siting of the new transmission line is at present embroiled in a difficult resource consent process. 22 Figures for the examples of constraint pricing here are taken from http://www.nzelectricity.co.nz/ electricity_prices/finals2003/August2003ReferencePrices.xls 23 See NZEM, Wholesale Electricity Prices Report 19 February 2002, at http://www.electricity.co.nz/ C2dPricesMonth/020219.htm 24 The main grid constraints can also bind in the opposite direction, at times when water shortages in the South Island require electricity to move south rather than north. For example, in August 2001 (a crisis period in a dry year with South Island hydro operating well below capacity) the average Otahuhu spot price was 9.93 c/kWh, the Haywards price was 11.13 c/kWh, and the Benmore price was 12.73 c/kWh.

216


300 250

Dry-year winter shortages June– August 2001

April 2003 price spike

$/MWh

200 150 100 50

N

ov Ma em y b 19 D J er 1 98 ec u 9 em ne 98 be 19 r 1 99 J Ja uly 99 nu 2 9 Au ary 000 gu 20 Se M st 01 pt arc 20 em h 01 be 200 r 2 O Apr 20 ct il 02 ob 2 e 0 N M r 2 03 ov a 0 em y 03 b 20 D J er 04 ec un 20 em e 04 be 200 r2 5 00 5

0

Otahuhu Haywards Benmore

Fig. 7.4. Monthly average spot price at three main nodes, 1999–2004. Source: Data from http:// www.stat.auckland.ac.nz/⬃geoff/elecprices/

These examples, however, are not typical of the day-to-day functioning of the system. In a month of normal operation, with no significant constraints apart from line losses, and with power moving north on the HVDC link, the three main nodal prices converge quite closely. In May 2005, for example, the Benmore spot price averaged 6.89 c/kWh, the Haywards spot price was 7.03 c/kWh, and the Otahuhu spot price was 7.09 c/kWh, an overall differential from south to north of only 3%. 7.3.5. Distribution and retail restructuring The Energy Companies Act of 1992 forced all ESAs to corporatize their operations, moving to a commercial company structure with shareholders and profit objectives. In the case of municipally owned networks this was a straightforward process, since they had well-defined owners and already operated on a commercial footing. In the case of the rural Electric Power Boards (EPBs), however, no defined owners existed. The Boards had been set up from 1918 on as “creatures of statute” which installed and managed their network assets on behalf of the consumers who elected the boards. Under corporatization, EPBs were deemed to be owned by all consumers served at the moment of the changeover. A variety of creative schemes for issuing shares were implemented in the early 1990s. Some Boards, transformed into joint-stock companies, issued shares to newly created elected trusts which held the shares on behalf of consumers in the same way as the EPBs had previous held their real assets. In other cases shares were gifted to individual consumers, many of whom took the opportunity to cash in by selling shares to private-sector interests, which quickly aggregated them into sizeable voting blocs. A period of consolidation by mergers and takeovers followed, as the more entrepreneurial of the new companies bought-up shares where possible, or took over control of trust-owned companies by direct acquisition where trust boards were willing. By early 2003, the four largest companies had captured 60%


217

Table 7.4. Consolidation of market shares in distribution networks: GWh carried.

Power New Zealand/United Networks Vector Ltd Powerco Orion Ltd Total, big four Other companies Total GWh Share of big four (%)

1995

1998

2001

2004

2569 4053 347 2416 9385 13,700 23,085 40.7

3384 4432 1019 2582 11,418 14,422 25,840 44.2

7120 4990 2083 2822 17,015 10,711 27,726 61.4

** 10,257 4074 3080 17,412 12,488* 29,900* 58.2

*Estimate. **Taken over 2003 by Vector and Powerco, who divided up the network assets. Source: MED disclosure statistics, at http://www.med.govt.nz/ers/inf_disc/disclosure-statistics/, plus company disclosures for 2004 financial year.

of the distribution lines business, up from 40% 10 years earlier. In 2003 a further merger reduced the number of industry leaders to three (see Table 7.4).25 The Electricity Industry Reform Act of 1998 forced ownership separation of electricity retailing from the operation of distribution networks. Most of the existing distributors opted to retain their natural-monopoly lines businesses and divest their retail arms. The retail businesses, with their customer bases, were quickly snapped-up in 1999– 2000 by the five main generators, which thereby achieved vertical integration of their generation plants with retail outlets.26 The supply of wholesale power to these retail affiliates then became an intra-firm transfer, largely removing any need for the large generators to enter into openmarket long-term contracts or sell more than a marginal part of their generation through the spot market. In the very light-handed New Zealand regulatory environment of the 1990s, vertically integrated generator retailers had a strong competitive advantage over stand-alone retail

25

It appeared to some observers in the 1990s that the new corporate culture of the major network companies, with its focus on mergers and acquisitions, might shift management priorities from ensuring reliability of supply to financial issues such as the market valuation of the enterprises. Claims of this sort were heard especially in relation to the failure of all the high-tension cables supplying the downtown Auckland area in 1998, due to a combination of improper installation and poor maintenance practice. An inquiry into the failure concluded that “Mercury (the relevant network company, since renamed Vector Ltd) does not have an adequate maintenance policy for 110 kV gas and oil filled cables. It did not comply with manufacturers’ recommendations in regard to the routine testing of gas pressure and oil pressure alarms and accuracy of their initiating devices, and electrical checking of the integrity of the outer coverings of the cables.” See Integral Energy Australia, Inquiry into the Auckland Power Supply Failure http://www.med.govt.nz/inquiry/publicsummary.html#P117_7323 conclusion xvii.) These failings, however, predated the corporatization process and at most it would seem that the new culture failed to remedy them. 26 Non-major retailers survived only in a few isolated rural areas such as the King Country in the central North Island. (King Country Energy’s independent-retailer status is buttressed by ownership of (and vertical integration with) local small hydro amounting to 50% of its retail load. It also has a 50% share in the large Mangahao hydro station in the Manawatu. See http://www.kcenergy.co.nz/ generation.html

218


businesses because of their ability to hold physical hedges27 within each company, whereas independent retailers had either to secure hedge contracts from generators on an extremely thin market, or face exposure to the spot price. Even faced with a dry-year crisis in 2001, the New Zealand Government took no steps to compel generators to offer hedge contracts on the open market. With no regulatory or statutory protection against the exercise of market power by the vertically integrated generator-retailers, almost all independent retailers were deprived of either profitable arbitrage opportunities or access to profitable long-term contracts, and quickly exited the market. Only a single large independent retailer remained by the end of 2000. In 1996 the Canadian company TransAlta had acquired a substantial share of the distribution and retail market, but in 1999 the company was unable to acquire a large enough generation portfolio to match its retail sales volume.28 Faced with large upstream exposure to the hedge and spot markets, TransAlta quickly sold its New Zealand business for $830 million29 to New Zealand’s dominant natural-gas pipeline and retail company, NGC. Possessing only 399 MW of generation capacity, and having failed to secure forward hedge contracts to cover the winter of 2001, NGC’s retail affiliate On Energy found itself in June 2001 in a critically dry winter with almost full exposure to the spot market for its supply of electricity.30 The company could not raise its retail price to cover the high wholesale prices, because its vertically integrated competitors kept their retail prices unchanged throughout the crisis. As NGC’s subsequent annual report ruefully noted, recording losses of $304 million from this classic cost–price squeeze:31 “Wholesale prices increased to up to four times their normal levels, placing a pronounced strain on NGC’s cash flows, profitability and financing arrangements, and raising serious questions about the operation of the market itself. NGC decided to withdraw from electricity retailing and completed its exit on August 1, 2001 following the sale of its retail electricity customers to two Government-owned energy companies. NGC’s withdrawal from that business closed off future retail exposure to the volatile wholesale electricity market and crystallized the resulting losses.” Of the retail customer base of 405,000 which NGC had acquired from TransAlta NZ Ltd the previous year, representing 23% of all electricity consumers, 115,000 were sold to Meridian Energy and 290,000 to Genesis Power Ltd. Since then the vertically integrated five generator oligopoly of retailing has been unchallenged. The elimination of non-generator parties from the retail market spelt a halt to the process of competition for retail customers, which had briefly flourished in the 2 years following the 1998 separation of lines and energy retail activities. Figure 7.5 shows that the new-entrant

27

Retailers can hedge their costs of future wholesale supply either by long-term contracts with generators, or by directly owning physical generating plant. The practice of physical hedging in New Zealand has foreclosed the emergence of a liquid hedge market; this in turn has constituted a major barrier to new entry by independent retailers. 28 TransAlta in 2000 held more than 20% of New Zealand’s electricity consumers but less then 5% of generating capacity. 29 NGC Becomes Majority Owner of TransAlta, media release dated 31 March 2000, http://www.ngc.co.nz/ article/articleprint/166/-1/21/. The price represented a $300 million tax-free capital gain for TransAlta. 30 The wholesale spot-price spike of June–August 2001 is dramatically apparent in Figure 7.3 above. 31 Natural Gas Corporation, Annual Report 2001, p. 5.


219

40 35 30 (%)

25 20 15 10 5 April 1994 October 1994 April 1995 October 1995 April 1996 October 1996 April 1997 October 1997 April 1998 October 1998 April 1999 October 1999 April 2000 October 2000 April 2001 October 2001 April 2002 October 2002 April 2003 October 2003 April 2004 October 2004

0

Fig. 7.5. Share of non-incumbent retailers in former franchise territories. Source: Stratagen.

share of retail sales in former franchise territories, following removal of franchises in 1993–1994, remained very low until the Electricity Industry Reform Act 1998 separated retail from distribution. Retail competition took off in 1999–2000, but froze again at around 30% as soon as On Energy had been driven out in mid-2001. Three years later, Murray and Stevenson (2004, p. 18) reported to the Electricity Commission that “customer switching figures seem to have declined and stabilized over a period when prices have been rising” and that “price trends suggest electricity prices are probably higher on average than they would be in a workably competitive market”. The 1989 Task Force vision of competitive retail markets served by a liquid market for forward hedge contracts, thus, ran aground on the reality of generators’ market power. The anti-competitive effect of vertical integration of generation with retail had not been foreseen at the time of the 1998 separation of retail from distribution networks. Consequently no consideration was given to requiring generators to transact with their retail affiliates via an arms-length contestable market for hedge contracts, and although proposals for such compulsory hedging were discussed during the 2001 crisis, Government took no steps to remedy the extreme thinness of the forward contracts market.32

7.4. Pricing, Profitability, and “Light-Handed Regulation” 7.4.1. “Efficient” pricing A dilemma over the meaning of “efficient prices” dogged the electricity reform process from the outset, and remains an unresolved issue two decades later. One interpretation in the mid-late 1980s was that since the electricity system was breaking even in cash terms at its existing prices, efficiency-enhancing reforms ought to bring down the prices paid by final consumers, and certainly ought not to lead to rising prices.

32

The issue now rests with the recently established regulator, the Electricity Commission.

220


A May 1990 press statement by the then Minister of Energy reassured consumers:33 “Lower real electricity prices resulting from the corporatization of the electricity distribution industry is the motivation for the latest Government decisions on electricity which were announced today … Savings in electricity bills represent an immediate improvement in living standards and help toward the restoration of full employment ….” An opposing view from the outset was that economic efficiency required prices to increase, to raise the industry’s return on capital to a commercial level (Ministry of Energy Financial Objectives and Pricing Review Team 1984). In addition, an across-the-board price increase was allegedly needed to signal to electricity users the marginal cost of new supply (Electricity Corporation Establishment Board, 1987; New Zealand Treasury, 1987). There was general agreement that low-cost generation options had been fully exploited by the 1980s, and that new generation and transmission capacity would be costly to install. Faced with an upward-sloping LRMC curve, the choice between average- and marginalcost pricing presented a political dilemma. If the restructured electricity industry were to be allowed to price at LRMC, the inevitable result would be higher prices to consumers and very large operating surpluses on the existing hydro generation plant, far in excess of the surpluses required to yield a competitive return on, and of, the book value of alreadyexisting capital (Bertram, 1988). If a lower average price were set to recover the full cost of supply, including a commercial rate of return on the book value of existing assets, then the resulting price signal would render new investments unattractive while encouraging excessive growth of demand. Two solutions to this dilemma were on offer. The consumer-oriented position was either to stick with an average-cost price and accept any consequent inefficiencies;34 or to adopt a non-linear tariff structure to achieve the same outcome of restricting existing generators’ total revenue, while providing efficient price signals at the margin. The latter solution was supported by a parliamentary select committee (New Zealand House of Representatives, 1992) and in a report commissioned by a group of major users (Terry et al., 1992).35 The other approach to wholesale pricing, championed by the Treasury and ECNZ, was to charge consumers the full LRMC price, and to legitimize the resulting cash surpluses that would accrue to generators, the grid operator, and the distribution networks, by revaluing their existing assets up to a level at which the rate of return on capital would appear to be no more than “normal”. In 1987 Treasury had estimated that the BST should be raised from less than 6 c/kWh to somewhere in the range 8–11 c/kWh (New Zealand Treasury, 1987, p. 4). The Labor Government, which initiated the reforms, was replaced at the 1990 general election by a National Party regime in which the Treasury view prevailed. In terms of

33

Hon David Butcher, press statement dated May 25, 1990. Advocates of this approach in the mid-1980s included the New Zealand Business Round Table (1985), Ernst and Whinney (1985), Frater et al. (1985) Jarden and Company (1985), McDonald (1985), Scott and Co (1985), and University of Waikato Interfirm Comparison Unit (1985). 35 Another pricing arrangement with the same basic thrust would have been to rebate to consumers any excess profits resulting from application of a uniform LRMC price, possibly by means of a lump-sum reduction in fixed lines charges funded from generation surpluses, along the lines later adopted in the UK by Scottish Hydro. 34


221

electricity-sector reform, this meant support for “full-cost uniform pricing” of electricity, which translated in practice into higher overall prices for consumers, with any efficiency gains that might result from restructuring being captured as additional profit. Treasury argued that electricity prices needed to rise rather than fall, to signal LRMC; that no regulatory barrier should be placed in the way of electricity suppliers pushing their prices up to the “limit prices” at which, in theory, the threat of entry by new competitors would cap prices; and that gains from increased prices and/or reduced costs, provided they fell below the contestability threshold, could legitimately be taken as profits and built into the asset valuations shown in the companies’ regulatory accounts. This tolerance for wealth transfers from consumers to suppliers36 meant that New Zealand’s regime of so-called “light-handed regulation” lacked any bright-line test for abuse of market power until all assets had been revalued up to the replacement-cost ceiling, and companies had adjusted their margins to match the higher ratebase. It also reveals the extent to which New Zealand policy-makers adopted without qualification some recent developments in economic and accounting theory, which other OECD governments have treated with more circumspection. 7.4.2. Economic and accounting theory and the New Zealand reforms Economic policy-making in New Zealand in the late 1980s and early 1990s was heavily influenced by three overseas developments in the economics and accountancy literature. These were: ●

●

●

36

The proposition, familiar from early UK debates over electricity restructuring, that electricity generation and retailing were potentially competitive activities and that in relation to those two levels of the electricity market, therefore, policy intervention could be limited to promoting competitive conditions, not to controlling prices. The theory of contestable markets set out in Baumol et al. (1982). Contestability theory was interpreted to mean that in a process of “competition for the market”, a natural monopolist would be unable to price above the limit at which a new entrant would be attracted. This, New Zealand officials reasoned, meant that if an incumbent monopolist’s assets were revalued up to replacement cost, no more than a competitive rate of return on that valuation would be achievable unless management could cut costs by improving efficiency. Hence, although electricity lines networks were natural monopolies, officials decided no regulatory restraint on price would be necessary, as market disciplines would do the job unaided; all that would be required would be transparent information disclosure. The newly fashionable method of accrual (current-cost) accounting, which prescribed that fixed assets should be continually revalued to market value, and that profit and loss statements ought to reflect changes in shareholder wealth accruing as a result of each year’s trading activity. In the hands of the New Zealand accounting profession, this methodology was incorporated into “generally accepted accounting practice” (GAAP) in a partial manner that opened the way to manipulation of asset valuations. To summarize a complex story, New Zealand’s Accounting Standard SSAP28 (later FRS3) prescribed that natural-monopoly entities whose assets do not (by definition) have a competitive

The two Government departments most closely associated with electricity regulation during the 1990s, Treasury and the Ministry of Commerce, adopted and promoted the so-called “total surplus standard” for regulation. This standard treats all pure transfers as welfare-neutral and hence of no concern to the regulatory authorities. See Bertram (2004).

222


arms-length market value, should value their fixed assets at optimized depreciated replacement cost (ODRC), which was claimed to approximate the capital cost of setting up from scratch a new supplier providing the same service, to the same standard, as the incumbent (Cooper, 1995). This valuation could then be used as the ratebase for setting and justifying prices. In the view of the officials overseeing the light-handed regulatory regime during the 1990s, no concern over excess profits could arise so long as no more than a competitive rate of return on the ODRC-valued assets was revealed in the regulatory accounts prepared for disclosure purposes by all transmission and distribution network owners. If assets were to be continually revalued to the hypothetical contestability limit, consistency required that the profit-and-loss account should record as income all wealth changes accruing to the shareholders, whether by virtue of current cash flows or of asset revaluations. New Zealand’s GAAP, however, did not (and still does not) require this to be done for upward revaluations. Gains and losses on the actual sale of particular assets are recorded in the profit-and-loss account, as are all negative revaluations (asset write-downs). The crucial omission is the treatment of upward asset revaluations (effectively, negative depreciation). Rather than being recorded as revenue in the profit-and-loss account, these are recorded separately in a “revaluation reserve”, usually hidden deep in the notes to the financial statements. Under this procedure, the accrual to a company’s books of hundreds of millions of dollars of revaluations of fixed assets need never be recognized as income, and so can be excluded from recorded profits for both taxation and regulatory purposes,37 while the revalued assets can be used as the ratebase for price setting and justification. 7.4.3. Generation and the wholesale spot price Figure 7.6 shows the generation supply curve for May 2004, constructed by stacking the various generation plants in merit order of variable operating cost. The large hydro plants, 11 9

c/kwh

7 5

Average monthly price ⫽ marginal cost of last plant in the stack

3

2004 supply curve May demand

1 -1 380

880

1380

1880 2380 GWh per month

2880

3380

Fig. 7.6. Generation supply curve 2004.

37

This practice is acceptable to the tax authorities because New Zealand does not have a capital gains tax.


223

whose operating cost is close to zero, crowd the higher-operating-cost thermal and geothermal units out to the marginal one-third of the market. The upward-sloping curve at the right of the diagram shows these various non-renewable units stacked in merit order. The market-clearing spot price, which provides, over the long run, the anchor for long-term wholesale supply contracts, is found at the point on the supply curve at which aggregate demand intersects the supply curve. The May 2004 demand, it can be seen, lay only about 400 GWh (14% of monthly demand) inside the point at which the supply curve turns sharply upwards. In this situation, radical price spikes can be anticipated if either demand rises, or supply falls, by this amount. The months following May are winter in New Zealand, when demand is higher and the system’s ability to meet demand without price shocks rests heavily on the volume of very low-operating-cost hydro generation made available by the owners of hydro plant. By withholding even a small part of the available water from use for generation at times of strong demand, the owners of large hydro plants can potentially pull the bidstack to the left, thereby (deliberately or inadvertently) driving up the spot price and raising their operating surplus – an opportunity for the exercise of market power mitigated in the New Zealand case only by the existence of a duopoly, rather than a monopoly, of major hydro generators with the necessary market leverage. The very steep profile of the supply curve beyond about 3000 GWh per month confers substantial market power on any hydro generator (or cartel of generators) with the ability to withhold capacity and thereby shift the bidstack to the left. To achieve such withholding, a hydro generator must either have unutilized water storage capacity which can be allowed to fill while generation is curtailed; or else must be able to dispose of unwanted water by hydro spill. New Zealand policy-makers became aware only in 2001 (5 years after the break-up of the ECNZ generation portfolio) of the possibility that hydro generators might game the spot price by spilling water to waste. The Government’s ex-post review of the 2001 dry-winter supply crisis brought to light the fact that in the summer of that year Meridian Energy had been spilling water from Lake Tekapo. Whether this was strategic behavior to drive up price (as one distributor alleged), or responsible management to avoid flood risk (as Meridian claimed),38 the issue was placed on the agenda for regulation, and new rules subsequently came into force requiring generators to report each month on the details of any spill.39 Since 2001 there has been very little hydro spill recorded. Use of empty storage capacity to withhold water, however, is not so subject to Government control. An example of the strategic importance of commercial generators’ restriction of hydro generation in order to build up (or protect) the level of storage lakes was the price spike of April 2003, visible in Figure 7.3. Rainfall in the early months of 2003 was below normal, and lake storage fell below the levels required to ensure ability to meet the forthcoming winter demand. The two large hydro generators in the South Island (Meridian and Contact) both cut back water use, citing the need to conserve water and maintain storage levels ahead of the coming winter. At the same time, Contact took its 357 MW gas-fired Stratford station offline in mid-April for

38 See Electricity Post-Winter Review, 2001, Section 2.2, http://www.winterreview.govt.nz/submissions/ summary/summary-03.html#P186_28826 39 Hydro spill reporting is now to the recently established Electricity Commission; see http://www. electricitycommission.govt.nz/opdev/secsupply/sos/overview/hydrospill1/

224


maintenance.40 Both actions shifted the bidstack significantly to the left. As lake storage levels dropped to 60% of normal for the time of year, the spot price was driven up sharply to an average for the month of 20 c/kWh, and Government solicited voluntary demand restraint by electricity users in order to avoid blackouts. Rainfall subsequently increased during May and June and the supply situation eased, bringing the spot price back down to 6 c/kWh by June. The extreme volatility of the spot market in the April 2003 event was attributable not only to the high shadow price implicitly assigned to water by Contact and Meridian Energy. It was worsened significantly by the fact that the New Zealand bidstack in 2003 had far less reserve thermal plant, and hence a much steeper right-hand end, than had been the case at the beginning of the reforms. In the dry year 1992 the ECNZ portfolio had included four high-operating-cost thermal plants, which were brought online to compensate for the water shortage. Because these plants’ capital costs were sunk, the only economic cost of bringing them online was the operating cost, primarily fuel. By their mere existence, these plants exercised a moderating influence over the spot market by placing a ceiling on the spot price over a range of several hundred MW of supply capacity. Table 7.5 shows the thermal high-cost reserve capacity that had been available in 1992 (when operation of the Marsden A oil-fired station during the dry-winter crisis reduced the scale of blackouts in the Auckland region), and compares this with the corresponding reserve capacity available in early 2003 before the April price spike. The difference is striking. Under commercial incentives and supposedly competitive conditions, the former owners of thermal reserve plant had decommissioned and/or demolished a total of 620 MW of reserve capacity.41 Over the same period, roughly 1000 MW of new thermal plant was commissioned, but none of this qualified as reserve capacity to cover dry years; the Southdown and Otahuhu B stations simply helped supply to keep up with growing demand, while the cogeneration stations perform no role in relation to dry-year firming, since their operation is tied to the steam requirements of the host facilities. Having failed to persuade any of the commercial generators to invest in new reserve plant, the Government opted in 2004 to spend $160 million on construction of a new 155 MW diesel-fired thermal station at Whirinaki, where Contact Energy had demolished an almost identical plant a couple of years previously. The station, although owned by the Crown, is maintained and operated by Contact Energy under contract, and is not to be dispatched at a price of less than 20 c/kWh42 (roughly the monthly average price during the April 2003 price spike, see Fig. 7.3). 7.4.4. Grid pricing The high-voltage transmission grid was transferred in 1994 to a new State-Owned Enterprise, Transpower Ltd, following several years of debate over asset valuation and pricing. 40

NZEM, Declining Storage Levels Fuel Rising Electricity Prices 7 May 2003, http://www.nzelectricity.co.nz/ C2dPricesMonth/030508.htm 41 Prior to 1992, the 133 MW Meremere coal-fired station in the Waikato had already been decommissioned by ECNZ in 1990. Marsden A (114 MW) was closed in mid-1992 and demolished in 1997. Stratford (200 MW) closed in late 1999. Otahuhu A (90 MW) and Whirinaki (216 MW) were decommissioned in 2002. 42 Electricity Commission, Explanatory Paper to the Initial Security of Supply Policy, June 2005, http:// www.electricitycommission.govt.nz/pdfs/opdev/secsupply/policy/Initial-SOS-Policy-ExplanPaper.pdf, Part VII p. 21.

225

Restructuring of the New Zealand Electricity sector 1984–2005 Table 7.5. Thermal generating capacity, March 1992 and March 2003 compared. Station New Plymouth Huntly Stratford TCC Otahuhu B Southdown Big thermal total Stratford Otahuhu A Marsden A Whirinaki Total high-cost dry-year reserve thermal

1992 capacity (MW)

Operating cost, c/kWh, 1991

2003 capacity (MW)

580 1000 198 0 0 1778

3.13 2.92 3.97 n.a. n.a.

400 1000 355 380 118 2253

200 90 114 216 620

3.97 6.27 7.43 18.5

0 0 0 0 0

0 0 0 0 0 0 0

n.a. n.a. n.a. n.a. n.a. n.a.

52 40 44 25 355 65 581

Te Awamutu cogen Kinleith Te Rapa Edgecumbe Kapuni Whareroa Cogen total Total thermal

2398

2834

Sources: 1992 capacity data from Annual Statistics in Relation to Electric Power Operation in New Zealand for the Year Ended March 31, 1992, pp. 57–59. 2003 capacities from Energy Data File July 2003, pp. 108–109. Operating-cost estimates from Terry et al. (1992), p. 128.

Following the 1987 transfer of the NZED generation and grid assets to ECNZ at a negotiated vesting value of $6.3 billion, ECNZ undertook the task of allocating this lump-sum valuation across its generation and grid assets. The transmission system was assigned a value of $2.1 billion, and generation and other fixed assets $4.2 billion.43 In July 1990 the Transpower Establishment Board was set up to oversee the separation of the grid from ECNZ. A central issue confronted by the Board was the valuation that should be assigned to the grid assets when they were fully vested in a new independent company. ECNZ management and Treasury were focused on achieving privatization of the generation assets at a high price, and this could best be achieved by off-loading as much as possible of the Corporation’s debt into the books of its grid subsidiary, allowing the generation assets to be sold relatively unencumbered by debt. In addition, a range of operating expenses formerly attributed to generation were transferred to Transpower prior to separation (Terry et al., 1992, p. 87), raising the reported profitability of ECNZ’s generation business in readiness for sale. A higher valuation of the grid assets was then required to bring Transpower’s debt– equity ratio down to a commercially sustainable level. The TPEB achieved this objective by having the grid assets revalued to “optimized deprival value” (ODV), a variant of depreciated replacement cost. This resulted in a valuation of $2.55 billion (Ernst et al., 1991). The higher asset value and increased operating costs were used to justify a real increase of 21% between 1989 and 1991 in the grid transmission charge per kWh conveyed. 43

ECNZ Annual Report 1989, p. 47.

226


Over the decade following its establishment as an state-owned enterprise, Transpower paid down its debt and wrote-down its ODV asset valuation in recognition that the longrun sustainability of the grid itself depended upon transmission prices low enough to compete with distributed generation connected directly to distribution networks, for which transmission service would not be required. To protect the grid’s pre-eminent position in the short term, Transpower used its market power to impose contract conditions on distributors which obliged them to collect transmission charges on all power delivered, whether it was taken from the grid or generated locally by suppliers connected only to the local network. These contract conditions, by imposing high fixed connection charges regardless of load changes, also eliminated the prospect that retailers might be able to profit from demand-side conservation initiatives. The resulting barrier against entry by small-scale distributed local generation, and the equally suffocating effect on local demand-side conservation initiatives, effectively foreclosed development of both for a decade. 7.4.5. Distribution networks Legislation to force through the corporatization of ESAs was passed in 1992, and the process was largely completed by April 1994. As the new companies were set up, the issues of asset valuation and price setting had again to be addressed. Following the Transpower precedent, the Minister of Energy and the Treasury planned to revalue all assets up to ODV prior to vesting, enabling the new distribution companies to start off with a new, higher ratebase against which their profitability could be monitored under a light-handed regulatory regime of information disclosure. It was obvious to all industry participants, including major users, that the historic-cost asset valuations in the books of the pre-corporatization ESAs were far below depreciated replacement cost. Roughly speaking, at 1994 the network assets of all networks combined had a book value of $2 billion, but a replacement-cost valuation would come to double that amount.44 If the new companies were gifted a $2 billion asset revaluation at the time the assets were vested, two politically significant groups stood to lose. One group was electricity users, who effectively would have to pay for the increased profits required if the distribution companies were to meet commercial rate-of-return targets on their revalued ratebases. The other group were private investors eager to make capital gains by acquiring distribution assets cheaply and then undertaking the revaluations themselves. Early in the restructuring process it became apparent that switching to a replacement-cost ratebase for pricing supply to customers in low-density areas would sharply increase electricity prices in low-density rural areas with a high ratio of line length per customer. A confidential survey undertaken by officials in 1989–1990 found that “full-cost pricing” would require price increases of up to 300% for rural electricity users.45 Faced with the prospect that the political fallout would halt the reform process at the outset, Treasury fell back on a modified form of replacement-cost valuation called ODV, which included the proviso that whenever the economic value of an asset (the discounted present value of expected revenues46) was below full ODRC, the asset would be written down and the users of the asset 44

Cabinet documents recently released under the Official Information Act reveal that these orders of magnitude were known to ministers and officials in 1991, 3 years before vesting took place. 45 Cabinet committee document SAS (90) 31, March 13, 1990, p. 10. 46 The circularity between asset values and revenues was well understood. The ODV technique enabled the revenues extracted from specific groups of consumers to be selectively capped, with the ratebase valuation of the assets serving that group written down accordingly, leaving an ostensibly competitive market return on the assets for disclosure purposes.

New Zealand $ billions

Restructuring of the New Zealand Electricity sector 1984–2005 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

227

Revaluations Depreciatedhistoric-cost estimate 1992 1994 1996 1998 2000 2002

Fig. 7.7. Asset book values of electricity distribution networks, 1992–2002. Source: Bertram and Terry (2000, p. 7).

thereby protected from rate shock. This ingenious solution became embedded thereafter in the valuation procedures for both grid and distribution networks. In the event the new National Government elected in late opted pragmatically to continue the time-honored practice of using revenues from densely populated parts of each ESA’s territory to cross-subsidize the prices charged in low-density areas – a procedure which had been generally accepted by consumers since the 1920s.47 Neither the ODV concept nor the decision to retain urban–rural cross-subsidies removed the looming prospect of a general price shock if network asset valuations were doubled across the board. In October 1991, Ministry of Commerce officials estimated that the ODV valuations would be 2.5 times the existing book values.48 Modeling carried out for the Government in April 1992 by a local accountancy practice suggested that a rate shock of 25% would be required to meet the required return on a revalued ratebase.49 Treasury at this stage proposed that the assets should be vested at book value but that the new companies be allowed to revalue to ODV without facing any regulatory restraint. It would then be the responsibility of the new corporate boards to decide whether to squeeze their customers or accept below commercial rates of return.50 Cabinet agreed,51 and the Establishment Boards of the new companies were instructed to adopt existing book values for their opening balance sheets. Figure 7.7 shows the subsequent process of increasing the regulatory ratebase by writing-up asset values to replacement cost. Figure 7.8 shows the evolution of prices and average costs of lines networks over that period. Free from regulatory restraint, the sector raised its aggregate Lerner Index from 0.36 at vesting to 0.68 by 2001. The loophole in the regulatory system was well known to, and understood by, industry insiders. It was equally obvious to analysts familiar with current-cost accounting theory. The procedure of vesting the assets at historic cost, while signaling to the new owners that ODV valuation would be the regulatory benchmark, transferred responsibility for 47

Corporatized ESAs are compelled, under the reform legislation, to maintain supply to all rural customers until 2013. Thereafter they will be allowed to disconnect unprofitable customers. 48 Ernst and Young, letter to Michael Lear, Ministry of Commerce, May 14, 1992, p. 1. 49 Ibid., p. 3 of appendix. Ernst and Young pointed out in this letter that recognizing asset revaluations as income would reduce the required rate shock to between 5% and 9%, but the point was not taken by officials. 50 Officials’ briefing document for Minister of Energy, May 8, 1992. 51 Cabinet State Sector Committee document STA (92) 96.

228

Electricity Market Reform 5.00 4.50 c/kWh conveyed

4.00 3.50 3.00 2.50 2.00

Average revenue

1.50

Average operating costs

1.00 0.50 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

0.00

Years to March Fig. 7.8. Price–cost margin of electricity distribution networks, 1991–2002. Source: Bertram and Twaddle (2005, p. 295), Figure 1(f).

subsequent increases in margins and prices from the Government to the distributors but left consumers unprotected. In a current-cost accounting framework, profitability should be measured with revaluations (changes in shareholder wealth) recorded in the profit-and-loss accounts. New Zealand’s GAAP did not require this to be done.52 Lines companies were therefore able to inflate the denominator and reduce the numerator in their profit calculations, as justification for an annual wealth transfer from consumers to distribution network owners of $200 million annually (0.2% of GDP) from the late 1990s on (Bertram and Twaddle, 2005). Perhaps ironically, the information disclosure regulations for electricity lines networks, promulgated in 1994, included a requirement for companies to disclose an “accounting rate of profit” which included the wealth effects of ratebase revaulations,53 and this requirement was complied with, resulting in the disclosure of profit rates often of 30–40%, and in one case as high as 90%,54 with no reaction from Government.55 52

This issue had been thoroughly discussed prior to the UK privatizations, and the regulatory accounting implications worked out, in the “Byatt Report”, Accounting for Economic Costs and Changing Prices: A Report to HM Treasury by an Advisory Group, London: HMSO, 1986, Volume 1. 53 Ernst and Young, as advisers to the Ministry of Commerce, set out the correct accounting procedures in a letter of May 14, 1992, and explained the correct interpretation of the Accounting Rate of Profit (later renamed the Return on Investment) in Ernst and Young (1994). 54 Far from recognizing the implications of these numbers, a 2000 Ministerial Inquiry rejected the calculation methodology itself and found no grounds for regulatory concern (Caygill et al., 2000, Table 7.3, p. 14, and p. 15 paragraph 75). 55 The largest lines company, United Networks, disclosed a return on equity of 235% for 2000, 347% for 2001 and 125% for 2002, without attracting attention from Parliament, media, or officials. See New Zealand Gazette 2000, No. 111 p. 2807 (http://www.dia.govt.nz/Pubforms.NSF/URL/UnitedNetwork 111Aug00.pdf/$file/UnitedNetwork111Aug00.pdf ); 2001, No. 104 p. 2665. (http://www.dia.govt.nz/ Pubforms.nsf/URL/Unitednetworks104Aug01.pdf/$file/Unitednetworks104Aug01.pdf); and 2002, No. 122 p. 3272. (http://www.dia.govt.nz/Pubforms.nsf/URL/UnitedNetwork122.pdf/$file/United Network122.pdf ). In fairness it should be noted that the taking of monopoly profits is not illegal under New Zealand competition law. Consumers have no legal redress against high prices, and the Electricity Complaints Commission set up in 2001 was barred from hearing complaints about pricing. See http:// www.electricitycomplaints.co.nz/faqs.htm


Contestability limit

Market value

229

Say 2.5 ⫻ ODV ⫽ $8.4 billion

$ billions

Full replacement cost

Depreciated replacement cost

Costless-entry limit

Optimised DRC/ODV

4.2 billion

Historic cost

$2 billion

Net realisable value 0 Fig. 7.9. Range of possible ratebase valuations for the distribution networks.

Figure 7.9 shows schematically the range of feasible asset valuations, any of which could have been arbitrarily chosen for ratebase purposes. The theoretical limit valuation under conditions of perfect contestability (zero costs of entry and exit) is represented by the ODV of $4.2 billion – more than double the pre-corporatization historic cost. Adding in the observed effects of barriers to entry (in particular, very high fixed costs of entry and exit) raises this further by a factor of 2.5 (based on the actual purchase price of distribution networks taken over as going concerns). In short, the New Zealand regulatory regime for lines businesses prior to 2003 encouraged ratebase revaluation up to ODV, which was achieved by the network businesses over the first 6 years of reform. Thereafter, as market expectations factored in the lack of credibility of the light-handed regime, network assets changed hands at “fair value” levels, which included the discounted value of expected future regulatory tolerance. The market judgment in these transactions suggested that an actual contestability limit valuation would be of the order of $8.4 billion for all networks aggregated. The essential issue raised by asset revaluations throughout the electricity sector was not the theoretical choice of valuation methodology per se; there are ample precedents around the world for both the historic-cost and the replacement-cost approach, with matching implications for the setting of the warranted rate of return on the resulting ratebase. The central issue was the New Zealand Government’s decision to radically change the ratebase valuation methodology in asset mid-life, causing a dramatic levy (several billions of dollars) on the aggregate wealth of consumers, for the benefit of electricity suppliers. No protection was provided for consumers against this wealth expropriation. In particular, no regulatory provision required suppliers to compensate consumers for the wealth transfer, whether by means of rebates or through allocation of shares in the newly created equity value of suppliers. In August 2001 Parliament passed a set of amendments to the Commerce Act 1986, giving the New Zealand Commerce Commission the task of regulating transmission and distribution lines networks. The Commission conducted lengthy hearings on the pricing practices of the electricity networks sector, and eventually decided to use the status quo of mid-2002 as its ratebase for future profit-cap regulation. The revaluations and widening margins of

230


the 1990s were thereby retrospectively legitimized.56 This regulatory outcome was inherited in 2003 by the new Electricity Commission.

7.5. The Electricity Commission: Back to An Industry Regulator After more than a decade of experimentation with light-handed regulation, the New Zealand Government finally in 2003 followed the example of most other OECD countries by setting up a specialized industry regulator to oversee the electricity industry. The Electricity Commission is charged with a wide array of tasks: managing the ongoing information disclosure regime, setting price and revenue caps, coordinating the investment plans of various industry players, maintaining reserve generation capacity, overseeing industry governance arrangements, and guiding new investment by the issuing of “statements of opportunity”, to name a few. Further privatization is off the policy agenda. Several regulatory issues, however, remain unresolved: ●

●

●

There is little prospect that the incumbent generators will be forced to divest their retail affiliates; yet without such divestment, new competitive retail entry remains foreclosed. Similarly, although Government has declared itself in favor of the rapid development of distributed generation, Transpower’s grid pricing practices, which foreclose most opportunities for such projects, remain in place. Since 2002 a rush by large incumbent generators to build wind farms is raising a raft of difficult coordination problems, since the location of favorable sites for wind farms, and of the hydro generation assets that can be used to back-up wind generators, does not always coincide with the existing grid infrastructure, presenting the grid’s operator, and the new regulator, with investment and coordination requirements not foreseen even a few years ago.

7.6. Conclusion: The State of Play at 2005 The structure of the industry in 2005 is shown in Figure 7.10. Of the 1989 Task Force recommendations, some have been implemented while others have been abandoned along the way. ECNZ has been broken into five separate generators (the Task Force had recommended against breakup). Only two of these generation companies are in private hands, while the Government continues to own 60% of generating capacity. The Task Force’s fear that generation breakup without an industry regulator might result in losses of efficiency in the coordination of scheduling and investment seemed to have been borne out by 2002, and partly in response to this a new industry regulator was introduced in 2003. Generation and transmission were separated early in the reform process, but the Task Force’s proposal for club ownership of Transpower was rejected early on by industry participants, leaving the grid in state ownership.57

56

The Commission’s deliberations are fully recorded at http://www.comcom.govt.nz/Industry Regulation/Electricity/ElectricityLinesBusinesses/Overview.aspx 57 The Government attempted to implement the club proposal in 1992, but distributors refused to take part in the formation of a club in which their interests would have been diametrically opposed to those of generators, but in which they would not have had sufficient voting power to form a blocking coalition. The Task Force had failed to appreciate the likely extent of these conflicts of interest.

231

Restructuring of the New Zealand Electricity sector 1984–2005 Share of generation Generation

Electricity commission

26.9%

31.2%

16.6%

12.6%

7.9%

4.8%

Contact Energy

Meridian Energy

Genesis Power

Mighty river Power

Other IPPs

Cogen

Electricity Governance Regulations and Rules 2003

Transpower

Transmission

Local network companies

Distribution

Electricity retailers

Retailing

Consumers

Large directsupply industrial consumers

Residential, commercial, agricultural, small and medium industrial consumers

Fig. 7.10. Electricity industry structure 2004.

Corporatization of ESAs has been carried through, but less than half of distribution network assets have been fully privatized, and the interim trust-ownership arrangement has become entrenched in many rural and small-town systems. Retail franchises have been abolished and retail operators separated from lines networks, but competition at retail level quickly stalled once retailers and generators became vertically integrated. No liquid market for hedge contracts has yet emerged – a defect still to be addressed by the new industry regulator. The main buyers in the wholesale market are the retail affiliates of generating companies, plus major manufacturers taking supply directly from the grid. Direct consumer exposure to spot market prices was estimated in 2003 to be no more than 10–15%,58 which is not surprising given that the great bulk of the wholesale market is intra-firm. Customer invoices continue to be presented without disaggregated line-item information that would enable consumers to identify the costs incurred at each stage of the supply chain – a level of information disclosure which the Task Force regarded as fundamental to retail competition, but which has never been mandated by Government.59 Possibly the most important lesson from the New Zealand experiment has been the failure of the Task Force’s preferred model of light-handed regulation. Industry self-regulation under information disclosure failed comprehensively over a full decade of attempted implementation. Generators and distributors proved unable to agree on club governance for the grid in 1992–1994. Generators, distributors, retailers, and Transpower were unable to agree 58

Commerce Commission, Decision number 491, www.comcom.govt.nz The New Zealand Commerce Commission, as de facto industry regulator from 2001 to 2003, repeatedly drew attention to this gap in the information disclosure arrangements, with no response from Government. 59

232


on an industry-wide governance arrangement and rules by 2003, when the Government ran out of patience and established the Electricity Commission.60 After a tentative start, lines companies pushed their margins up from 30% to 70% without triggering any regulatory response from Government. The Commerce Commission (following an inquiry in 2002) retrospectively validated the practice of revaluing the networks’ asset ratebases and hence validated also the radically increased price–cost margins in that sector. The Information Disclosure Regulations introduced in 1994 obliged lines businesses to disclose their financial statements, but New Zealand’s GAAP allowed true rates of return to be hidden in the notes to the accounts, leaving lay members of the public (including, apparently, officials and ministers responsible for oversight of the regulatory regime) in the dark on key issues of pricing and profitability. Looking forward, major new challenges loom on the horizon. New Zealand’s sole large gas field (Maui) is expected to be exhausted by 2007, and only relatively small fields have been located to replace it, raising the possibility that thermal generation will shift to reliance on liquefied natural gas (LNG) or coal. Coal will then be the cheaper thermal option61 unless New Zealand’s compliance with the Kyoto Protocol leads to substantial carbon taxes. In addition, the past 2 years have witnessed large-scale investment in wind farms, which will transform the nature of demands on the grid as wind is matched to (mainly hydro) back-up. Installed wind generation reached 168 MW by the end of 200462 and a further 700 MW of projects are in the planning stage,63 raising the prospect that wind turbines will soon make up over 10% of total generating capacity. Key policy challenges facing New Zealand in the next decade involve dealing with these new issues as well as matters that were ignored or left unresolved in the first round of restructuring. These include the implementation of the Kyoto Protocol to which New Zealand is a party; opening up the demand side of the electricity market to new initiatives such as smallscale distributed generation, time-of-use metering and charging, and net metering of customers with their own generation capability; and breaking the logjam in retail competition. With an electricity regulator at last firmly established, there is an opportunity to make progress on these items of unfinished business.

References Baumol, W.J, Panzar, J.C. and Willig, R.D. (1982). Contestable Markets and the Theory of Industry Structure. Harcourt, Brace and Jovanovich, New York. Bertram, G. (1988). Rents in the New Zealand Energy Sector. Royal Commission on Social Policy, The April Report, Vol. IV. Government Printer, Wellington. Bertram, G. (2003). New Zealand since 1984: elite succession, income distribution, and economic growth in a small trading economy. Geojournal, 59, 93–106. Bertram, G. (2004). What’s wrong with New Zealand’s public benefits test? New Zealand Economic Papers, 38(2), 265–278. Bertram, G. and Terry, S. (2000). Lining Up the Charges: Electricity Line Charges and ODV. Simon Terry Associates, Wellington. 60

Here again there are parallels with the German experience, discussed in Chapter 8. Mighty River Power is in the process of seeking planning permission to convert the mothballed oilfired Marsden B power station to coal. 62 http://www.windenergy.org.nz/FAQ/proj_dom.htm 63 http://www.windenergy.org.nz/ 61


233

Bertram, G. and Twaddle, D. (2005). Price–cost margins and profit rates in New Zealand electricity distribution networks: the cost of light handed regulation. Journal of Regulatory Economics, 27(3), 281–307. Caygill, D., Wakefield, S. and Kelly, S. (2000). Inquiry into the Electricity Industry: Report to the Minister of Energy. Ministry of Commerce, Wellington, June, http://www.electricityinquiry.govt.nz/reports/ final/index.html Cooper, K. (1995). Valuation techniques and problems: continuing education paper no 439. In proceedings of New Zealand Society of Accountants Infrastructure Assets Forum, Wellington. Easton, B. (1994). Economic and other ideas behind the New Zealand reforms. Oxford Review of Economic Policy, 10(3), 78–94. Electricity Corporation Establishment Board (1987). Electricity Pricing: An Invitation to Contribute. ECNZ, Wellington. Ernst and Whinney (1985). Policy Recommendations for Commercial Efficiency in the New Zealand Electricity Industry. Unpublished consultants’ report, Wellington, August. Ernst and Young. (1994). Rationale for Financial Performance Measures in the Information Disclosure Regime, Including Use of Optimised Deprival Values and Avoidance of Double Counting of Asset Related Expenses: A Report to the Energy Policy Group. Energy and Resources Division, Ministry of Commerce, by Ernst and Young for Briefing ESANZ. Unpublished consultants’ report, Wellington, August. Ernst and Young, Ewbank Preece and State Electricity Commission of Victoria (1991). Valuation of TransPower New Zealand Ltd, Stage 3: Asset Valuation. Wellington, September. Frater, P.R. et al. (1985). Summary Report of Energy Pricing Study and Technical Report of Energy Pricing Study, Business and Economic Research Ltd (BERL), Wellington, June. Hawke, G.R. (1969). Economic decisions and political ossification: The New Zealand retail electricity tariff. In P. Munz (ed.), The Feel of Truth: Essays in New Zealand and Pacific History. Wellington, A.H. & A.W. Reed, pp. 219–233. Hogan, W.W. (1992). Contract networks for electric power transmission. Journal of Regulatory Economics, 4(3), 211–242. Hogan, W.W. (1999). The Nodal Zone Debate Revisited, http://ksghome.harvard.edu/⬃whogan/ nezn0227.pdf Jackson, K.E. (1988). Government and enterprise: early days of electricity generation and supply in New Zealand. British Review of New Zealand Studies, 1, 101–121. Jackson, K.E. (1990). Electricity provision and the concept of service in New Zealand. In M. Trédé (ed.), Électricité et Électrification dans le Monde: Actes du Deuxième Colloque International d’Histoire de l’Électricité, Organisé par l’Association pour l’Histoire de l’Électricité en France. Presses Universitaires de France, Paris, July, pp. 411–419. Jarden and Co. (1985). Energy Pricing and Commercialisation. Jarden and Co., Wellington. Martin, J.E. (1998). People, Politics and Power Stations: Electric Power Generation in New Zealand 1880–1998. Electricity Corporation and Department of Internal Affairs, Wellington. Ministry of Energy Financial Objectives and Pricing Review Team (1984). Financial Objectives and Pricing Review for Ministry of Energy (Trading). Wellington. Murray, K. and Stevenson, T. (2004). Analysis of the State of Competition and Investment and Entry Barriers to New Zealand’s Wholesale and Retail Electricity Markets: Report Prepared for the Electricity Commission. LECG and TWSCL, Wellington, August. http://www.electricitycommission.govt.nz/pdfs/opdev/ retail/consultationdocs/pdfsconsultation/pdfscompetition/competition-report.pdf New Zealand Business Round Table (1985). Overview Report: Policy Recommendations for Commercial Efficiency in the New Zealand Electricity Industry. Wellington, August. New Zealand House of Representatives (1992). Report of the Commerce and Marketing Committee: Inquiry into the Proposed Increases of Wholesale and Retail Electricity Prices. Government Print, Wellington, February. New Zealand Treasury (1987). Valuation of Electricity Generation and Transmission Assets. Wellington. Rennie, N. (1988). Power to the People: 100 Years of Public Electricity Supply in New Zealand. Electricity Supply Association of New Zealand, Wellington. Ring, B.J. and Read, E.G. (1993a). Nodal pricing and extensions to the theory. Proceedings of the First ECNZ Optimal Generation Scheduling Workshop, Wellington. Ring, B.J., Drayton, G.R. and Read, E.G. (1993b). Transmission System Pricing Model: Record of Experimental Tests. EMRG Report to Trans Power, Wellington.

234


Scott, W.D. and Co. (1985). Key Issues Arising from Proposals to Change the Level of Electricity Prices for Bulk Supply in New Zealand. W.D. Scott & Co (NZ) Ltd., Wellington, March. Terry, S., Bertram, G., Dempster, I. and Gale, S. (1992). Hydro New Zealand: Providing for Progressive Pricing of Electricity. Electricity Reform Coalition, Wellington, March. University of Waikato Interfirm Comparison Unit (1985). Comments on Financial Objectives and Pricing Review for the Ministry of Energy. Hamilton, July. Videbeck, S. (2004). Economic Market Segmentation of an Electricity Pool. New Zealand Institute for the Study of Competition and Regulation, Wellington, http://www.nzae.org.nz/conferences/2004/89Videbeck.pdf

Chapter 8 Energy Policy and Investment in the German Power Market G. BRUNEKREEFT1 AND D. BAUKNECHT2 1 Tilburg Law and Economics Center (TILEC), Tilburg University, Tilburg, The Netherlands; 2OekoInstitut – Institute for Applied Ecology, Freiburg, Germany

Summary The authors study the investment incentives of energy policy in Germany and how this affects competition, the environment and supply adequacy. First, after a long period of “self-regulation”, the new Energy Act of 2005 installs a regulator and network regulation. Second, Germany has a strong agenda for the environment. Furthermore, the CO2 emission trading scheme has significant effects. Third, despite international debate, Germany does not have an explicit policy on generation adequacy. The key conclusions are threefold. The initial position of Germany to refrain from regulating network access did not work satisfactorily. The recent creation of regulation can be welcomed and expected to stimulate competition and generation investment. As elsewhere, CO2 permits are allocated free-of-charge, both for existing and new plants. This may be inefficient, but promotes new investment and thus benefits competition and generation adequacy. The data suggests that new generation investment will be required, but also that the market is active. Apart from wind and combined heat and power, coal seems to have a brighter future than sometimes thought. 8.1. Introduction The debate on electric power markets seems to be shifting toward the long-term perspective: does the market provide timely, adequate and efficient investment? Investment (covering both generation and network assets) affects competition, the environment and supply adequacy. In this contribution we analyze and discuss German energy policy with precisely this focus in mind. Evidently, German energy policy does not stand alone, but rather strongly relies on European policy. The policies we examine are threefold. First, the competition and network regulation of the electric power market, which changed direction in 2005. Whereas Germany initially took an exceptional position within Europe, it is now more in line with neighboring countries. Second, environmental policy and in particular the start of the CO2 emission trading scheme deserves attention. Germany has a strong agenda for promoting environmentally friendly technologies and energy efficiency. The question is how much scope remains for pursuing these objectives. Third, it is noteworthy that there is no policy on generation adequacy; in the light of experience and growing concern in other countries the question is whether this is justifiable. 235

236


The capacity margins are now declining but still comfortable, stemming as they do from excess generation capacity before liberalization. Investment levels have been picking up again in the last 2 years after a serious decline in the previous 5 years which followed on from an all-time high as a result of modernizing the former east following reunification in 1990. This may reflect better competitive opportunities indicated by higher wholesale prices. More worrying is that generation assets are old and need to be replaced, while it is unclear what should replace these. The share of coal and lignite in the generation mix is already large. The CO2 reductions required by the Kyoto protocol raise doubt about the future of coal. The share of nuclear is already 30% and is being phased out, while new build lacks support. Support for renewables (RES) is strong but it is unclear whether this has sufficient scope. Our key conclusions are also threefold. The initial exceptional position of Germany to refrain from (ex-ante, sector specific) regulating network access did not work satisfactorily. Clearly, the institutions were not in equilibrium. The recent creation of a regulator (the Bundesnetzagentur, BNA) and regulation (with the new Energy Act of July 13, 2005) can only be welcomed. For a variety of reasons, network regulation must be expected to promote competition and thereby stimulate new investment by newcomers. We note that the wholesale margin increases. As elsewhere, CO2 permits are allocated free-of-charge (instead of auctioned), both for existing plants as well as for new investment. Evidently this has a political background and cannot be supported on economic grounds. Still, a free-of-charge allocation does promote new investment, and thus it may be inefficient, but benefits competition and generation adequacy. A curiosity is the so-called transfer rule which grants new (CO2 poor) plant the number CO2 permits of the CO2 rich plant it replaces. This is good for the environment, but sets new entrants at a serious disadvantage. With respect to generation adequacy (taking into account the points made above), we note that data suggests that new investment is required, but also that (as elsewhere) there appears to be a lot of new construction plans. Apart from wind and combined heat and power (CHP), new coal seems to have, perhaps surprisingly, a brighter future than sometimes thought. The organization of this contribution is as follows. Section 8.2 provides an overview of the recent history and the current state of the German electricity supply industry (ESI). Section 8.3 gives an in-depth examination of various energy policies distinguishing between the Energy Act, environment policies and generation adequacy. Section 8.3 includes analysis of the investment effects. Section 8.4 is the conclusion.

8.2. The German Electricity Supply Industry 8.2.1. How the sector is structured With a population of 82 million, Germany has the largest power market in Europe. Total net electricity consumption is around 500 TWh/year; installed gross capacity is around 140 GW including more than 15 GW wind capacity – more than any other country in absolute terms. Peak load in 2004 was at 77.2 GW. Total gross revenues in the sector are roughly €60 billion/ year and investment amounts to around €4 billion/year. Germany’s transmission network is integrated with that of nine neighboring countries. Imports and exports are more or less balanced, with a total of 44 TWh imports and 51 TWh exports in 2004. The exchange with France, the Netherlands and the Austrian and Swiss hydro systems is especially significant (see Table 8.1).

237

Energy Policy and Investment in the German Power Market Table 8.1. German power imports and exports (in TWh). 2004

Austria Switzerland France Luxembourg The Netherlands Denmark Czech Republic Poland Sweden Total

2003

Imports

Exports

Imports

Exports

4.4 2.8 15.5 0.8 0.6 5.3 13.1 0.4 1.3 44.2

8.9 11.8 0.4 4.9 17.3 3.4 0.1 3.2 1.5 51.5

3.3 3.1 20.2 0.8 0.6 4.0 12.8 0.3 0.6 45.7

9.9 13.2 0.2 5.0 15.0 5.4 0.1 2.8 2.2 53.8

Source: VDN Berlin. 700 600 Other 500

Gas

TWh

Oil 400 300

Coal Lignite Nuclear

200

Hydro

100 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Fig. 8.1. Generation mix. Source: Brunekreeft and Twelemann (2005).

The main energy source for power generation is coal, which accounts for around 50% of electricity production, with hard coal and lignite each accounting for about half of this (Fig. 8.1). Germany has substantial domestic coal reserves. Yet German hard coal is about three to four times as expensive as imported coal and relies on state subsidies. Since the number of miners has become too small to be a serious interest group1 and the hard coal subsidy is a state aid and is for this reason not favored by the European Union (EU) commission,2 the subsidy is gradually reduced. This will not influence the generation mix or investment decisions, which are based on the price of imported coal, but will reduce domestic coal consumption and increase coal imports. Lignite is the only other major domestic energy source in Germany. As it can be accessed through open-cast mines, it is relatively cheap and does not require state subsidies.3 The 1

In 2003, there were just over 50,000 people working in hard coal mining and processing and around 15,000 in lignite. The number of employees is bound to decrease further. 2 Cf. EU Regulation, No. 1407/2002; July 23, 2002.

238


downside is that open-cast mines consume vast chunks of land, leading to significant public opposition. The RWE utility in the West and Vattenfall utility in the East are the main lignite producers and generators. While there was a major overhaul of plants in Eastern Germany after reunification, RWE operates a much older fleet of lignite plants, with quite a few plants approaching their 50th anniversary. Nuclear plants generate about one-third of the total power production. However, the red–green government coalition agreed on a nuclear phase-out program which was laid down in the 2001 amendment of the nuclear law. The agreement stipulates a generation limit based on a 32-year plant operation, which means that nuclear generation will be phased out at around 2020 according to this plan. However, companies have the option to shift generation allowances between plants to increase the output in more efficient plants. In mid-2005, only two plants: Stade and Obrigheim, were closed. The conservative party has announced that it wants to do away with this agreement and extend the plants’ lifetime. However, with the likely new conservative-socialist coalition it is unclear what will happen. There has been no “dash for gas” yet, with gas still accounting for only around 10% of power generation. There are only a few combined cycle gas turbines (CCGT) plants. With only minor domestic gas supplies and a high dependency on gas imports, mainly from Russia, the Netherlands and Norway, there is some concern that an increasing share of gas may undermine supply security. Despite this, generation from gas is forecasted to increase in most scenarios. For instance, a report for the Ministry of Economics (BMWA, 2005, p. 33) projects a share of gas of about 33% by 2030. The remaining power is generated from hydro plants and a rapidly increasing share of “new” RES, especially wind. The above-mentioned report for the Ministry of Economics also projects a share of about 33% for RES in 2030. Demand growth is relatively low at around 1% per year and is expected to remain at this level.4 CHP has a production share of about 10%,5 of which 60% is gas fueled and 40% hard coal (and lignite) fueled (Fig. 8.2). There are no official statistics on distributed generation, but the generation share is reported to be at around 18% (Wade, 2005). 160 CHP Power only

140 120 TWh

100 80 60 40 20 0

Gas

Hard coal

Lignite

Others

Fig. 8.2. CHP generation. Not including industrial plants. Source: Destatis. 3

There are no official and explicit subsides, yet it can be argued that there is hidden financial support (Wuppertal Institut, 2004). 4 Source: VDEW (www.strom.de). 5 This does not include industrial plants, which are often CHP plants.

Energy Policy and Investment in the German Power Market

239

Implementing the EU Directive of 1996, the German ESI was liberalized in 1998 with the Energy Act of 1998. The sector was never institutionally monopolized (like for instance the UK). Instead competition in a relatively unconcentrated and fragmented industry was excluded by cartels agreements, which were stabilized by legally enforced demarcation contracts. The main step of liberalization was to invalidate these cartel agreements after which the ESI fell under the authority of the Cartel Office and the Competition Act. The industry structure, which was artificially stabilized by the demarcation contracts, strongly reflected historical institutional lines; roughly speaking, generation and transmission reflected the position of the states within the federal structure of Germany, while distribution and retail reflected the strong position of the communities. Not surprisingly and as we have seen elsewhere, competitive pressure and commercial interests enforced significant changes in the industry structure after liberalization, most notably toward more concentration. Vertically, the ESI was strongly integrated between networks and commercial businesses and if anything this has increased since liberalization; it is thus remarkable that the Energy Act of 1998 only required minimal vertical unbundling requirements. We will discuss this in more detail below. The German ESI is strongly vertically integrated. There are basically two blocks, which is depicted in Figure 8.4. On the one hand, four predominantly privately owned big utilities own and operate the high-voltage transmission grids (plus the interconnectors) and most of the power plants (usually in their own control area). These firms are both transmission system operators (TSOs) and dominant generators. They operate the balancing market in their own control area. There is an ongoing discussion about separating these balancing markets and merging them into a national balancing market. Together these four companies own about 90% of total generation capacity (Table 8.2). Moreover, they also have majority shares in many distribution networks and retail activities. These four utilities are RWE, E.On, EnBW and Vattenfall Europe. On the other hand, a vast number of predominantly municipality-owned firms (Stadtwerke) own and operate the distribution networks and, as end user switching away from the incumbent retailer has been low, mostly the retail activities in the subsequent host areas.6 As we will argue below, and as also noted by Haas et al. (2005) the high degree of vertical integration led to cross-subsidization between networks and generation, stifling competition. Most of the municipal utilities were considered to be too small and expected to disappear quickly after liberalization. However, most of them have done much better than expected, setting up various alliances to defend their market position and to be able to take part in wholesale trading and realize economies of scale, for example, in billing. The vertically integrated “Stadtwerke” also responded to market opening by lowering the retail margins (being the difference between the end user price and network charge plus wholesale price), making life for new third-party retailers difficult. Cumulative domestic switching rates are reported to be 5%. Mergers and acquisitions have increased concentration in generation since the beginning of liberalization (see Table 8.2). Around 2000, two big mergers, creating the current firms RWE and E.On, pushed the Hirschman Herfindahl Index (HHI) to more than 2500.7 6

The exact number is unclear. VDN, the association of network operators, has 390 members. Also, VDN lists in its publication of network charges 700 different networks some of which fall under the same holding though. VDEW, another industry association mentions 900 firms. 7 In European merger control a post-merger HHI of 2000 and in the USA of 1800 are crucial thresholds. Note however that these are very rough indications, which neglect many details.

240


Berlin 7 2 4 Dortmund

3

Bayreuth 1

Stuttgart 3

1 EnBW Transportnetze AG 2 E.ON Netz GmbH 3 RWE Transportnetz Strom GmbH 4 Vattenfall Europe Transmission GmbH Fig. 8.3. German transmission system operators.

8.2.2. The institutional steps Germany implemented the 1996 EU Electricity Directive (EU E-Directive) with the Energy Act of 1998, of which three aspects stood out. First, full market opening. For generation this is unsurprising, but 100% end user eligibility from the start was exceptional in 1998. Even

241


Generation

4 Verbundunternehmen (VUs)

Transmission

~900 communal distributors

Distribution

Distribution

Retail

Retail

Fig. 8.4. A stylized representation of the ESI in Germany. Table 8.2. Market shares in generation (percentages of output). 1994

VEBA VIAG RWE VEW EVS Badenwerk HEW BEWAG VEAG Other Total HHI

} E.ON } RWE } EnBW } V’FALL

2000

A

B

Pre A

Pre B

Post

16.92 11.23 31.38 7.24 4.89 4.91 3.55 2.87 – 17.00 100 1807

13.96 8.27 28.42 6.65 4.30 4.32 2.96 2.28 11.84 17.00 100 1595

21.36 12.55 31.53 8.84

18.77 9.97 28.94 8.33

} 28.74

} 9.64

} 8.60

} 8.60

3.09 2.65 – 10.35 100 1903

2.57 2.13 10.33 10.35 100 1658

} 37.27

} 15.03 10.35 100 2622

HHI: Herfindahl Hirschman Index. Notes: The shares have been corrected for participation rates. Pre means Pre-merger and Post means Post-merger. V’Fall is Vattenfall. Source: Brunekreeft and Twelemann (2005, p. 103).

early 2005, only nine European member states had full retail market opening. And although the EU E-Directive 2003 aims at full retail market opening by 2007, we expect that there will be a debate on whether this will be pursued. Full retail competition in Germany worked well technically, but competition developed only slowly for domestic and small commercial end users. Second, whilst the degree of vertical integration of monopolistic networks and commercial businesses is high and increasing, the rules on unbundling were weak and were not enforced. Third, being the exception within Europe, Germany opted for negotiated Third Party Access (TPA), instead of regulated TPA.8 Negotiated TPA implied that, despite the monopolistic networks, the sector was left without sector-specific regulation and regulator. The government trusted the ESI to resolve 8

Cf. Haas et al. (2006) for a European overview. Moreover, Brunekreeft (2003, pp. 208 ff.) contemplates on possible explanations for this exceptional position; it is rather likely that the reunification in 1990 contributes to an explanation.

242


network access and network charges by voluntary negotiations controlled by the Cartel Office. Network access had to be arranged collectively in the so-called association agreements (VV). Initially, these arranged the (technical and administrative) rules but not the price of network access. Later, a set of accounting principles to calculate the network access charges was added to the VV. At no stage though was the precise level of network charges agreed upon or laid down. These were the sole responsibility of the individual network owners. Both the network access and the network charges were controlled by the Cartel Office. To facilitate this task, the Competition Act was strengthened with an essential facilities doctrine in 1998, which requires that access to the network should be provided on nondiscriminatory terms and at a fair and reasonable charge. This one clause in the Competition Act was the main regulatory instrument.9 Control was not strong and network charges were (and in fact still are) persistently high. The Cartel Office faced a number of problems (Bundeskartellamt, 2001). First, it is allowed to act only after a justified suspicion of abuse; hence, it can act only ex-post. Second, with up to 900 networks to be controlled, the Cartel Office was seriously understaffed for this task. Third, many of communal and regional networks enjoy political support from the states and communities. Fourth, the Competition Act is well suited to address discriminatory behavior; the more persistent problem, however, turned out to be the high level of the network charges, which is difficult to address with the Competition Act. Lastly, accounting according to the association agreement received legal validity, which in practice weakened the position of the Cartel Office. After a series of events and reports, the so-called Monitoring Report of the Ministry of Economics paved the way to stronger regulation in 2003. Parallel to this, the German government gave up its resistance in Brussels and the European Commission seized the opportunity to remove negotiated TPA from the directive. Hence, the new EU E-Directive 2003 exclusively allows regulated TPA. This is also the key development which led to the new Energy Act which entered into force on July 13, 2005. 8.2.3. Past, present and future Figure 8.5 plots an interesting development. Shortly after liberalization end user prices (net of taxes) fell strongly, but started rising again shortly afterwards and quite steeply since a year or so.10 The figure depicts the representative domestic user of Eurostat and relies on Eurostat data. This implies that it only captures the non-switching part of the market, which are still under the old tariff regime. It does not capture the prices of the competitors and thus the prices for switching end users. As shown elsewhere (Brunekreeft, 2003, p. 220), the best-practice alternative offers undercut the incumbent price severely at first, but then started to increase and converge. Meanwhile the difference is small. Further we should remark that the domestic market excludes the industrial users. The pattern for industrial prices is the same but far more dramatic. Industrial prices have fallen severely but are now increasing steeply as well. A steady increase of the electricity tax is an important contributor to higher prices. It has been raised gradually over the last 6 years to up to €2.05 c/kWh, which makes up somewhat

9

This is not unlike the situation in the USA under the Energy Policy Act 1992. Order 888 of 1996 made a strong move toward regulation of network charges (cf. Joskow, 2005a). 10 Cf. also Growitsch and Müsgens (2005) for more details.

243

Energy Policy and Investment in the German Power Market 17

c/kWh

16 15 14 13 12

Germany w/o tax EU 15 w/o tax

11

Germany with E-tax January 2005

January 2004

January 2003

January 2002

January 2001

January 2000

January 1999

January 1998

January 1997

January 1996

January 1995

January 1994

January 1993

January 1992

January 1991

10

ct/kWh

Fig. 8.5. Residential end user prices in Germany and Europe-15. Source: Eurostat data, various years. Note: This is for an average domestic end user; eurocent/kWh. Nominal prices. 8 7 6 5 4 3 2 1 0

VAT (16%) Concession fee RES CHP E-tax 1998

1999

2000

2001

2002

2003

2004

Fig. 8.6. Taxes in the German electricity price. Source: VDEW.

more than 11% of the total price. Further substantial parts are the communal concession fee and the federal value added tax; however, these are stable and do not explain the increase. While these taxes are substantial, they are unambiguous. More ambiguous are two levies induced by support for CHP11 and RES.12 These are not strictly speaking taxes, but the costs of CHP and RES are socialized over network users and electricity-consumers, respectively. It is clear from Figure 8.6 that whereas these costs are increasing steeply in relative terms, they are unsubstantial in absolute terms. Yet, the industry sometimes justifies the price increases, especially the increased network charges, with these “tax” increases. The development of the network charges is ambiguous. Up to mid-2005 the network charges were unregulated. At best, one could argue that rather loose self-regulation was enforced by either some threat of ex-post control by the Cartel Office, or by the threat of a change toward ex-ante regulation, which indeed happened in mid-2005. In the course of

11 12

Combined heat and power. Renewable energies.

244


65.00 55.00

/MWh

45.00 35.00 25.00

Peak Peak minus CO2 CO2 price CO2 coal markup

15.00

September 2005

July 2005

August 2005

May 2005

June 2005

April 2005

March 2005

January 2005

February 2005

December 2004

October 2004

November 2004

September 2004

July 2004

August 2004

May 2004

June 2004

April 2004

March 2004

5.00

Fig. 8.7. Wholesale prices and CO2 mark-up at EEX. Source: EEX.

self-regulation the association of network operators (VDN) started to publish a standard format of a sample of network charges for different voltage levels twice a year. Examination of the LV network charges reveals that they were high in international comparison (cf. e.g. EC, 2005) and high relative to end user prices (cf. Brunekreeft, 2002). However, they have been stable since at least 2002. Only the HV level has seen an increase of about 10%, which is unsubstantial in absolute terms. According to the network operators, this increase is just the cost-pass-through of higher balancing costs due to an increase in intermittent RES generation. A study of the association of industrial users, VIK13 suggests a recent increase in the network charges for industrial customers for selected networks. It is unclear whether the changes are representative for the sector. Growitsch and Wein (2004) calculate a reduction of the spread in network charges among various operators as a result of the introduction of the self-regulation in VVII⫹.14 This suggests that the increases by some are leveled out by decreases by others. All things considered, we should conclude that the recent increase in end user prices cannot be explained by changes in network charges. Much of the recent price increase can be explained by the wholesale price development. We note that more than 90% is traded “over the counter”, for which we do not know the prices. However, we assume that the spot price at the European Energy Exchange (EEX) in Leipzig is a sufficiently good indicator for contract prices. Figure 8.7 gives the EEX price development and clearly shows that the price is rising. The wholesale prices used to be very low. In fact, after liberalization the prices went down to almost short-run marginal costs and could not recover total costs. This is changing. With well over €35/MWh, it is believed that full cost recovery has been restored. Why the recent increase? There are three plausible explanations.

13 14

Cf. www.vik-online.de, April 28, 2005. VV abbreviates the German word Verbaendevereinbarung.

245

Energy Policy and Investment in the German Power Market 60

Demand

/MWh

50

Merit order stylized CO2: 20 /tonne

40

Price

30 Gas

20 10 0

Coal Hydro Nuclear 0

20

Lignite

GW

40

60

80

100

Fig. 8.8. The merit order and subsequent marginal prices in Germany with a CO2 price of €20/tonne CO2. Source: Bremer Energie Institut (this was handed to the authors by Wolfgang Pfaffenberger). 30.00 25.00

/MWh

20.00 15.00 10.00

Gas (th) Coal (th) Gas (el) Coal (el)

5.00 0.00 1998

1999

2000

2001

2002

2003

2004

2005

Fig. 8.9. Price development of gas and coal. Source: BAFA (efficiency gas: 59% and coal: 45%).

First, with the start of emission trading (see below) the electricity wholesale price now includes a CO2 mark-up. With a CO2 price of €20/tonne CO2 and an emission factor for a coal plant of 0.75 t/MWh, this amounts to a mark-up of €15/MWh on the wholesale electricity price. If gas is marginal, then the mark-up is about €7/MWh because the emission factor of gas plants is about 0.35. If generation is reasonably competitive then the CO2 markup would by and large be passed through into the EEX price. In either case, the effect of CO2 prices is substantial, as shown for real values in Figure 8.8. More importantly, Figure 8.7 suggests that if the CO2 mark-up is subtracted the net wholesale price fluctuates around €35/MWh, which corresponds by and large to the prices in 2004. Second, the fuel prices have increased as depicted in Figure 8.9. For Germany the increase in gas and oil prices has less influence on wholesale prices because gas is not often marginal, but the world coal price has increased recently as well. Third, for different reasons, the price increase may reflect a stronger control of the producers on the market, which at the very least seems to ease competitive pressure. Using an elaborate electricity market model, Müsgens (2004) makes an in-depth examination of costs, bidding and prices in the period from 2000 to mid-2003. He finds a structural break in early

246


2002. Before that prices were closely in line with system marginal costs, while from then on prices started to diverge from system marginal costs. Notably, the divergence is mostly in high demand periods. It does not follow though that these prices are excessive; as already noted, the early prices were too low to recover costs, whereas after early 2002, they might just recover full cost. Why would prices start to diverge from system marginal costs? Arguably, installed capacity in Germany was excessive, which was likely to increase competitive pressure and suppressed prices down to short-run marginal costs. Around 2000, the two big players, E.On and RWE announced that they would shut down some 10 GW of plant capacity because prices were too low; part of this was decommissioned and part mothballed. In addition, to other reasons a decline of excess capacity and the resulting relative scarcity are likely to have given the firms some grip on the prices.15 Alternatively, the higher prices might simply reflect emerging scarcity and actually signal that new investment is needed. Further, as shown in Table 8.2, concentration has increased around 2000 with the RWE–VEW (now RWE) and VEBA–VIAG (now E.On) mergers. With the HHI increasing from 1700 to 2600, theoretical insight and experience abroad would suggest a potential weakening of competitive pressure, as also suggested by Haas et al. (2005). Cross-border trade will certainly increase competitive pressure, but this is still limited. Table 8.1 shows that Germany is a net exporter, mainly because the prices in the Netherlands are higher than in Germany. Furthermore, cross-border capacity is only some 14% of total installed capacity. We would like to stress though that current wholesale prices (less of the CO2 mark up) do not seem to establish an abuse of market power.16 Lastly, in 2001 a report of the Cartel Office made clear that network charges were high (Bundeskartellamt, 2001). As a result the industry attempted to strengthen industrial selfregulation and started to publish the network charges systematically and in a comparable way. Pressure to regulate network access and network charges started to increase. The EU E-Directive 2003 removed negotiated TPA altogether and required regulation, which has now taken shape in Germany (see below). As argued extensively in Brunekreeft (2002, 2004), given vertical integration and the lack of effective regulation of network revenues, the rational strategy was to concentrate on the network while keeping the margins in the competitive businesses low and thereby retaining market shares. As regulation of the network takes shape we would expect the reverse, that is, an increase in both the wholesale and the retail margins. The implications for new entry and investment follow swiftly. After the German market was liberalized in 1998, foreign companies showed especial interest in entering the German market. Although the German generation market was by far not as attractive as, for example, the Spanish or Italien markets in terms of wholesale price, the need for additional capacity or expected demand growth, many companies considered it strategically important to be present in the largest European electricity market. The large number of small- and medium-sized utilities provided good take-over candidates. It took only a few years for this excitement to die down. This was partly because many new players like Enron or Dynegy either completely disappeared or were on the brink of disappearing. More importantly, the German market turned out to be hostile toward new

15

In formal terms, mothballing capacity can be interpreted as a credible commitment not to use this capacity and it can thereby ease competitive pressure. 16 This leaves the question open what exactly is market power and how prices above marginal costs can be stable in a competitive environment.


247

entry for a number of reasons. First, it was difficult to get hold of new plant sites. Second, as indicated above, wholesale prices were unattractively low. Third, in the first few years after liberalization the arrangements on network access were biased against third parties. Fourth, there have been persistent complaints about discrimination of third parties. Fifth, as far as new entry from gas-fueled CCGT is concerned, the big electricity–gas merger E.On-Ruhrgas (controversially approved in 2002) did not improve competitive conditions. Sixth, a gas tax represented a significant entry barrier for new gas plants. The gas tax increased the costs of a new CCGT plant by ca. €3/MWh or 10%. There were limited exemptions from the tax for new CCGT plants with an efficiency of more than 57.5%. Following a new European directive on energy taxation,17 this tax was abolished in 2005. Many new plant projects were either given up or had to look for new investors who were willing to keep these projects alive and wait for better times. New firms entered the market mainly through acquisitions; for instance EdF bought a minority stake in EnBW and Vattenfall took over Bewag, HEW and VEAG. This did not always improve competition, but rather increased the concentration in the wholesale market. While Germany has not seen many new plant projects that actually came on-line since market opening, investment activity is now picking up again as we will indicate in Section 8.3.3. The wider wholesale margin is definitely helpful in this respect.

8.3. Energy Policies and the Investment Effect Before discussing energy policies in Germany in detail, Table 8.3 gives an overview of the main acts and events as they affect the ESI. 8.3.1. The Energy Act 2005, regulation and the regulator BNA 8.3.1.1. The Energy Act 2005 As explained above, for a variety of reasons, the Energy Act 1998 was replaced by the Energy Act 2005. The key elements in the new Energy Act 2005 are the following. First, the rather artificial hybrid approach of ex-ante approval of the methodology and ex-post control of the level did not survive the debate, and a clear step has been taken toward ex-ante regulation of the network charges. Second, although starting off with a cost-based approach, it is an explicit intention of the authorities to switch to incentive-based regulation. Third, there will be a sector-specific regulator: the BNA. Fourth, the rules on unbundling are strengthened but they still only minimally fulfil the directive’s requirements. We will discuss these points in detail in this section.18 Ex-ante regulation of the network charges. Art. 23(2) of the EU E-Directive 2003 requires “fixing or approving, prior to their entry into force, at least the methodologies used to calculate” the network charges. The precise phrasing reflects the German wish to stick to ex-post control of the level of the charges. In fact, this type of ex-ante/ex-post hybrid regulation has been practiced in, for instance, Sweden (with mixed success) and Finland (where it worked well). After a long debate, it has been decided in Germany that the by-pass in the directive

17

Directive 2003/96/EG. Haas et al. (2005) are somewhat sceptical about the new Energy Act and point out that the legislator might have taken the opportunity to put in place a more pro-competitive market design. 18

248


Table 8.3. Major events in the German electricity supply industry. Date

Event

Comments

April 1998

Amendment Energy Act, start of liberalization

100% market opening in generation and supply.

May 1998

First Association Agreement VVI

Self-regulation of networks.

April 1999

Introduction of the electricity tax

Starting with 1.02 c/kWh and gradually increased to 2.05 c/kWh.

December 1999

Second Association Agreement VVII

New retailers (e.g. Yello) enter the market; prices drop severely and surprisingly, but only short lived.

April 2000

Renewable Energy Act

Fixed feed-in tariffs for renewables.

August 2000

EEX in Frankfurt

December 2001

New Nuclear Act

Stipulates phase out of nuclear plants in Germany, prohibits new nuclear plants.

April 2001

Report of the federal Cartel Office

Indicates that network charges are high and difficult to control by Cartel Office by Competition Act.

December 2001

Third Association Agreement VVII⫹

Stronger emphasis on industrial self-regulation.

2000/2001

Mergers

RWE and VEW: RWE VIAG and VEBA: E.On

2002

Merger

E.On and Ruhrgas

December 2003

Monitoring report by the Ministry of Economics

This report confirms “officially” that negotiated TPA in the ESI and GSI failed. It paves the way to regulated TPA.

January 2005

Emission trading starts

July 2005

Amendment of Energy Act, Regulatory Authority (BNA) takes over electricity network regulation

Implementing EU Directive, ending 7 years of self-regulation of the network.

GSI: gas supply industry.

be ignored and the ex-ante regulation of the level of the access charges be applied. Thereby the regulation of the network access charges is finally as it should be.19 There has been some debate about controlling price increases only. This would imply that all current levels would be beyond the authority of the regulator. Since some network operators have increased their charges quite significantly over the last year (i.e. before regulation would take effect), this restriction in regulatory authority was unacceptable and was overthrown. The regulator has now been authorized to look back and control the recent price increases. Cost-based versus incentive-based regulation. The main debate has been on the type of regulation. The formal current state is that regulation is cost based (para. 21), which reflects 19

Increases of domestic end user prices required approval of state authorities relying on a federal decree. The enforcement of this decree has always been questioned. In any case, as a result of the new regulation of the network charges, this decree on end user prices will expire in mid-2007.


249

business as usual. Previously, the “self-regulation” followed the accounting principles laid down in the association agreement. This is nothing else than a rate-of-return regulation, with the difference that it will now be enforced and resulting charges will have to approved before they enter into force.20 The legislator explicitly allows the option to switch to incentive-based regulation (para. 21a), which can be a price-cap or revenue-cap regulation. The regulator has been given the task to develop an incentive-regulation mechanism. However, whether, how and when this will be implemented is to be determined by the government in an ordinance (i.e. not by the regulator).21 The choice between cost-based and incentive-based regulation deserves attention. Incentive-based regulation aims at improving the incentives of the regulated firm to produce efficiently (i.e. cut costs). The means to do so is to allow the firm to keep the profits resulting from efficiency improvements. Not having to lower the (ex-ante allowed) prices after lowering costs for some predetermined period is the incentive. The German Energy Act explicitly mentions incentive-based regulation, but as has been pointed out by Joskow (1989, 2005b), it is not so clear what this means and how incentive-based differs from cost-based regulation.22 Three points seem important. The name rate of return regulation (being a typical form of cost-based regulation) suggests that prices should always be adjusted to costs so as to allow a reasonable rate of return. In practice this is not the case as rate-of-return regulation typically also fixes (weighted average) prices for some period of time: the regulatory lag. During this period prices can deviate from the “fair” rate of return. The difference in emphasis is that under typical cost-based regulation the regulatory lag is endogenous and relatively short. As Joskow (1974) explains well, typically in the USA, (weighted average) prices remained fixed until either the firm or the regulator requested a rate hearing. An important innovation of the incentive-based regulation first introduced by Littlechild in 1983 was to make the regulatory lag explicit and exogenous as a closed regulatory contract (cf. Beesley and Littlechild, 1989). A second point is that a cost-based approach typically adds a mark-up to the firm’s own costs. This is fair and reasonable but does little for the incentives to keep costs low. An incentive-based approach steps away from this and tries to avoid the use of the firm’s own costs as the benchmark. Instead it might use an industry benchmark. This retains the incentives but may lead to unreasonable results. These are theoretical polar cases; in practice the difference is blurred. Typically, with cost-based regulation, the regulator will look at whether the underlying costs are reasonable and thereby use comparators. Also, in incentive-based regulation any regulator will always check whether the outcome for an individual firm is reasonable.

20

The interested reader may refer to Brunekreeft and Twelemann (2005, p. 109) for details. The new Energy Act is a step away from the controversial accounting of replacement value. Presumably the practical background is that replacement value can lead to a high RAB whereas the network may have completely depreciated. 21 It may be noted as an aside that the policy uncertainty is striking: it is, at best, likely that regulation will be incentive based in the future, but we do not know when or what it will look like. This may be compared to Norwegian legislation where the switch to incentive-based regulation in 1997 was laid down in 1991 in the law. 22 To be precise, incentive-based regulation is the overarching term of which pure cost-pass-through and pure price-cap are the polar cases (cf. Joskow, 2005b). So, it is not really appropriate to contrast cost based with incentive based. We will assume that in the Energy Act incentive-regulation means a move toward price or revenue capping.

250


A third point is that pure cost-based or pure incentive-based mechanisms only exist in textbooks. In practice details matter and we find all kinds of adjustments and modifications and we see that the polar cases converge. Two examples are important. Rate-of-return regulation can be modified by a use-and-useful clause (cf. Joskow, 1989), which basically says that the investment costs may only be passed through in the rate base if the investment is used and useful, on which the regulator decides. Clearly this steps away from pure costbased regulation. Incentive-based regulation can be adjusted by profit-sharing rules, basically saying that if under the incentives-based constraint profits get either too large or too small, prices can (should) be adjusted. This clearly adds a cost-based element. Illustratively, Grout and Zalewska (2003) define a profit-sharing rule as a weighted average of the outcomes under cost-based and incentive-based regulation.23 The Energy Act (para. 21a.2) highlights the ex-ante determination of the average revenue cap as the decisive point. The control period will be between 2 and 5 years. Furthermore, relative efficiency will be determined by benchmarking with comparable firms.24 It should be noted though that this also holds for the current cost-based approach, where the reasonableness of the firm’s own costs can be checked by comparing with other firms. Moreover, only the costs components, which are under control of the firms will be subject to efficiency incentives. Although without explicit details, the Act touches upon the following aspects. First, presumably the price-cap regulation will be tariff basket, capping the weighted average price of a basket of products and leaving individual prices to the firms.25 Second, the regulation will explicitly be quality adjusted, presumably with a penalty-and-reward system. Third, it seems unlikely that there will be a yardstick; the Xi will be firm individual or firms will be collected in comparable groups. The discussion on X versus Xi is non-trivial in the face of up to 900 networks. Faced with so many firms, a yardstick X for all is very attractive but seems unreasonable. Bundesnetzagentur. The EU E-Directive 2003 requires with art. 23(1) regulatory authorities, “wholly independent from the interest of the electricity industry.” This excludes industrial self-regulation as it was practiced in the German ESI, especially by means of the VVII⫹. The new Energy Act creates the sector-specific regulator BNA, which will include the regulator for gas, telecommunications and postal services and which will also cover railways. Authority has been split though. The federal regulator BNA is responsible for all network operators with more than 100,000 customers (and for network owners with less than 100,000 customers that operate in more than one state). The states are in charge of regulating smaller network operators. However, if desired, the states can hand over the regulation to the federal BNA.26 This follows art. 15(2) in the EU E-Directive 2003, which exempts network operators with less than 100,000 customers from unbundling rules (except separate accounts). At least 500 networks are the responsibility of the states.27 As the communal lobby is very strong, and states and municipalities are the main stakeholders in the distribution network operators (DNOs), we may expect that state regulation of the DNOs will be weaker than federal regulation. 23

The distribution price control 2005–2010 in the UK provides interesting examples. There is some discussion to apply a virtual network approach as in Sweden to pre-select some very highly priced networks. 25 This stands out against the regulation of telecommunications, which has a stronger leg in the regulation of individual prices. 26 In Summer 2005, most states have decided to keep a state regulator. 27 However, the aggregate market covered by these firms will be small.

24


251

It seems that all the regulators will have to follow the same federal ordinance concerning the choice of regulation. This is a missed opportunity. As pointed out the idea is to switch to incentive-based regulation. One of the problems is how to manage the regulation of 900, mostly very small DNOs. Exactly this problem could be by-passed by applying different types of regulation: a strict incentive-based regulation at federal level and a “loose” costplus approach at state level for many small utilities. If all the small DNOs are also regulated by the same type of incentive regulation, it is unclear what is gained with splitting up the authority, while it opens the door for regulatory capture at state level. Unbundling. The unbundling requirements correspond to the EU Directive. Hence, the Energy Act requires legal (and functional and management) separation of TSO and DSO (with the art. 15(2) exemptions as mentioned above), confidentiality of information and accounting separation. This has by and large been implemented. The more urgent point is whether it will be enforced and controlled. As the regulator will pick up its task, we are confident that this will indeed be serious and that firewalls will start to be pressing. The more interesting question is whether ownership unbundling has any prospect. This question is aimed at the TSO in first instance. The legal problem is that the four TSOs are largely in private hands and that ownership unbundling is expropriation and violates the constitution. However, there are signs that ownership unbundling may re-emerge as an issue. First, there is some debate to split off the system operators (SO) from the rest of the firm and thus leave the transmission ownership (TO) to the current owners. This would also allow the creation of both one national SO and one national balancing market. The SO has no assets and this approach would therefore most likely not to be regarded as expropriation. Alternatively, all current firms could have the national SO in collective ownership. Second, experience in the UK, for instance, suggests that very strict firewalls can make “voluntary” unbundling an attractive option for the companies. Typically, this requires very strict monitoring by the “watchdog” and hence depends a great deal on the BNA. 8.3.1.2. The institutional disequilibrium Why did the government decide not to regulate from the beginning of liberalization? Recall that the German telecommunication sector does have sector-specific regulation by a regulator. Although speculative, four arguments are apparent. First, the legislator may not have been completely benevolent. The sector’s influence on politics is considerable. Second, the energy sector (gas and electricity) is considered to be strategic. Faced with counterparties like Gazprom, the government hesitates to fragment the industry too heavily and tries to balance between different goals (in this case, in particular between competition and countervailing power). Third, there has also been the desire to create and support “national champions” able to compete on a European scale. Fourth, after reunification the firms from the West committed to investing heavily in the former East in order to modernize both plants and networks. Oddly, this did not result in stranded-costs claims when liberalization started, unless we should interpret the lack of regulation as such. In any case, only 7 years after liberalization, the institutional framework was adjusted to adopt ex-ante regulation of the network charges. Hence, we may conclude that the framework was not in equilibrium and that something went wrong.28 Changes in the ESI are at least partly a spin-off of the gas supply industry (GSI). The GSI as well as the ESI was supposed to develop an association agreement for network access. 28

We have studied this in detail elsewhere (cf. e.g. Brunekreeft, 2002, 2003), and we will summarize it here briefly.

252


Whereas this by and large succeeded for the ESI, this failed in the GSI, leaving the government no option but to intervene. However, as this contribution is on the ESI, we will continue with an examination of the developments in the ESI.29 High network charges are against consumers’ interests, but as long as they are within a reasonable range they are unlikely to arouse too much political attention. More important is that competition died off after a first wave of excitement. Retail competition for domestic users is problematic. Switching rates are low and third-party suppliers are in financial distress. Although consumers perhaps do not switch because they are satisfied with their incumbent supplier and although potential competition may work, we observe that active competition is not a great success. Müller and Wienken (2004) estimate that roughly 40% of the household market is effectively closed, because the margin is below cost. The developments on the wholesale market were similar and highly remarkable (see above). Undoubtedly there has been quite strong competition, which in the beginning resulted in renegotiation of old contracts by large users (industry and retailers). The presence of competition and traders acted as a threat in the bargaining game. The prices for large users, which are an indicator for wholesale prices, came down strongly, presumably squeezing out the air resulting from productivity increases made in the 1990s and which had not been passed through. As shown in Figure 8.7 above, wholesale prices at the power exchange in Leipzig were very low; as low as fuel costs and substantially below cost recovery. This short-lived success has depressed entry: the first 6 or 7 years of liberalization have not or have hardly seen third parties in generation and most planned projects were never realized. If anything, firms left the market, while the assets in the market became more concentrated through mergers and acquisitions. The low entry activity reflects different issues. The low wholesale price, policy uncertainty about next institutional steps (regulation or not) and discriminatory behavior by the network operators will all have contributed to hesitant new entry. As we will discuss in Section 8.3.3, this is now changing. Weak regulation of a strongly vertically integrated industry (and weak enforcement of unbundling) implied difficult times for competition. Complaints about discrimination against third parties have been persistent. Indeed, the first association agreement was most certainly not pro-competitive. Moreover, during the first years after market opening, the Cartel Office has been active to settle unresolved issues and pursue abusive behavior. Moreover, the institutional framework of vertical integration without effective regulation of the network charges created the incentives for a margin squeeze: in case of doubt, the integrated firms will make (excess) profits on the network, not on the commercial business. The resulting low margins were unattractive for third parties. Summing up all the points above, we conclude that among other effects, effective regulation will widen the retail and generation margin (i.e. higher wholesale prices) and make abuse of the network or system-operation more difficult. All in all, effective regulation will promote new entry in generation and retail and thereby promote new investment. 8.3.1.3. Regulation and network investment Regulation and the choice between cost based or incentive based have potentially substantial effects on network investment. Incentive based may be good for short-run efficiency but may impede long-run network investment. Recall from Section 8.3.1.2 that the difference between cost based and incentive based is not clear-cut but rather a gradual matter of accents; the same applies for the reflections below. 29

The interested reader may refer to Brunekreeft and Twelemann (2005, Section 2.2) and references quoted therein for further details on the GSI.


253

It is not implausible that the explicit step of creating regulation and a regulator as already carried out by the Energy Act decreases (policy or regulatory) uncertainty. This can have a stimulating effect on investment. A second effect concerns the institutional choice for the BNA, which is part of the Ministry of Economics. Independence, being one of the leading regulatory principles, is thereby violated. How this could work out depends on the interests of the Ministry. Although it does have stakes, the federal Ministry is not a major shareholder in the power industry; the Ministry’s primary interest will be the consumer. This implies that the regulator might be under political pressure to lower the network charges if this is politically opportune.30 Other effects concern the choice between cost based and incentive based regulation. Though being still ambiguous in an empirical sense, cost-based approaches are seen as inefficient and generally wasteful of resources (gold-plating) and dependent on details biased toward over-capitalization; this was one of the drivers to move away from cost-based approaches (cf. Beesley and Littlechild, 1989, p. 456). The long-run perspective reverses the argument. Gold-plating may be inefficient but might be good for investment. Gilbert and Newbery (1994) point out that, in an uncertain world, the expected deviations from the reasonable outcome are smaller under cost-based regulation than under incentivebased regulation. Importantly, this increases the credibility of the regulator to stick to previously announced policies. In other words, incentive-based regulation can impede network investment as it reduces the regulator’s credibility. Peltzman (1976) pointed out the “buffering hypothesis”, which means that rate-of-return regulation reduces the firm’s exposure to market risk as compared to no regulation. Wright et al. (2003) extend the argument for price-cap regulation. In terms of demand uncertainty, risk under the price-cap regulation is lower than without regulation, similar to rate-of-return regulation. In contrast, in terms of cost uncertainty, risk under price-cap regulation is higher than without regulation. The arguments imply that the firm’s risk-adjusted cost of capital might be higher under price-cap regulation. All else equal, this means that investment may be lower under price-cap regulation. Lastly, as Spence (1975) pointed out incentive-based regulation has poor incentives for investing in quality. A price-cap regulated firm can increase profits at the expense of quality. Regulators in the Netherlands, Norway and the UK, for instance, have adjusted the price-cap regulation for quality incentives. The new Energy Act in Germany allows the BNA to make the necessary quality adjustments. As argued in Section 8.3.1.2, what we would expect is that if long-term network investment becomes more important relative to short-term efficiency, incentive-based regulation will be modified to cost-based type of regulation and quality-adjusted regulation. It appears that this is happening in the UK where the new distribution price controls which, came into force in April 2005, included sliding scales and used-and-useful test for capital overspending. 8.3.2. The policy on RES, CHP and the CO2 emission trading scheme German environment-related policy has the following targets: ●

30

Under the EU burden-sharing agreement to implement the Kyoto climate protocol, Germany has committed itself to reducing its greenhouse gas emissions by 21% between 2008 and 2010, as compared to the 1990/1995 emission levels.

This contrasts to telecommunications, where the government was the major shareholder of Deutsche Telekom, although its stakes reduced gradually by floating the shares. Furthermore, the situation is in contrast to the state level; the states are stakeholders in the power industry, and there we might see the directions go the other way around.

254

●

●


CHP generation is to play an important role in achieving these targets. In accordance with the national CO2 reduction strategy the ESI committed itself to reducing CO2 emissions through an increase of CHP generation by at least 20 million tonnes by 2010, which would mean nearly a doubling of CHP generation to 20% in 2010 compared to 2002 levels. According to the CHP Act, CO2 emissions are to be reduced through an increase of CHP generation (10 million tonnes by 2005; 23 million tonnes by 2010). The red–green government aimed at doubling the share of RES by 2010 from 5% to 10%.

8.3.2.1. Support for RES and CHP RES and CHP are supported by the Renewable Energy Act (EEG) and the CHP Act, respectively. The support mechanisms for EEG and CHP plants are basically a subsidy, although different in detail. While EEG plants get a fixed remuneration depending on technology and plant size, the payment for CHP plants varies with the market price. CHP in Germany has a production share of 10%, of which 60% is gas- and 40% coal- or lignite-fueled (Fig. 8.2). Support for CHP is a plain subsidy over and above the market wholesale price. The arrangement is the result of stranded cost compensation, after it turned out that CHP became unprofitable after liberalization. The CHP Act applies only to CHP plants that were in operation when the Act entered into force and will be phased out. The Act does only apply to new plants if they replace exiting plants (modernization) and for small CHP plants below 2 MWel and fuel cells. As a result, the CHP Act does little for investment which expands CHP capacity. RES are promoted by the RES Act (EEG), which arranges a feed-in charge system with a take-off obligation: a “take and pay” system. Like the CHP support now, the pre-liberalization system used to be a predetermined subsidy on the “market price”. With liberalization the market prices and thereby the feed-in charging fell substantially, pushing the RES plants into financial distress and suppressing new investment. The government decided to change the system by fixing the feed-in charges independent of market developments. The feed-in payments are generous, with a minimum payment of €5.5 c/kWh for wind and €43.4 c/kWh for solar. The support mechanism distinguishes between technologies, vintages and sites, thereby increasing overall efficiency. The costs of the feed-in mechanism are socialized over all end users. Whilst under the old pre-liberalization mechanism, each DNO had to bear the total costs of RES in their area individually, the EEG has established a mechanism whereby the costs are spread countrywide. The DNO, to which the RES plant is connected, is obliged to take-off the energy, but passes this on to the TSO (TNO) to which it is connected. The TSOs spread the burden equally among themselves and calculate a nationwide compensation charge. They then pass it on proportionally to the suppliers in their region, who in turn pay the compensation charge and pass through the costs into the end user price. In 2004, the share of RES was about 9% and the calculated compensation charge €9 c/kWh; with a wholesale price of €3.3 c/kWh, this amounts to a “RES tax” of €0.51 c/kWh.31 A “take and pay” system of feed-in charges and take-off obligation affects competition only in an indirect fashion. The system implies that RES and conventional sources do not compete directly. Indirectly, the conventional suppliers face reduced residual demand (which is total demand minus the exogenous supply of RES), which brings the price down. Also, we should expect that as the capacity-load margin increases (excess capacity), competitive 31

Compare Haas et al. (2006) for an overview and impression of European policies and experiences.


255

pressure increases, which further reduces prices. Moreover, under a system of fixed feed-in charges and take-off obligation the RES do not directly compete with each other. If the share of RES is moderate this is acceptable, but if the share is substantial a large part of the market will effectively be exempted from competition. If the RES share grows, the designs of the market, network regulation and a RES support mechanism might need to be reconsidered. Network connection charges are shallow, meaning that new generation assets only pay for the cost of connecting to the first network connection point, whereas the costs of network upgrades beyond this connection point are borne by the network operator. The network operators are obliged to connect new plant as long as the request is reasonable and, if necessary, undertake network enforcements. There is a new debate, however, about an estimated €800 million for an HV network upgrade which would be necessary to facilitate offshore wind projects. Evidently, the industry argues to pass through these costs. Plants connected to the distribution network (distributed generation), including CHP and RES, receive a network charge rebate. The calculation of grid charges is based on the assumption that all electricity is fed into the high-voltage transmission grid. The payment from the DNO to the TSO, however, is based on the actual annual peak load which the DNO gets from the TSO, reduced by a coincidence factor. As a result, if there are plants connected to the distribution network, the payment which the DNO receives from grid users may exceed the charges he has to pass through to the TSO. DG receives these avoided network charges. More generally though, while the support for RES does lead to more decentralized generation, there is no explicit policy on distributed generation yet. 8.3.2.2. CO2 emission trading Emission trading started at the beginning of 2005, as part of a European Union-wide emission trading scheme. The first trading period is a pilot phase, lasting until 2007. The second trading period will last from 2008 until 2012. Each EU member state had to draw up a national allocation plan, defining the overall emission targets for the various sectors (macro plan), including the targets for those sectors covered by the emissions trading scheme (ETS) (industry, energy), and the method of allocating CO2 permits to individual plants (micro plan). The allocation of permits to individual plants is based on two principles: grandfathering based on historical values for existing plants and a kind of benchmarking for new plants. Permits are allocated to existing plants on the basis of historical emissions multiplied by a reduction factor, whereas new plants receive the permits based on their expected emissions with an upper limit set by modern coal plants and a lower limit set by CCGTs. In both cases, permits are allocated free of charge. How will the allocation plan affect investment in new power plants and thereby the environment, generation adequacy and competition? The leading principle is that irrespective of whether the CO2 permits are auctioned or are free of charge, there is an opportunity cost corresponding to the market price of the permits, pushing up marginal costs. They will thus be passed through into the electricity wholesale price. If the permits are auctioned then evidently they are real (variable) costs. If the firms receive the permits free-of-charge they earn windfall profits equal to the quantity of permits times the price. The more relevant effect is on new investment. The CO2 price as such has a merit-order effect: it makes gas less expensive compared to coal in terms of marginal costs, which may increase the load factor of gas plants and thereby reduces their average costs. For an investment decision the windfall profit translates into lower investment cost. Brunekreeft and Twelemann (2005) calculate the entry price, which is the price at which a new investment just

256


recovers full cost.32 Receiving a number of permits works out as lowering fixed investment costs and increasing variable (opportunity) costs.33 Lower effective investment costs make it more likely that new plants will be able to compete against existing plants. For existing machines the lower investment costs are bygones and the windfall profits can be passed through to the shareholders. For new entrants it makes a difference in the investment decision. For this reason, the free allocation of permits stimulates entry with new investment. If the initial allocation is free of charge, money flows into the system. As long as entry is possible and rewarded with new permits, this leads to excessive entry and capacity. In the long run, the profitability of plants is brought back to a normal rate of return by (inefficiently) low load factors. At least initially new entry is likely to be more efficient with lower specific CO2 emissions and thus existing plants are likely to face the lower load factor; one would anticipate the early retirement of these plants. Although auctioning the permits is superior from an efficiency point of view, allocating the permits free of charge stimulates new investment (more and sooner) which is good for competition and supply security. The effect on technology is less optimistic. As soon as allocation deviates from “best practice” (product benchmark) and instead differentiates between different technologies (technology benchmark), the technology choice is distorted. The relative advantage new RES should have under a system of auctioned CO2 permits vanishes, if the permits are allocated free of charge and according to a technology benchmark. Furthermore, if the permits are allocated according to a technology benchmark (by and large corresponding to the emission rate of state-of-the-art machines), then replacing an old inefficient, high carbon machine with a new efficient, low carbon machine implies less permits, which in turn means smaller windfalls. At the very least, this postpones the replacement. Art. 10 of the German National Allocation Act specifies a transfer rule, which addresses this problem. If an old plant is replaced by a new plant, the permits of the old plant can be transferred to the new plant for 4 years. In an insightful study, Bode et al. (2005) argue that transfer rule heavily distorts competition as it puts new entrants (who cannot “replace old machines”) at a disadvantage: for the same investment an incumbent replacing its old machine would get more CO2 permits then an entrant not replacing an old machine. Somewhat surprising though, Bode et al. (2005) also claim that (with unlimited validity of the transfer rule) the transfer rule does not speed up replacement. This counterintuitive claim seems to be due to fact that the analysis lacks an explicit dynamic factor and thus a timing problem. Explicitly including a dynamic factor and timing (e.g. demand growth or costreducing learning) repairs this point and causes the transfer rule to speed up replacement. Overall we conclude that a system allocating CO2 permits free of charge (inefficiently) supports competition with new entry and generation adequacy. The effect on the environment is less clear. Having a CO2 system at all evidently supports the environment, but technology benchmarking may well have detrimental effects. The transfer rule is good for the environment but may damage competition too much.

32 There are two key numbers, reflecting the merit-order effect. For an CO2 price of less than €30/tonne CO2 the entry price is about €52/MWh due to a low load factor. With an CO2 price of more than €30/tonne CO2 the entry price is about €36/MWh, with a high load factor. The numbers are sensitive to assumed fuel prices and efficiency levels. Compare also Pfaffenberger and Hille (2003) for similar findings. 33 An alternative way of reasoning (leading to the same result) is to argue via the revenue side of net present value. Allocating the permits free of charge does not have an effect on expenses, but it does increase market price. Hence, the system will make new investment more attractive than it otherwise would be.


257

8.3.3. Generation adequacy 8.3.3.1. Hands-off policy on generation adequacy The power crisis in California in 2000/2001, the power black-outs in New York, London and Italy in 2003 and many near black-outs in recent years triggered concerns about the incentives of the liberalized power markets to provide adequate capacity. The overall issue is reliability, including both generation and networks. We concentrate on generation here; an impression of the network side has been given in Section 2.1.4. Reliability in turn covers two aspects: security and adequacy. Supply security is the ability of the system to respond to short-term disturbances; this requires sufficient reserve capacity and is typically the SO’ task. Supply adequacy (or, in our case generation adequacy) reflects sufficient long-term investment in such a way that the system functions under standard conditions; this begs the question as to whether the market provides sufficient incentives to invest. This is controversial and the impression is that policy-makers in the USA are less confident in the market than in Europe. A primary problem is fluctuating and uncertain demand (and of course, the fact that electricity power cannot be stored and the fact that supply should meet demand at all times), which implies that there will be peaking units with a very low load factor. In other words, the costs of a peaking plant should be recovered in only a couple of hours. If we assume, for instance, that a peaking plant has annualized costs of €40,000/MW/year, we need 10 hours per year and a price as high €4000/MWh to recover costs. If generation units are paid only for real production, then the prices are called energy-only prices. A system of energy-only prices is typically what the spontaneous market design will be. Theory predicts that scarcity will push up prices, which will attract new investment which in turn will reduce scarcity and so on, until an equilibrium is found. This is convincing, yet there are reasons to be cautious. Individual consumers cannot (at least not in current circumstances) be shut down individually; hence consumers have weak incentives to contract for (reserve) capacity. Two points weaken this argument. First, if large consumers can be shut down individually, the total market may be sufficiently responsive; the question is what is sufficient? Second, developments with so-called smart meters, which can be used to disconnect individual households, are fast. A further argument why markets may be slow to respond to scarcity prices is these very high prices are simply unrealistic (Joskow, 2003). Joskow and Tirole (2004) point out that even such very extreme situations and subsequent extremely high prices are very sensitive to the discretionary behavior of the TSO. Furthermore, most systems have a maximum price; in many parts of the USA, bids are capped at $1000/MWh. And even if they are not capped explicitly, there is justified concern that prices higher than this might trigger government interference. As pointed out in Brunekreeft and McDaniel (2005) this may be a vicious circle ending in a low-capacity equilibrium. We see the academic controversy reflected in policy, where there is a wide variety of policy measures. In the USA, a system with capacity obligations is popular. In Europe, concern has been expressed by the European Commission with its supply security package of December 2003. Some countries in Europe have explicit policies like capacity payments (e.g. Spain) or reserve contracting (Sweden and the Netherlands34). Most countries, however, have a hands-off policy: that is, explicitly doing nothing (except perhaps monitoring) and leaving it to the market (e.g. Norway and the UK). The German approach is also hands-off although it has not been made explicit. The background is more practical; Germany has a long tradition of excess capacity on which the system still relies and the investment question is not yet urgent. Para. 51 of the Energy Act requires the monitoring of supply security by the Ministry of Economic Affairs, and in the case of installed capacity (taking account of interruptible contracts) not being adequate, para. 53 then allows the government to organize a tender for additional capacity, in line

258

Electricity Market Reform Table 8.4. Age of German generation plants (in MW). Type Hard coal Lignite Nuclear Gas Oil Other

⬎30 years

30–10 years

⬍10 years

10,635 9570 2223 7291 4879 183

17,457 6207 21,340 6980 2044 1109

768 5465 0 3293 39 1853

Source: Ziesing and Matthes, “Energiepolitik”, DIW-Wochenbericht 48/2003. 9000 8000 Million /year

7000 6000 5000 4000 Generation T&D Total Total west Total east

3000 2000 1000 2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

0

Fig. 8.10. Investment in the German ESI. Source: Data from Karl, “Ifo Schnelldienst”, various years.

with art. 7 of the EU E-Directive 2003. It is safe to conclude that generation adequacy is not a policy issue in Germany yet. 8.3.3.2. Generation capacity and investment in Germany Is this justified? Basically we observe that (as elsewhere) both investment levels and generation reserve margins have dropped in the last 5–10 years. However, in recent years, both have been restored. The reserve margin in Germany has been studied closely in Brunekreeft and Twelemann (2005) and it seems that all that has happened is that excess capacity has been reduced without endangering continuity of supply. Yet at some point new investment is required. First, to meet new demand. Second, to adjust to technological progress (certainly eyeing the environment). Third, provided that phasing out goes ahead as planned, to replace the nuclear power plants which are to be decommissioned. Fourth, to replace old and depreciated machines. Table 8.4 below suggests that a large share of the generation plants in Germany is rather old and will need to be replaced soon. As mentioned before, though, investment activity is picking up.35 Figure 8.10 shows how investment fell steeply after the all-time high following reunification. But clearly, the fall halted 34

The system in the Netherlands has been designed but the required additional reserve capacity has currently been set to zero, and hence the system is inactive at the moment. 35 See, for example, the August 2005 investment survey among 200 industry done by ZEW (www. zew.de).


259

and in the meantime and investment is increasing again. Further, the increased wholesale prices make new projects attractive and indeed attract new entry. There are projects by companies like Statkraft, yet the most interesting development comes from mainly municipal distributors/retailers joining forces and investing in new generation plants. Examples are Trianel and SüdWestStrom. A reason which is sometimes heard, is to reduce the dependence on the big producers, from which we may conclude that competition has not been working all that well. A problem for new entrants has been the availability of sites for power plants. While completely new sites are difficult to find and often meet the resistance of the local public, existing sites are difficult to get hold of because they are in most cases controlled by the incumbents. For example, there was interest from municipal utilities in the south to build a CCGT plant on the Obrigheim site, a decommissioned nuclear plant. Yet EnBW who owns the site, refused to make it available for such a project, putting forward grid-related arguments. Lastly, the regulator and regulation should be expected to ease new entrants’ lives and as explained the CO2 ETS as well as the RES policy appears to support new investment. All in all, the German ESI may need new investment but it is likely to come. 8.3.3.3. New capacity: gas or coal? For several years after liberalization, gas-fueled CCGT was seen as basically the only option for new plants. In 1991, the European Commission lifted restrictions on the use of gas for electricity production. Gas prices were relatively low. With CCGT, new gas-fueled technology was highly efficient. The relatively low capital costs and short construction times and life duration made new gas an attractive investment in the liberalized market. However, coal and lignite are on the rise again, at least in terms of announcements and expectations. Even EnBW in the south of Germany ponders the possibility of building new coal plants, although not that long ago transportation costs in this region, which is far from both domestic and imported coal, were thought to be prohibitive. CCGT plants still benefit from high fuel efficiency and low capital costs, but the high gas price has turned against it. Gas projects have also suffered from the lack of gas network regulation and problems with third-party access, a situation that can be expected to be improved by the new Energy Regulator. This particular problem was presumably worsened by the controversially approved E.On-Ruhrgas merger, in as far as new entry was expected to be with gas-fueled plant. The European Commission is worried about a high European gas import dependence (from northern-Africa, Russia and the Middle-East). Alternatives are for instance to rely more strongly on liquefied natural gas (LNG) and more on indigenous sources like coal. On the other hand, as becomes clear from Figure 8.9 hard coal prices are increasing as well. Heavily increased demand from especially China has increased upward pressure on the coal price. There are also increasing costs of transportation, apparently especially due to the Chinese claim on shipping of steel. It is sometimes expected that the high coal price will not last, due, for instance, to exploitation of new mines and transportation capacity, while gas prices are expected to remain high. The CO2 ETS makes coal relatively more expensive compared to gas.36 However, the CO2 price must be rather high to have a significant effect. On the other hand, if fuel efficiency increases, ceteris paribus CO2 emission per MWh goes down. Although the same holds for 36 However, coal does not contain methane (CH4) which puts coal at a relative advantage if methane should be part of an emission scheme.

260


Table 8.5. Power plant projects in Germany. Project name/ location

Company

Plant type/ fuel

Size in MW

Comments

Hamm

Trianel

Gas CCGT

800

Under construction. 28 municipal utilities from Germany, Austria and the Netherlands have shares in this project, to go on-online in September 2007.

HuerthKnapsack

Statkraft

Gas CCGT

800

Project was taken over from intergen. Under construction.

Lubmin

Concord Power

Gas CCGT

1200

Irsching/ Ingolstadt

Eon/N-Ergie/ Mainova

Gas CCGT

800

Herdecke

Mark-E/Statkraft

Gas CCGT

400

Lingen

RWE

Gas CCGT

800–1000

Boxberg

Vattenfall Europe

Lignite

700

Additional capacity, not replacing old lignite plants.

Neurath

RWE

Lignite

1100 or 2200

Replacing old lignite units.

Karlsruhe/ Heilbronn

EnBW

Hard coal and/or gas

SüdWest-Strom

Hard coal or gas

400–800

No final decision on fuel yet, coal would have relatively high transport costs, municipalitybased company.

Trianel

Hard coal

700–800

As with the Trianel CCGT project, municipal utilities can buy shares in this project. Replacing a 300 MW coal plant.

No final decision on fuel yet, coal would have relatively high transport costs.

Datteln

Eon

Hard coal

1000

Hamm

RWE

Hard coal

2 * 750

Hamburg

Vattenfall Europe

Hard coal

700

Source: Company information, various sources.

CCGT, rather strong technological advances are expected in more efficient coal and lignite plants (supercritical plant and clean-coal technology) (cf. e.g. Bode et al., 2005). While both hard coal and gas can be bought on the market and are potential options for new entrants, lignite is a different game. As it is too expensive to transport lignite over long distances, due to its low energy density, lignite plants are generally located right next to the mine, the mines are owned by the generators and the fuel is shoveled from the mine into the plant without going through any form of market. Consequently, the marginal costs of lignite plants are anyone’s guess. For new entrants, there is no way they can get access to lignite and new lignite plants will be built by the incumbents. Currently, it is mainly RWE that is about to replace its old plants by new ones. We conclude that the future of coal as a fuel for the German ESI looks brighter than is sometimes thought. Looking at projects under construction and announced projects from mid-2005, gas and coal have about the same share (see Table 8.5).


261

8.4. Concluding Remarks This contribution examines energy policy in Germany. The primary focus is on the effect of various policies, which directly or indirectly relate to the energy market, on investment in the ESI in Germany. Investment in turn affects competition, environment and supply adequacy. The policies we examine are threefold. First, we study the policy related to liberalization and regulation of the ESI. On July 13, 2005 a new Energy Act entered into force implementing the EU Directive of 2003. The key point of the new Energy Act is to remove negotiated TPA and instead establish regulated TPA. It can be concluded that the previous system, which did not have effective regulation did not work. Network charges are high both in international comparison and relative to the end user price. Competitive margins are low, which impedes effective competition. The new system installs a sector-specific regulator (BNA) and regulation. The regulation is as yet cost based, but the Act explicitly allows the option to switch to ex-ante, incentive-based regulation. In practice this means a shift of emphasis toward ex-ante, forward-looking capping of revenues for a predetermined period and presumably a stronger reliance on benchmarking of different firms. As set out extensively in this contribution, we expect that the regulation of network access will strongly support the development of competition in both generation and retail; we already observe that investment activity in generation assets is starting to take off. On the other hand, it can be argued for a variety of reasons that the shift toward incentivebased regulation, which aims at short-run efficiency, tends to have detrimental effects on (long-run) network investment. This can be defended though, because currently the networks are viewed as being in good shape albeit inefficient and the Energy Act does allow quality-adjusted regulation. Second, Germany has a strong tradition of supporting the environment. The policies on RES, CHP and the CO2 emission trading are dominating the debate at the moment. The support schemes and network connection arrangements for RES and CHP, although with different background, are generous and should be expected to support further new investment. The costs of the schemes are passed through to end users and network charges, respectively. Examination reveals that the numbers are too small to make RES and CHP responsible for the recent increase of end user prices. The CO2 price is surprisingly high and as the German power production relies on coal the CO2 mark up in Germany is high as well. Exactly this seems to explain much of the recent price increase. The system of allocating the CO2 permits free of charge, whilst inefficient, stimulates new investment and thereby promotes competition and supply adequacy. Oddly, as a consequence of having a technology benchmark, the new investment need not be in clean technology. Third, despite controversial debate on generation adequacy elsewhere, Germany has no explicit policy on generation adequacy. Leaving the theoretical question as to whether the energy-only market will provide sufficient capacity aside, we observe that capacity margins and investment levels have dropped in the last 6 or 7 years, but have been restored recently. Moreover, generation assets in Germany are old and replacement and modernization are required soon. At the same time, we observe that investment activity (at least announced) is definitely picking up again. Challenging the conventional wisdom that gas will dominate the future, it seems that hard coal has a brighter future than sometimes thought.

Acknowledgment The authors would like to thank Paul Joskow, Wolfgang Pfaffenberger and Perry Sioshansi for useful comments and remarks.

262


References Beesley, M.E. and Littlechild, S.C. (1989). The regulation of privatized monopolies in the United Kingdom. Rand Journal of Economics, 20(3), 454–472. Bergman, L., Brunekreeft, G., Doyle, C., von der Fehr, N.-H., Newbery, D.M., Pollitt, M. and Régibeau, P. (1999). A European Market for Electricity? Monitoring European Deregulation 2. CEPR/SNS, London/ Stockholm. BMWA (2005). EWI/Prognos-Studie: Die Entwicklung der Energiemärkte bis zum Jahr 2030. BMWA Dokumentation Nr. 545, May 2005, Berlin. Bode, S., Hübl, L., Schaffner, J. and Twelemann, S. (2005). Oekologische und wettbewerbliche Wirkungen der Uebertragungs- und der Kompensationsregel des Zuteilungsgesetzes 2007 auf die Stromerzeugung. HWWAReport 252. Brunekreeft, G. (2002). Regulation and third party discrimination in the German electricity supply industry. European Journal of Law and Economics, 13(2), 203–220. Brunekreeft, G. (2003). Regulation and Competition Policy in the Electricity Market; Economic Analysis and German Experience. Nomos, Baden-Baden. Brunekreeft, G. (2004). Regulatory threat in vertically related markets: the case of German electricity. European Journal of Law and Economics, 17(2), 285–305. Brunekreeft, G. and McDaniel, T.M. (2005). Policy uncertainty and supply adequacy in electric power markets. Oxford Review of Economic Policy, 21(1), 111–127. Brunekreeft, G. and Twelemann, S. (2005). Regulation, competition and investment in the German electricity market: RegTP or REGTP. Energy Journal, Special issue, 99–126. Bundeskartellamt, (2001). Bericht der Arbeitsgruppe Netznutzung Strom der Kartellbehörden des Bundes und der Länder. 19 April 2001, Bundeskartellamt, Bonn. Clingendael Institute (2004). Study on the Energy Supply Security and Geopolitics. Report prepared for DG TREN, January 2004, The Hague. EC (2005). 4th benchmarking Report. European Commission, DG TREN, Brussels. Gilbert, R.J. and Newbery, D.M. (1994). The dynamic efficiency of regulatory constitutions. Rand Journal of Economics, 25(4), 538–554. Grout, P.A. and Zalewska, A. (2003). Do regulatory changes affect market risk? LIFE Working Paper 03-022, LIFE, Maastricht University, Maastricht. Growitsch, C. and Müsgens, F. (2005). The economics of restructuring the German electricity sector. Mimeo, University of Cologne, Cologne. Growitsch, C. and Wein, T. (2004). Negotiated third party access – an industrial organization perspective. Working Paper No. 312, University of Lueneburg. Haas, R., Glachant, J.-M., Keseric, N. and Perez, Y. (2006). Perspectives for long term competition in the continental European electricity market. In F.P. Sioshansi and W. Pfaffenberger (eds.), Electricity Market Reform: An International Perspective, Elseviers Scientific, 2006 (forthcoming). Joskow, P.L. (1974). Inflation and environmental concern: structural change in the process of public utility price regulation. Journal of Law and Economics, 17(1), 291–327. Joskow, P.L. (1989). Regulatory failure, regulatory reform, and structural change in the electric power industry. Brookings Papers on Economic Activity; Microeconomics, Vol. 1989, 125–208. Joskow, P.L. (2003). The difficult transition to competitive electricity markets in the U.S. MIT, Boston, Mass, May 2003. Joskow, P.L. (2005a). Transmission policy in the United States. Utilities Policy, 13(2), 95–115. Joskow, P.L. (2005b). Incentive regulation in theory and practice: electricity distribution and transmission networks. Mimeo draft August 8, 2005, MIT, Boston. Joskow, P.L. and Tirole, J. (2004). Reliability and competitive electricity markets. Working Paper CMI EP 53, University of Cambridge, Cambridge. Müller, Chr. and Wienken, W. (2004). Measuring the degree of economic opening in the German electricity market. Utilities Policy, 12(4), 283–290. Müsgens, F. (2004). Market power in the German wholesale electricity market. EWI Working Paper, No. 04.03, EWI Cologne.


263

Peltzman, S. (1976). Toward a more general theory of regulation. Journal of Law and Economics, 19(2), 211–240. Pfaffenberger, W. and Hille, M. (2003). Zukünftige Energieoptionen: Sicherung der Investitionen in die Elektrizitätsversorgung. Report, Bremer Energie Institut, Bremen. Spence, A.M. (1975). Monopoly, quality and regulation. Bell Journal of Economics, Vol. 16, 417–429. Wade (World Alliance of Decentralised Energy) (2005). World Survey of Decentralised Energy 2005, Edinburgh. Wright, S., Mason, R. and Miles, D. (2003). A Study into Certain Aspects of the Cost of Capital for Regulated Utilities in the U.K. Report, February 13, 2003, Smithers & Co, London. Wuppertal Institut (Hg.) (2004). Braunkohle-ein subventionsfreier Energieträger? Kurzstudie. Wuppertal.


Chapter 9 Competition in the Continental European Electricity Market: Despair or Work in Progress? REINHARD HAAS,1 JEAN-MICHEL GLACHANT,2 NENAD KESERIC1 AND YANNICK PEREZ3 1

Institute of Power Systems and Energy Economics, Vienna University of Technology, Vienna, Austria; Head of the Department of Economics, University Paris XI, Cedex, France; 3Department of Economics, University Paris-Sud 11, Cedex, France

2

Summary This chapter examines the perspectives for competitive electricity markets in Continental Europe. In most Continental European countries restructuring of the electricity market started in the late 1990s and is still going on. The object of this chapter is to investigate past developments in this market and to analyze which conditions are necessary to enhance competition in the long run. Currently, the major obstacle for one common European electricity market is a general lack of competition in virtually all local and national wholesale as well as retail electricity markets because the number of competitors is too low, or because barriers to entry and incentives to collude remain too high. Our major conclusion is that several conditions are necessary to bring about effective competition in the Continental European electricity market: (i) a complete separation of ownership of the transmission grid and the generation and supply in all countries and submarkets; (ii) sufficient transmission capacity for creating a larger market; (iii) adequate margins in generation capacity; (iv) a sufficiently large number of generators to share this capacity; (v) a secure and competitive supply with primary fuels (notably natural gas). As it is not likely that these conditions will be fulfilled the prospects for a vibrant competition in Continental Europe are bleak.

9.1. Introduction The restructuring of electricity markets in most Continental European (CE) countries started in the late 1990s, and is still going on. This process was triggered by the European Commission (EC) directive, 1996, “Directive for a common electricity market”. The major 265

266

Electricity Market Reform European electricity sub-markets

2

1

6

5

1. UK & Ireland 2. Nordiccountries 3. Iberian peninsula 4. Italy 5. Western Europe 6. Eastern Europe 7. South-Eastern Europe

7

3 4

Fig. 9.1. Electricity sub-markets in Europe in 2005.

motivation for this directive was the EC’s conviction that liberalization, price deregulation and privatization would directly lead to competition in generating, as well as supply which would then result in lower prices for the whole of Europe. The EC’s main expectation in the directive was the belief that “market forces (would) produce a better allocation of resources and greater effectiveness in the supply of services”.1 In June 1996, after years of discussion, the European Council of Energy Ministers reached an agreement with the European Parliament on a market liberalization directive, and 6 months later passed the full directive concerning common rules for the internal market in electricity which, with the intention of restructuring the European power industry, became law in February 1997. The initial intention of the EC was the creation of a common European electricity market, but this area still consists of at least seven different sub-markets which are separated by insufficient transmission capacities, and differences in conditions for access to the grid (Fig. 9.1). Furthermore “the evidence from Europe and the USA suggests that there are a number of conditions for successfully liberalizing electricity markets.” (Newbery, 2002; Glachant and Finon, 2003). We have identified conditions which could bring about a common competitive European electricity market, thus leading to competitive electricity prices. They are: ●

●

●

1

access to the grid, this requires unbundling of generation from transmission, and of supply from distribution; supply adequacy, adequate capacity in generation and transmission as well as access to primary energy sources (e.g. natural gas); market structure, ownership and number of generators and suppliers;

EC Communication Services of general interest in Europe, OJ C 281, 26. September 1996, p. 3.

Competition in the Continental European Electricity Market

● ● ●

267

design of the market place, notably the ease of entry for new players; regulatory governance; environmental issues, which are playing an increasingly important role.

The goal of this chapter is to analyze the evolution of the European electricity markets and to discuss future developments with respect to competition (see former treatments in Bergman et al., 1999; de Paoli, 2001; Glachant and Finon, 2003; Jamasb and Politt, 2005; Glachant and Lévêque, 2005; as well as the special issue of the Energy Journal, 2005). This chapter covers most of what is currently called “Continental Europe” (CE): Austria, Belgium, Czech Republic, France, Germany, Hungary, Poland, Portugal, Slovenia, Slovakia, Spain and Switzerland. It is organized as follows: Section 9.2 provides the historical context with major data and facts for the liberalization of the CE markets. Section 9.3 describes EC and national governments’ market liberalization initiatives. Section 9.4 presents major changes country by country. Section 9.5 discusses the evolution of the markets corresponding to government initiatives. Section 9.6 describes the market’s remaining problems; followed by conclusions in Section 9.7.

9.2. Background: Facts, Figures and History Before 1990, almost every electricity supply industry in Continental Europe was vertically integrated with a captive franchise market, either state owned (the majority of cases) or under price-regulated mixed private/public ownership (as in Belgium, Germany and Switzerland) (see Chapter 1). Regulated area monopolies prevailed in all countries. Until the end of the 1990s, the standard model was “an effectively vertically integrated franchise monopoly under either public ownership or cost-of-service regulation” (Newbery, 2005). Although electricity networks were typically synchronized over wide areas, interconnections of areas under different transmission system operators (TSOs) were frequently guided by security rather than economic considerations. However, most trade in the past was due to economic benefits of arbitrage during peak load hours. Real electricity liberalization in Europe started with Britain’s restructuring and privatization in 1990, demonstrating that vertical unbundling and the creation of wholesale electricity markets was actually feasible (see Newbery, 2005). Jamasb and Pollitt (2005) argue that the centralized approach to market liberalization because of European Electricity Directives has succeeded in maintaining the pace of reform in the original EU-15, and in a number of associated and accession countries, and, as well as achieving a certain degree of standardization of structures, institutions and rules in national markets. However, the problems created by initially concentrated market structures have been reinforced by a wave of subsequent mergers, and the low level of interconnection that reduces the scope for fostering competition by imports (Glachant and Lévêque, 2005). Yet, ownership structures and degree of vertical integration were quite different among the following countries: ●

● ●

In France, Italy, Portugal, the former Czech-Slovak Republic, Poland, Hungary and Slovenia a strong state-owned vertically integrated monopoly dominated the ESI. This centralized structure typically led to a single dominant player, such as Electricité de France. In Spain and Switzerland, vertical integration was strong, but with a handful of companies. In Germany there were about 10 generators integrated with transmission, but only partially integrated with supply.

268

●

●

●

●


In Austria there was one large generator which was integrated with transmission, and about 14 regional suppliers fully integrated with distribution. In the Netherlands there was an upward vertical integration with the distribution companies controlling the grid and the generators. In Belgium, the large majority of the power sector has been private for decades. The private generator Electrabel is supervized and controlled by the mother company, Tractebel. Belgium, Germany, Spain and Switzerland were the only countries in the mid-1990s where private ownership among generators prevailed (tempered in Germany and Switzerland by the local public ownership of distribution and supply, and the former “State enterprise” nature of Endesa in Spain). It contrasted with the state-owned enterprises in France, Italy, Portugal and the remaining Central and Eastern countries.

9.2.1. Development of demand and supply About 2300 TWh were consumed in the CE area in 2004. The largest electricity markets are currently in Germany, France, Italy and Spain. Highest per capita demand was in Luxemburg, Belgium and Switzerland. The lowest per capita demand was in Poland, Hungary, Portugal and Slovakia. Demand growth per year was strongest in Spain (⫹5.0%), Portugal (⫹4.9%), and Austria (⫹3.1%). In Poland and Germany demand increased by about only 1%/year. In the whole of the CE, electricity consumption grew from 1% to 3% per year between 1999 and 2004. Details are depicted in Figures 9.2–9.4.

Poland The Netherlands (145 TWh) (111 TWh) Germany NL (554 TWh) Czech Republic Belgium (61 TWh) Slovakia (88 TWh) Austria CZ (26 TWh) (52 TWh) France Hungary CH Switzerland (445 TWh) SI (38 TWh) Slovenia (65 TWh) (12 TWh) Spain (235 TWh)

Italy (322 TWh)

Portugal (46 TWh)

Fig. 9.2. Electricity consumption in CE countries in 2004.

269


In 2004, generation in CE countries was mainly from fossil thermal power plants (mainly coal) with 51%, followed by nuclear (34%) and hydro (12%). Other renewable (mainly wind) have contributed 3%. As shown in Figure 9.5 the distribution of generation sources across CE countries is rather uneven. In most countries thermal power dominates, in Italy and the Netherlands with more than 80%, in Poland almost 100%. In France, Belgium and Slovakia nuclear power plays the most important role. Only in Austria and Switzerland does hydro power prevail. Specific electricity consumption CH ES SK SI PT PL NL LU IT HU DE FR CZ BE AT 0

2000

4000

6000 8000 kWh/cap

10000

12000

14000

Fig. 9.3. Comparison of electricity consumption per capita in CE countries in 2004. Demand increase (2000–2004) CH ES SL SK PO PL NL LU IT HU DE FR CZ BE AT 0.0

1.0

2.0 3.0 Average (%/year)

4.0

Fig. 9.4. Growth of electricity demand in CE countries (average 2000–2004).

5.0

270

Electricity Market Reform Generation mix of continental european countries (TWh in 2004) 100 90 80 70

%

60 50 40 30 20 10 0 AT

BE

FR

DE Hydro

IT

LU

Nuclear

NL

PO

Thermal

ES

CZ

HU

Renewables

PL

SK

SL

CH

Others

Fig. 9.5. Comparison of the fuel mix for generation in CE countries in 2004. Source: UCTE (2005).

9.2.2. Generation capacity and load Capacity margin is different among countries as can be seen from Figure 9.6 . However not all gross capacity are available for generation. This is especially true for hydro capacity (Austria, Spain) and old fossil plants (Italy). For example, Italy, Austria and the Netherlands which are net importers of energy also exhibit such an apparent excess capacity margin. Figure 9.7 depicts the evolution of generation capacity over the last 10 years in CE. The growth in capacity is mainly from wind power and fossil fuel power plants. 9.2.3. Development of imports and exports In 2004, the total amount of electricity exchanges between CE countries stood at about 300 TWh. This is equal to about 13% of consumption and is frequently limited by constrained cross-border transmission capacity. Figure 9.9 show the physical2 electricity exchanges between CE countries. France is the biggest net exporter among CE countries with net exports of almost 67 TWh, followed by Czech Republic and Poland. The major importing countries are Italy with 51 TWh followed by The Netherlands and Hungary with 17 TWh and 7 TWh, respectively. The percentage of imports and exports of total electricity demand in CE countries is depicted in Figure 9.8. Smaller countries like Switzerland, the Czech Republic, Slovakia and 2

To some extent these flows are not due to contracts between countries but just because of loop flows (e.g. from Germany to Poland to Czech Republic back to Germany).

271

Competition in the Continental European Electricity Market Installed gross capacity in CE countries versus peak load in 2004 130 Others Renewables Thermal Nuclear Hydro Available capacity Peak load 2004 (MW)

120 110 100 90 GW

80 70 60 50 40 30 20 10 0 AT

BE

FR

DE

IT

LU

NL

PO

SE

CZ

HU

PL

SK

SL

CH

Fig. 9.6. Installed gross and net generation capacity (except auto producers) and maximum load in CE countries 2004.

Historical development of gross generation capacities in CE 600,000 550,000 500,000 450,000 400,000

MW

350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 1995

1996 Hydro

1997

1998

Nuclear

1999

2000

Thermal conv.

2001

2002

Renewables

2003 Others

Fig. 9.7. Evolution of generation capacity in CE 1995–2004. Source: UCTE, national reports.

2004

272

Electricity Market Reform Percentage of imports/exports CH ES SL SK PO

Imports

Exports

PL NL LU

⫺108%

IT HU DE FR CZ BE AT

⫺60

⫺40

⫺20

0

20

40

60

Percent of total consumption Fig. 9.8. Imports/exports as percentage of electricity demand in CE countries. Source: UCTE (2005)

Luxemburg, with around 20% of net exports related to domestic consumption, transfer the largest shares of their electricity to and from other countries. Figure 9.9 shows the evolution of imports and exports since the start of liberalization. There is a trend toward a slight increase, but no dramatic boost. 9.2.4. Past and current issues of transmission The bulk of the transmission and distribution network in Europe was built between 1950 and 1990 prior to the introduction of market liberalization and has had few additions in recent years. Figure 9.10 shows the highest percentages of net transfer capacity (NTC) used in 2004 between CE countries.3 Due to the operating complexity of the European meshed network, commercial capacity and physical capacity differ. Hence, the interconnection capacity is defined by European transmission system operator (ETSO) as “NTC”. The most congested lines are between Italy and its neighboring countries; and between Spain and Portugal. But the borders between Germany, Austria and the Czech Republic are next. In principle the congested lines need a special mechanism so as to be managed in an economic way (see Section 9.5.6). The existing CE network was built to guarantee a good level of technical reliability and to give some room for managing peak load problems. Now, it is supposed, it is to be used in a more economic way, under optimization processes of scarce capacity, and to produce price convergence in a single European market perspective. 3

Further details are documented in Table 9A.3 in the Annex. It shows current cross-border transmission capacity (Net Transfer Capacity, NTC), as published by the European Transmission System Operators (ETSO) for winter 2004/2005, physical flows 2004 and maximal possible (theoretical) annual energy flows between the countries.

273

Competition in the Continental European Electricity Market Electricity net imports in CE (1998–2004) 100 Austria

90 Spain

Spain

Luxembourg Austria Austria Luxembourg Portugal Spain Luxembourg Luxembourg Luxembourg Hungary Hungary Portugal Hungary Portugal Spain Belgium Belgium Hungary Belgium Hungary Luxembourg Belgium Spain Belgium Belgium Netherlands Netherlands Luxembourg Netherlands Netherlands Belgium Netherlands Netherlands

80 70 60 TWh

Austria Spain

50

Netherlands

40 30 Italy

Italy

Italy

Italy

Italy

Italy

Italy

1998

1999

2000

2001

2002

2003

2004

20 10 0

Austria Spain Luxembourg Portugal Switzerland Slovenia Hungary Germany Belgium The Netherlands Italy

Electricity net exports in CE (1998–2004) 110 Switzerland

100 90 Austria

80

TWh

70 60

Switzerland Austria Poland Switzerland Czeh.Rep. Germany Poland Czeh.Rep.

Austria Switzerland Poland

Switzerland Germany Poland Germany Poland

Czeh.Rep.

Switzerland Switzerland Germany Germany

Poland

Poland Czeh.Rep. Czeh.Rep.

Czeh.Rep.

Czeh.Rep.

Austria

50

Switzerland Slovenia

40 30

Slovakia France

France

France

France

France

France

France

Germany Belgium

20

Poland Czeh.Rep.

10

France

0 1998

1999

2000

2001

2002

2003

2004

Fig. 9.9. Evolution of net imports and net exports over the period 1998–2004. Source: UCTE (2005).

274

Electricity Market Reform Maximum percent use of transmission capacity SLO–IT ES–PO SK–HU CZ–DE AT–IT CH–IT FR–DE FR–IT CZ–AT DE–AT PL–CZ 0

50

100 (%/year)

150

200

Fig. 9.10. Major bottlenecks in CE transmission grid measured as percentage of use of transmission capacity per year in 2004. Source: UCTE (2005) (for details see Table 9A.3 in the Annex).

Basically, in the new competitive system, European interconnectors have to allocate electricity flows from low-cost regions to high-cost regions, and by doing so, they are expected to produce both a price convergence and a redistribution of stakeholders’ welfare. 9.3. How the System Changed: Political Issues Of Restructuring The restructuring of the CE electricity market was triggered by the EU Directive on ‘Common Rules for the Internal Market in Electricity’ which came into force in February 1999.4 The major intention was to create a common European electricity market, EC (1997).The major issues of this Directive (officially named 96/92) were: ● ● ●

Minimal requirements for the unbundling of generation and transmission. Minimal market opening, expressed by the consumption size of “eligible customers”. Different approaches for the access to the grid (negotiated or regulated, TPA or single buyer).

However, each national government within the EU had to “transpose” the EU Directive into national law and national rules. An overview on the major milestones is provided in Table 9.1. In practice, the major area of action within the European liberalization project was “providing access to the market”. Far less attention was paid to the issues of restructuring generation & supply and designing market places as well as ensuring adequate generation and transmission capacity. Independent energy regulators were introduced in all countries

4

As already mentioned in some countries in CE (Germany, Poland and Spain) steps towards liberalization were set already before the EU Directive went into force.


275

Table 9.1. Milestones of reforming in Continental Europe. 1996

EU-15

February 1997

EU-15

1998 1998 1998 February 1999

Spain Poland Germany EU-15

2001 2001

Austria EU-15

2003

EU-25

2003 2004

Spain EU-15 ⫹ 10

2004

EU-25

2005

Portugal, The Netherlands EU-25

2007

European Council of Energy Ministers and Parliament reached agreement on a market liberalization directive This “Directive concerning common rules for the internal market in electricity” (Directive 96/92/EC) became valid while waiting up to two more years for its transposition by countries Introduction of a Spanish centralized pool Introduction of TPA (market opening: 22%) 100% market opening in one step Directive went into force after a 2 years transposition delay: market opening due the directive in Austria, Belgium, France, Italy, Spain, Portugal and the Netherlands between 30% and 35% 100% market opening (in a second step) Approval of the “Directive of the European Parliament and the Council on the promotion of electricity from renewable energy sources in the internal electricity market (RES-E Directive)” (European Parliament and Council, 2001 – Directive 2001/77/EC) Approval of the “Directive concerning common rules for the internal market in electricity” (officially Directive 2003/54; usually named “the Second Directive”) 100% market opening Extension of the EU to 25 member countries, new CE member countries to open their market with 30% minimum Electricity Directive 2003/54 due to be transposed by member states All non-domestic customers made eligible in the EU in July 2004 An EU Regulation on cross-border electricity trade came into effect (Regulation 1228/2003) in July 2004 100% market opening Due to Electricity Directive 2003/54, 100% market opening in all EU-25 countries in July 2007

except Germany (and Switzerland which is not part of the EU). In addition, environmental issues were also treated very prominently. On the contrary, aside from a minimal unbundling, the restructuring of utilities and the design of market places was not tackled comprehensively by governments in most countries (few exceptions: Spain created a centralized pool, and Italy divested generation capacities). The milestones of reforming the electricity sector in Continental Europe are given in Table 9.1. 9.3.1. Providing non-discriminatory access to the market and to the grid The first important requirement for a competitive electricity market is non-discriminatory access to the grid. Therefore, a prerequisite for competition is the unbundling of generation

276


Table 9.2. Types of unbundling of transmission system operators and access to the grid in CE (as of December 31, 2004a. Country

Unbundling TSOb

TSO

Ownership

Access to the grid 2004

Austria

Legal (APG); Management (TIWAG, VKW)

APG (90%), TIWAG (6%), VKW (4%)

100% public, 100% public, 51% public

rTPA

Belgium

Legal (2005: ownership)

ELIA

100% Electrabel (2005: floated)

rTPA

Czech Republic

Legal

CEPS

51% CEZ, 49% public

rTPA

France

Legal

RTE

100% EdF

rTPA

Germany

Legal

RWE Netz, E-ON-Net, EnBW-Net, Vattenfall Transmission

100% RWE 100% E.ON 100% EnBW 100% Vattenfall Europe

nTPA

Hungary

Legal

MAVIR

100% public,

rTPA

Italy

Ownership

GRTN

100% public

rTPA … eligible customers SB(rTPA) – captive customers

Luxembourg

Management

ELIA (BE) RWE-Netz (DE)

100% ELIA 100% RWE

rTPA

The Netherlands

Ownership

TenneT

100% public

rTPA

Poland

Legal

PSE (Polskie Sieci Elektroenergetyczne S.A.)

100% public

rTPA

Portugal

Ownership

REN

100% public

rTPA … eligible customers SB(rTPA) … captive customers

100% public

rTPA

100% public

rTPA

Slovenia

Ownership

ELES

Slovakia

Legal

SEPS

Spain

Ownership

REE

Switzerland

No

Regional vertically integrated companies

a

rTPA No

rTPA: regulated third-party access; nTPA: negotiated third-party access; SB: single buyer model. Legal: legal separation of transmission and generation. Source: CEC, 2005; CEC, 2004, company reports, Power in Europe, own investigations. b


277

Table 9.3. Electricity directive implementation in CE countries. Country Austria Belgium Czech Republic France Germany Hungary Italy Luxembourg The Netherlands Poland Portugal Slovenia Slovakia Spain Switzerland

Market opening (%) (February 19, 1999)

Market opening (%) (January 1, 2005)

30 35 0 30 100 0 30 0 33 22 30 0 0 45 0

100 90* 55 68 100 67 79 57 100 80 100 75 66 100 0

Eligible customers (January 1, 2005) All >10 GWh* >0.1 GWh All non-households All All non-households >0.1 GWh >20 GWh All All non-households All All non-households All non-households All No final customers

*Figures for Wallonia. Full market opening in Flanders region. Source: CEC (2001); CEC (2005).

and supply from transmission. This means that access to transmission and distribution should be offered to all market participants at reasonable and non-discriminatory prices. So far, the experiences with respect to unbundling between generation and transmission in CE have been different. In Belgium, Spain, Portugal and Italy unbundling of generation and transmission by ownership was achieved either by full independence of the transmission company or by the flotation of a transmission subsidiary. In other countries, especially in Germany and France, only legal unbundling took place. In Switzerland, so far unbundling has only been done by means of internal management measures. These give no structural guarantees for avoiding discrimination in access to the grid, in particular when no independent regulator is able to monitor the behavior of grid managers. Table 9.2 provides the current status of unbundling. The second issue is the regime of access to the grid. Table 9.2 shows access to the transmission grid in various Western European countries (CEC, 2005). Access to the grid has been regulated in all countries except Germany where it was introduced in June 2005. The third issue is market opening. The geographically, and timely different opening of the markets led, at least to some distortions regarding free choice of supplier. Table 9.3 and Figure 9.11 depict the opening of the market in different CE countries between 1999 and 2005. Some countries like Germany, the Netherlands, Spain, Portugal and Austria have legally fully opened their markets, while others, like France, Luxemburg and Czech Republic have only partially opened their’s. In Switzerland (which is not member of the EU) there is currently no competition in supply.

9.3.2. The new institutional and regulatory environment In all countries, except Germany and Switzerland, independent regulatory authorities have been founded. An overview of these regulatory authorities and their staff and budget is

278

Electricity Market Reform Market opening 100 2005 2003 2001 1999

90

70 60 50 No market opening

Percent market opening

80

40 30 20 10 0 DE AT ES NL PO BE PL

IT

SL FR HU SK LU CZ CH

Fig. 9.11. Market opening in CE as of January 1, 2005. Source: CEC (2005) and earlier benchmarking reports.

given in Table 9.4.5 Powers vary widely from one country to another, but common core tasks are: ● ● ●

to ensure that unbundling is achieved; to regulate access to the grid; to regulate tariffs for the use of the transmission and distribution grid.

In practice, the current European regulatory governance consists of a decentralized framework on national levels in an incomplete process of convergence. Countries have6 established nationally based regulatory authorities which are administered by nationals. Access to the national TSO’s grid and operating system is regulated nationally. All this is done legally and with recourse to courts, while the European Directives and Regulations provide only a broad common frame. However the EC or the European Court of Justice can block this or that excess on a case-by-case basis. For example, in summer 2005 the European Court deemed illegal the “grandfathering” priority given to incumbent Foreign suppliers of the Dutch grid interconnections. 9.3.3. The promotion of renewables Currently, the promotion of electricity from renewable energy sources (RES-E) plays an important role in the energy policy of the EU. The major policy reasons are: (i) reducing the

5

It would be interesting to analyze whether there is any correlation to the size or budget of the regulator and the working of the market. Yet, unfortunately, such an analysis would go beyond the scope of this chapter. 6 Except Germany and Switzerland.

279

Competition in the Continental European Electricity Market Table 9.4. Budget and staff of regulatory authorities in CE. Country

Name (year of foundation)

Austria Belgium Czech Republic Germany France Hungary Italy Luxembourg The Netherlands Poland Portugal Slovenia Slovakia Spain Switzerland

E-Control (2001) CREG ERU (Bundesnetzagentur, 2005) CREG HEO AEEG ILR DTE URE ERSE Energy Agency URSO CNE No

Budget 2004 (Million euro)

Staff 2004

8.3 11.3 3.8 – 13.8 6.2 18 0.7 5.1 7.7 7 1.5 1.5 20.7 No

66 74 88 (180 in 2005) 108 95 100 32 55 267 51 22 57 175 No

Origin of budget P L P – E P P P P P L/P P P P No

L:levy on operators; P: public budge; No: does not exist. Source: European regulators, AIE, CEC (2004); Kaderjak (2005). Table 9.5. Renewable electricity targets as share of electricity consumption in the EU-25 member states. Country Austria Belgium Czech Republic France Germany Hungary Italy Luxembourg The Netherlands Poland Portugal Slovak Republic Slovenia Spain Switzerland

RES-E penetration 1997 (%)

RES-E target for 2010 (%)

70 1 4 15 4.5 0.7 16 2.1 3 1 38 18 30 20 68

78 6 8 21 13 3.6 25 5.7 9 7 39 31 34 30 No

dependence on energy imports; (ii) reduction of greenhouse gas emissions. To meet this target the EU has defined ambitious objectives which have been formalized in the “Directive of the European Parliament and the Council on the promotion of electricity from renewable energy sources in the internal electricity market (RES-E Directive)” (EC 2000). According to this directive, RES-E generation should reach a total share of 22% of electric production in 2010 from a level of 12% in 1998 (EC, 2000). Table 9.5 specifies the indicative targets for the share of RES-E for every CE country to be met by 2010.

Electricity generation (TWh/year)

280


60 50 40 Biogas Solid biomass Bio waste Geoth Wind onshore Wind offshore Hydro small scale Hydro

30 20 10 0 AT BE CZ FR DE HU IT LU NL PL PT ES SK SI CH

Fig. 9.12. Electricity generation from renewables in 2003 by country. Source: EEG TU Wien, GREEN-X database.

Figure 9.12 depicts the amounts of various RES-E technologies, country by country. Hydropower is the dominant source, but ‘new’ RES-E’s such as biomass and wind are starting to play a role. Wind energy has had a yearly growth rate of about 35% per year over the last decade. Biomass is especially popular in countries like Poland, where it is commonly cofired in existing coal power plants to meet the negotiated renewable energy share. Yet, the higher costs of RES-E technologies, compared to existing conventional power plants, advocate financial support. As the choice of instruments has not been prescribed or harmonized in Europe, every country has adopted its own. In Table 9.6 an overview is provided of promotion schemes for RES-E in EU-15 countries for the Year 2004. Feed-in tariffs are currently used in most of the CE countries. This instrument has so far turned out to be the most effective for a fast deployment of significant shares of RES-E.7 The promotion of wind energy has so far been the most successful in this context. As Figure 9.13 depicts, that due to feed-in tariffs in Germany and Spain, considerable capacity of wind power was constructed up until 2004.

9.4. Comparison of Developments by Country The developments toward competition in the countries and sub-markets have been quite different so far, as can be seen in Table 9.7. Germany started with a 100% market opening without any restructuring of the industry. Later on, a rapid merger process took place, resulting in the disappearance of half the generating, transmission companies. Moreover, the German idea of competition was unique because no regulatory authority was created. It soon became evident that high grid charges, discrimination with respect to access to the distribution network, and high transaction costs of the negotiated TPA were major problems for this model, in particular, because of the hundreds of regional or local distribution grid companies. Finally, in 2005 a regulatory body was created. 7

For a comprehensive comparison of the relative efficiency of guaranteed feed-in tariffs, bidding system and exchangeable quotas systems, see Finon and Perez (2006).

Table 9.6. Overview of the main policies for the promotion of RES-E in CE countries (as of end of 2004).{not in reference} Country

Main electricity support schemes

Comments

Austria

Feed-in tariffs (presently terminated) combined with regional investment incentives

Feed-in tariffs have been guaranteed for 13 years. The instrument was only effective for new installations with permission until December 2004. The active period of the system has not been extended nor has the instrument been replaced by an alternative one.

Belgium

Quota obligation system/TGC combined with minimum prices for electricity from RES

Federal government has set minimum prices for electricity from RES. Flanders and Wallonia have introduced a quota obligation system (based on TGCs) with obligation on electricity suppliers. In Brussels no support scheme has been implemented yet. Wind offshore is supported on the federal level.

France

Feed-in tariffs

Germany

Feed-in tariffs

For power plants ⬍12 MW feed-in tariffs are guaranteed for 15 years or 20 years (hydro and PV). For power plants ⬎12 MW a tendering scheme is in place. Feed-in tariffs are guaranteed for 20 years (Renewable Energy Act). Furthermore soft loans and tax incentives are available.

Italy

Quota obligation system/TGC

Obligation (based on TGCs) on electricity suppliers. Certificates are only issued for new RES-E capacity during the first 8 years of operation.

Luxembourg

Feed-in tariffs

Feed-in tariffs guaranteed for 10 years (for PV for 20 years). Also investment incentives available.

The Netherlands

Feed-in tariffs

Feed-in tariffs guaranteed for 10 years. Fiscal incentives for investments in RES are available. The energy tax exemption on electricity from RES was finished on January 1, 2005.

Portugal

Feed-in tariffs combined with investment incentives.

Investment incentives up to 40%.

Spain

Feed-in tariffs

Electricity producers can choose between a fixed feed-in tariff or a premium on top of the conventional electricity price; both are available during the whole life time of the RES power plant. Soft loans, tax incentives and regional investment incentives are available.

Czech Republic

Feed-in tariffs (since 2002), supported by investment grants revision and improvement of the tariffs in February 2005.

Relatively high feed-in tariffs with 15 year guaranteed duration of support. Producer can choose between fixed feed-in tariff or premium tariff (green bonus). For biomass cogeneration only green bonus applies.

Hungary

Feed in tariff (since January 2003) combined with purchase obligation and tenders for grants

Medium tariffs (6–6.8 c/kWh) but no differentiation among technologies. Actions to support RES are not coordinated, and political support varies. All this results in high investment risks and low penetration.

Poland

Green power purchase obligation with targets specified until 2010. In addition renewable exempted from the (small) excise tax

No penalties defined and lack of target enforcement.

Slovak Republic

Program supporting RES and EE, including feed-in tariffs and tax incentives

Very little support for renewable. Main support program runs from 2000, but no certainty on time frame or tariffs. Low support, lack of funding and lack of longer-term certainty make investors very reluctant.

Slovenia

Attractive feed-in system combined with long-term guaranteed contracts, CO2 taxation and public funds for environmental investments

None

Switzerland

Source: Huber et al. (2005).

282

Electricity Market Reform Installed wind capacity 2004 SL HU SK CZ LU CH PL BE FR PT AT NL IT ES DE 0

2,000 4,000

6,000 8,000 10,000 12,000 14,000 16,000 18,000 MW

Fig. 9.13. Wind capacity in CE by the end of 2004. Source: EWEA.

In Austria the market was legally opened in two steps: 33% in 1999 and 100% in 2001. In 2001 a voluntary spot market (EXAA) was founded. Since 2000 a discussion has been ongoing concerning several models of national and cross-border mergers and takeovers. Yet, so far only minority shares of some suppliers have been sold to the French EdF, or the German EnBW and RWE. In France, more than 90% of capacity is concentrated in EdF, with two potential competitors who have been institutionally linked to it. These links have been weakened in order to make them independent in the near future, and have been opened to new entrants, notably Electrabel and ENEL. These “fringe generators” are CNR, a hydro generator, and SNET, a subsidiary of Charbonnages de France which produces 8.5 TWh by dispatchable coal plants. The transmission business was made a subsidiary in the second half of 2005, and could be floated as soon as 2006. EdF, itself will put around 20% of its shares on the market before the end of 2005. The major feature, in the Czech Republic and the Slovak Republic, of restructuring was the break-up of the former vertically integrated public utility into generation, grid and supply companies. Furthermore, in the meantime, parts of the generation and supply companies have been privatized. In 1993, the Czech Republic spread about 31% of CEZ shares among investors (individuals and funds). As an attractive offer was not received for the rest of CEZ, further privatization has been delayed so far. In the Slovak Republic, 66% of the generator SE is being privatized (2005). In Hungary, Slovenia and Italy steps were taken to reduce the power of the former generation monopoly. Currently, however, it appears that in these countries the former monopolists still have a strong position in the market (ENEL kept 50% of the Italian generation capacity, plus the cash made by selling the rest of its plants – as “Gencos” or by the sale of transmission and distribution grid shares). In the Netherlands, until 1998, generation was dominated by four large regional companies: EPZ, EPON, UNA and EZH, who jointly owned the generator Sistema Electríco Publico (SEP). The Dutch government’s initial idea was to combine liberalization in supply with the concentration in generation by merging the four companies and SEP. This attempt should have

283

Competition in the Continental European Electricity Market Table 9.7. Differences in reforming and market design in various countries. Process of market opening

Mandatory pool

Voluntary day ahead exchange

Futures market

Privatization process

Divestment of generation capacity

Takeover, merger within the country

AT

Fast (2 years)

No

Yes (EXAA)

Yes (EEX)

Moderate

No

Under discussion

BE

Slow

No

No

No

*

No

No

CZ

Moderate

No

Yes (2004)

No

No

No

No

DE

Very fast

No

Yes

Yes

*

No

Yes, half electricity generation plus Ruhrgas

FR

Slow

No

Yes

No

No

No

Yes, two fringe generators

HU

Moderate

No

No

No

Moderate

No

No

IT

Slow

No

Yes (since 2004)

No

Yes

Yes

Yes, mainly abroad (ENEL in SK)

LU

Slow

No

No

No

N.A.

No

No

NL

Moderate

No

Yes (APX)

No

Yes

No

Yes, mainly from abroad

PL

Fast

No

PO

Moderate

SK

Moderate

No

No

No

Yes

No

No

SL

Moderate

No

Yes (2003)

No

Moderate

Moderate

No

ES

Moderate

Yes

No

No

*

No

No

CH

No

No

No

Yes (EEX)

*

No

No

Yes

No

Moderate

Yes

Moderate

No, but intended with Spain

No

No

Yes, moderate

Moderate abroad

*Major generators were already largely private before liberalization started.

created a “national champion” that would be able to compete on the European scene (van Damme, 2005). Yet, the merger failed because these companies could not agree. The major restructuring feature was then the sell out of half of the former largely public-owned generator to companies from abroad (Electrabel, Reliant, E.ON). Another trend is the vertical reintegration of generators and suppliers for example, by the purchase of power plants by suppliers. After a series of mergers and takeovers, two large

284


Dutch companies survived and are now integrated into generation, distribution and supply (ESSENT and NUON). The TSO TEnnET and its subsidiary, the power exchange (PX) of Amsterdam, have been 100% state owned for some years. In Belgium, the process has been dominated by the incumbent company Electrabel, which is controlled by the Suez group (France) through the intermediate engineering contractor Tractebel. A “second” Electrabel was developed outside Belgium by collecting 15,000 MWe plant capacity, mainly in Europe (the Netherlands, Poland, Hungary, Italy, France, Spain). In spring 1999, Tractebel pretended to become a liberalization champion. They split their companies into parts while keeping control over all of them all. In 2005 however, Electrabel and Tractebel were merged to increase their stock market size. They understood that being one of the oligopolistic players on the European and worldwide market was more profitable than to stay linked to the limited Belgian market (Verbruggen and Vanderstappen, 1999). The Spanish approach initially looked like being one of the most ambitious. However, the structure of the industry with two dominant producers integrated in distribution and supply was never changed. As a result, after the introduction of a centralized pool8 in 1998, the issue of market power exerted by the two largest incumbent generators was very soon raised. Crampes and Fabra state in (2005): “The 1997 reform did not succeed in introducing effective competition but retained an opaque regulation which has been subject to continuous governmental interventionism. …” Note, that due to scarce interconnection capacity between Spain and neighboring countries foreign utilities have not been very influential in the Spanish pool so far. The issue of market power is still – in 2005 – the major problem in Spain and could be reinforced by the take over attempt of the first gas company Gas Natural of the first electrician Endesa. In 2005, an investigation in the competitiveness of the Spanish market was conducted by Perez-Arriaga, and the new government is reviewing the rules with the view to changing them. As well as transmission, there were only four significant companies, all largely private and vertically integrated. While the former government blocked the merger of the two largest utilities (Endesa and Iberdrola), it allowed the takeover of Hidrocantabrico by EDP. Furthermore, when Endesa put 5% of its activities up for sale, it was bought by ENEL of Italy, as Endesa had just taken control of Elettrogen in Italy (Soares, 2003). In Portugal, the hard process of the privatization of the EDP, and the creation of a competitive affiliate (Sistema Electrico Nao-Vinculado, SENV) has been shaping the reform process so far. The idea was to split the national electricity system into two sub-systems: the public utility system SEP and the independent system SENV. SEP and SENV are not generator s, but sub-systems of the national electrical system. The former has to satisfy demand under the principal of a uniform tariff on the mainland, which moderates the application of market rules. It also has centralized planning. The latter has no responsibility for public service and comprises of two sub-systems: the non-binding system (SENV) and the independent producers. The SENV operates according to market rules and comprises of producers, distributers and eligible customers. Non-binding producers and customers are allowed to use the public utility system grid for a fee (Soares, 2003). Other objectives, since the start of reform in Portugal, have been to create a “national champion” by merging gas and electricity monopolies (which was refused by the European Competition Authority) and a joint Iberian market with Spain (The MIBEL project). Yet, so far this Mibel has been repeatedly postponed and currently, it is being planned to put it into practice in 2006. One problem is that “without substantial enhancements to interconnection, 8

While the participation in this pool is in fact mandatory, market participants are also allowed to enter into physical bilateral contracts (Crampes and Fabra, 2005).


285

it should be clear that the impact of the Spanish market on the highly concentrated, Portuguese market can only be marginal, and the impact of the Portuguese wholesale market on the Spanish minimal” (PiE 437, p. 3). With respect to divestment of capacity, Italy was the only country in Continental Europe where the former state-owned champion had been privatized and had to give away generation capacity (Lorenzoni, 2003). Currently, however, ENEL is in a comfortable position because it is still the largest electricity producer in a market with congested borders and a congested internal grid and can act as a private company with the cash generated by its divestiture. ENEL has now a market share of 50% of generation capacity, and an Italian PX has been opened. In Switzerland a draft law providing for ultimately complete opening of the Swiss electricity market was rejected by the Swiss population in a referendum in 2002. Another draft law providing for market opening for larger industrial customers was provided for discussion in 2004. Given the legislative procedure and a possible new referendum, first steps of market opening can be expected, in the case that the law is finally approved, at the earliest in 2008 (CEC, 2005). Eastern European countries are physically integrated within the western European grid, and have taken the first steps toward adopting the “western model” with regulated third-party access for larger customers. There has been partial privatization of companies within the industry (except in Slovenia) and the reduction of barriers to international trade. But, like the rest of Europe, each reform is unfinished in regard to its market design and the existing market power of the dominant player. The typical Eastern European market structure is made up of a dominant wholesaler and a competitive fringe. The competitive fringe is strongly limited by long-term contract structures that often allow the dominant wholesaler to deploy the generators, so being able to deny other companies’ access to surplus capacity that has not been contracted in advance (Kaderjak, 2005). It is also the case concerning the support for renewable energy which often takes the form of a feed-in tariff under which the power is sold to the dominant wholesaler, thus consolidating its position even more. Poland and Hungary were the forerunners of reform in Eastern Europe. Poland introduced TPA in 1998, and the Czech Republic and Hungary conducted unbundling of generation and transmission in the early 1990s. Hungary established a regulator in 1994 and started privatization of supply and most of generation in 1995. At the same time the gradual removal of price subsidies was started (Kaderjak, 2005).

9.5. The Markets: Structures and Performances The markets’ structures and performances after the start of liberalization can be measured in different ways. However, the evolution of electricity prices is, presumably, the most important indicator. A desirable outcome of a single European electricity market is the achievement of a lower price and a price convergence through wholesale and retail competition (Jamasb and Politt, 2005). Hence, in this section, after having examined the characteristics of the markets and the markets’ structures, focus will be put on prices changes for differing groups of customers and in various regions. 9.5.1. Characteristics of the markets Table 9.8 depicts the markets in CE. In particular, the degree of liquidity in spot markets and bilateral markets is indicated. As the European Commission states (CEC, 2005): “Ideally spot

286

Electricity Market Reform Table 9.8. Trading in CE in 2003/2004: spot markets, centralized versus mandatory pools and OTC.

Total supply 2004 (TWh) AT BE CZ DE FR HU IT NL PL PO SK SL ES CH

51.8 87.5 61.4 445.1 554 38.2 322 6.3 110.9 144.8 45.5 26 12.3 234.5

Spot market/ centralized mandatory or voluntary pool EXAA No OTE EEX (Sp.m.) Powernext (Sp. M.) No Yes (2005) APX (Sp. M.) Pol-PX No No SLOex OMEL (C.p.) No

Volume (TWh 2004) 1 0.3 39 7.5 2 15 1.1

0.36 204

OTC (TWh 2004) N.A. N.A. N.A. 342 300 N.A. 56 240 N.A. N.A. N.A. N.A. 5 N.A.

Source: CEC (2005) and own investigations.

markets should have enough liquidity to give a reliable and transparent price signal (… .). The normal benchmark from other commodity markets is that the volume of trade (of long term contracts) should be roughly 10 times the amount of physical delivery.”9 As Table 9.8 shows, no CE market is approaching this level. As can be seen from Chapter 1 of this book no CE country is in the list of the 12 most competitive countries. Most of the CE countries have few and relatively illiquid organized markets (PXs) for electricity. Such spot markets which exist, as in Poland and Slovenia, trade less than 5% of the total electricity consumption. Bilateral contracts are the most frequent form of “trade” arrangement in new EU members’ states. 9.5.2. Mergers, takeovers and market concentration The industrial reference model for electricity completely changed between 1995 and 2001. It has shifted from a preference for vertical disintegration between generation, trading, and sales to final consumers toward a preference for vertical reintegration of production, trading and final sales. Among the best illustrations of the changing “industrial paradigm” are the shifting attitudes of financial markets, financial analysts, rating agencies and banks vis-à-vis disintegrated structures, especially concerning “pure” trading and “pure” generation as in merchant plants. Bankers and financiers have finally joined force with stockholders and managers of firms operating in competitive energy markets, and concluded that vertical integration is the best protection against volatility and the cyclical nature of markets. 9

In the EU, there are differences regarding the mutual role of bilateral trade (with or without use of a broker) and PXs. We cannot guarantee the data in Table 9.8 which is mainly based on an EC report (CEC, 2005). The OTC figures are likely to be higher. There is no real transparency in the markets outside PXs and therefore it is not clear what conclusions can be drawn.

287

Competition in the Continental European Electricity Market Electricity sector national and cross-border M&As in the EU 60 50

National M&As Cross-border M&As 10

40 30

10

14 13

42

20 3

10 0

3 5

10

1998

1999

22

2000

27 19

2001

2002

2003

Fig. 9.14. Number of mergers within European electricity companies from 1998 to 2003. Source: Codognet et al. (2005).

Hence, for effective competition, a large number of companies is required. This has been clearly proven by the English and Welsh examples, where the number of generators has been increased several times by the regulatory authority (as well as by investors, notably the regional distribution and supply companies, the RECs). The “merger-mania” within the CE after the start of liberalization indicates that the major strategy of the bigger incumbent utilities is competing by merging so as to purchase market shares. Figure 9.14 depicts the mergers within the EU. These activities reached a maximum number in 2004, 5 years after liberalization started. In addition to vertical reintegration, we also observed intense activity in horizontal mergers and acquisitions. The most significant example is doubtlessly Germany, where the 10 biggest electrical and gas concerns that existed at the time the European directive was adopted in 1996 have now become four. As in the German example, integration and concentration between electricity and gas is another defining feature of this new “consolidation” phase in Europe’s energy industry. Among the seven biggest electricity firms in Europe, Vattenfall and EDF have proven themselves to be anomalies because they are notably less involved in gas to date. Finally, while gas wholesale markets and concerns have persisted in courting the entry of large European and North-American petroleum and gas companies, electricity wholesale markets and electricity and gas retail markets, have not experienced any comparable influx. Thus, the upshot is a net “consolidation” of the industry on the pan-European scale, with an increasingly concentrated small number of international European firms in the sector, sometimes mockingly called the “seven brothers” in a transparent reference to the “seven sisters” of the international petroleum industry in the 20th century. Nonetheless, on a country-by-country basis, the European Union often comes across as juxtaposing domestic markets of monopolies or duopolies with a small competitive fringe in which one, two or three fringe new entrants operate. In many Eastern European countries, national companies have been sold to strategic investors from abroad, with EdF E.On, RWE, Electrabel and Vattenfall particularly active. In reaction, some countries like Czech Republic, Slovakia and Slovenia have been concerned with the retention of national champions. These national champions have the size to survive

288


among the larger European groups with their unfortunate consequences for the level of competition within their national market and the European competitive game. The vested interests of the dominant incumbents in this region are encouraging them to fight against greater competition which is being pushed by further reforms. How should these mergers be seen in the light of competition? In principle, mergers and acquisitions should not be a major preoccupation. On the one hand, this issue is “old hat” in European competition policy, and, on the other hand, it is an excellent lever for directly obtaining structural remedies on a European scale that would be otherwise unattainable. If, nonetheless, certain “real” problems emerge, this more likely reflects on the deficiency of certain national rulings, especially when governments or “ordinary” judges can deliberately ignore the anticompetitive effects of their decisions. This would result at the very least, in a lack of harmonization between national decisions and those taken at the European level. The E.ONRuhrgas merger in Germany which created the biggest gas and electricity concern in the western world, will remain a bone of contention and a source of confusion for a long time. However, we cannot see any simple workable solution given the unwillingness of national governments to remedy the situation. The recent strengthening of the harmonization and cooperation between national and European authorities affects only the competition authorities, and not the other national third parties that possess other real decision-making powers. See how Portugal, and more recently Spain paid for having national gas and electricity mergers. With respect to market shares in CE, in 1998, 10 generators owned 60% of the generation capacities, in 2002 it was only six (see Codognet et al., 2005). Thomas (2003) suspects that finally European-wide only “seven brothers” will remain as large generators. Of particular concern, with respect to competition, is the situation in Central Europe (France, Germany, the Benelux countries and Austria). The concentration process in the electricity generation market was especially fulminous in Germany. Mez (2003) provides an impressing and detailed description of this process. A different but converging picture is described in Finon (2003). He portrays how a dominant player like EdF in France can benefit from liberalization by exerting market power in the home market, while at the same time is pursuing an aggressive acquisition policy abroad. Verbruggen and Vanderstappen (1999) show the same for Electrabel – Distrigas group in Belgium. As can be seen from Figure 9.15, of the 13 largest generators which existed in 1999 – the year liberalization started – in CE 5 years later only nine remained. Now in Continental Europe seven large concerns dominate the market: EdF-EnBW, RWE, E.ON, Vattenfall, Endesa, ENEL and Electrabel (Haas et al., 2002). Another interesting fact is (Table 9.9) that in the ranking of the largest generators public ownership still prevails. Of special interest is that the larger European groups put special focus on extension of their interest spheres to regions which are adjacent or separated by low transmission capacity from their home area (see Fig. 9.16). Table 9.10 depicts the current market structure in CE countries. In most countries market structure is highly problematic particularly when the national grid is poorly connected with adjacent markets. It is of specific interest that potential imports vary considerably. The small countries Luxemburg, Slovakia, Slovenia, Austria and Hungary have a potential of more than 70%. In the large countries Spain, France and Italy the potential is less than 20%. 9.5.3. Wholesale electricity price evolution How electricity prices developed after restructuring is of special interest. Figure 9.17 depicts the price evolution in CE in 1999–2004. With the exception of Italy in 2004 there was some

289

Competition in the Continental European Electricity Market Number of large generators in CE VEW

Largest CE generators 1999

VEAG EnBW CEZ Iberdrola

Iberdrola

Endesa

CEZ

Elektrabel

Endesa

Bayernwerk

Electrabel Vattenfall Europe ENEL

Vattenfall Preussen Elektra RWE

Largest CE generators 2005

Major mergers and acquisitions

RWE

ENEL

E-ON

EdF

EdF/EnBW 0

100

200 (TWh)

300

400

0

100

200

300

400

500

600

(TWh) 13

9!

Fig. 9.15. Largest European electricity generators in 1999 and 2005. Source: Own investigations.

convergence of wholesale electricity spot market prices. Moreover, while volatility in 2002 and 2003 was rather high it became moderate during 2004. In the first half of 2005, prices in Western markets increased, while prices in Poland remained stable. From Figure 9.17 we derived the following effects: (i) in Western Europe, prices increases were relative to the frame and timing of liberalization; (ii) the price level is highest in areas where capacity margin is smaller, and cross-border transmission capacity is congested (Italy, the Netherlands); (iii) prices have been highest in years when there was low hydro or low nuclear availability; (iv) however, wholesale prices are increasing and are converging at the top in markets which are connected by sufficient transmission capacity. Therefore, a major question is, are these prices a result of competition? That is to say, do these prices reflect the marginal costs of the generation set or are they influenced by some kind of market power. As for example, Muesgens (2004) shows from 2001 to 2003 in Germany, the difference between wholesale electricity prices and short-term marginal generation costs have increased continuously, possibly due to increasing exercise of market power (see Fig. 9.18). It compares with the historical data marginal cost model. As can be seen since 2001, the gap between prices and short-term marginal costs had been continuously widening until 2003. 9.5.4. Transmission prices Transmission and distribution tariffs represent a significant share of final customers’ electricity prices but are not subject to competition pressures. Figure 9.19 compares the prices for transmission within the EU countries. These prices vary considerably: between €3.5/MWh in the Netherlands and €13.8/MWh in Poland. These huge differences are currently still under investigation (ETSO, 2005). As can be seen from Figure 9.19, one important part of the tariff is the “indirect component” which reflects burdens like stranded costs, public interest contributions, fees to promote renewable energy and others. The indirect cost component contributes to almost 50% of the transmission price in Poland and Italy.

290


Table 9.9. Largest generators in various countries in 2004 and their ownership structure. Generation (TWh) [Wholesale (TWh)]

Public ownership (in 2005) (%)

Capacity (GW)

EdF

487

119

E-ON

137

36

5

56% financial sectora

RWE

192

43

33

50%(?) banks …a

ENEL

165

45

32

Vattenfall DE (Germany only)

81

16

89% (Swedish government)

Electrabel (BE) BE only?

75

28

0

100

Endesa (ES)

70

28

42

Iberdrola (ES)

61

16

0

CEZ (CZ)

57

12.3

69

EnBW (DE)

55

12

46.6

8

Union Fenosa (ES)

30.8

5.2

Verbund (AT)

28.5

7.3

51

MVM (HU)

28.3

N.A.

98

27

EdP (PO) PKE (PL)

19.5

Hidrocantabrico (ES)

18.0

ElectraBel NED (NL)

17.9

HSE (SL)

7.1

Suez (51%) (100% in 2005) EdP (4%)

45% EdF (French government) and 50% German public associations

BOT (PL)

Slovenske Elektrarne (SK)

Other significant ownership shares

100 Wienstrom (10%), EVN (10%)

6.95

34

66% ENELb

12

51

Iberdrola (4%)

5

85 51% EdP (Portuguese

4.7

government) 100

a

E.ON and RWE are not ready to reveal more detailed information on the ownership structures of their companies. b In 2005 in process of privatization. Source: Company reports, power in Europe, personal information.

9.5.5. Retail electricity price evolution The major expectation of final customers, with respect to the liberalization of electricity markets, was that prices would drop substantially. Figures 9.20 and 9.21 depict the price evolution in CE from 1999 to 2004 for households and large industrial customers. As can be seen from Figure 9.21, large electricity users were seeing – at least temporarily – indeed lower prices. Yet as Figure 9.20 shows households electricity prices in 2004 had already reached the same level as pre-liberalization, or were even higher. With the exception of Poland (and for


291

Mains congestions and majors M&A in EU-15 Vattenfal

PowergenTXU L.E. Innogy

E.ON

EB

RWE EnBw

EDF

Brazil Argentina

SNET

Iberdrola Latin America

HC Endesa

CNR Interpower Edison ENEL Elettrogen

Fig. 9.16. Mains congestion and major M&As in CE. Source: Parthenay and Perez (2005); EU and companies annual reports.

most countries even earlier) prices have been increasing since 2003. Moreover, neither for households nor for industrial customers did any remarkable convergence in prices take place. This was one of the expectations of the common European market. Figures 9.22 and 9.23 show that for both groups of final customers a wide range of price levels still prevails in different member states.10 The prices for households vary between €8 c/kWh in Poland and €16 c/kWh in its neighbor country Germany. Electricity prices for industry range from €4.1 c/kWh in the Czech Republic to €10.3 c/kWh in Italy. Figure 9.24 shows that after liberalization was announced prices fell. But soon suppliers started to increase prices. Of course, there are many reasons for price increases and outside competition effects, for example, transaction cost of market creation (e.g. splitting of distributor into two legal companies) for distribution and for supply, new power plants that have to meet new ecological legislation (emission limits, minimum thermal efficiency, etc.), which will mean utilization of expensive technologies (especially in Eastern Europe), emission allowances for CO2, consumer tax imposed to fossil fuels from 2007 (according to EU rules), fees for increasing share of RES-E production. Figures 9.20–9.23 require more in-depth investigation. 10

A sound comparison of prices over 10 years would require the expression of them in real terms because of different rates of inflation. Obviously this is difficult for so many countries. Another caveat remains: EUROSTAT figures are based on tariffs, whereas this may have given a valid representation during time of monopoly, it does not after liberalization because new pricing schemes outwith the former-regulated tariffs have been offered with the tariffs representing the maximum. Hence the figures cannot give the true picture of the development. This is particularly true for large-scale consumers. They have many options for buying electricity. Private contracts may not be represented in EUROSTAT figures, and therefore may give a slightly distorted picture.

292


Table 9.10. Market shares of largest generators in various countries 2004. Largest (%)

Three largest (%)

Import potential (TWh, %)

Largest generator

Second largest generator

Third largest generator

Fourth largest generator

AT

53

76

37.7 (73%)

Verbund (53%, 29.8 TWh)

Tiwag (13%, 6.7 TWh)

Wienstrom (10%, 5.8 TWh)

EStAG (9%, 5.0 TWh)

BE

85

94

40.3 (46%)

ELECTRABEL (85%, 75 TWh)

73

82

30.7 (50%)

CEZ (73%)

SPE (9%, 8 TWh) Prazˇ ská teplárenská (5%)

CZ

Energotrans (4%)

Dalkia (3%)

DE

34

71

122.6 (28%)

RWE (34%)

E-ON (23%)

Vattenfall (14%)

EnBW (10%)

FR

89

94

106.9 (19%)

EdF (89%, 487 TWh)

CNR (3%, 16 TWh)

SNET (2%, 9 TWh)

HU

46

65

27.2 (71%)

MVM (46%)

IT

46

65

52.6 (16%)

ENEL (46%, 165 TWh)

Edison (12%, 20 TWh)

Edipower (7%, 10 TWh)

Endesa (6%, 5 TWh)

LU

65

90

8.8 (139%)

Cegedel (65%)

Sotel (25%)

NL

25

80

41.2 (37%)

ElectrabelNed (17.9 TWh)

Essent (14.65 TWh)

Nuon (14.5 TWh)

E-ON Benelux (9.9 TWh)

PL

30

52

30.7 (21%)

BOT (30%)

PKE (13%)

Kozienice (9%)

PAK (9%)

PO

65

80

8.8 (19%)

SEP

SENV

SK

84

89

26.3 (101%)

Slovenske Elektrarne (26 TWh, 84%)

PPC (3.5%)

TEKO (1.4%)

SL

54

98

18.4 (150%)

HSE (7.1 TWh, 54%)

ELES/GEN (5.2 TWh, 39%)

TET (0.6 TWh, 5%)

ES

39

78

19.3 (8%)

Endesa (39%)

Iberdrola (28%)

Union Fenosa (11%)

Hidrocantabrico (7%)

CH

26

53

74.9 (137%)

NOK (25%, 15.9 TWh)

BKW (15%, 9.4 TWh)

ATEL (13%, 8.3 TWh)

EWZ (7%, 4.3 TWh)

Source: Company reports, power in Europe, personal information.

9.5.6. Evolution of capacity margin As in many liberalized electricity markets, many CE countries started this process with significant excess capacities in generation which had been built up during the time of regulated area monopolies. This was a common motivator and driver for competition introduction. Yet, excess in generation capacity played a central role in the restructuring process of ESI. Excess capacity in generation depends on transmission capacity – the price

ar Au y 1 gu 99 9 s M t 19 ar 9 O ch 9 ct 2 ob 00 er 0 2 N Apr 000 ov il em 20 be 01 r2 00 M D ec ay 1 em 20 be 02 r Ju 20 0 Ja ne 2 nu 20 03 a Au ry 2 0 g Fe ust 04 br 2 Se ua 00 pt ry 4 em 20 be 05 r2 00 5

br u

Fe

/MWh

ua ry 19 Au 99 gu st 1 M ar 999 ch O 20 ct 00 ob er 20 Ap 00 N ov ril 2 em 00 1 be r2 0 M ay 01 D ec em 200 2 be r2 Ju 00 2 n Ja e 2 00 nu 3 ar Au y 20 gu 04 Fe st 2 b 00 Se rua 4 pt ry em 20 be 05 r2 00 5

br

Fe

/MWh


90 Spot market prices (Central Europe)

80

70

60

50

40

30

20

10

0 EE XMix /MWh

90 APX (NL) /MWh

Spain /MWh POL-X /MWh

Spot market prices (Southern Europe)

80

70

60

50

40

30

20

10

0

Italy /MWh

Fig. 9.17. Evolution of electricity prices in CE 1999–2005. Source: Homepages of the PXs.

293

294

Electricity Market Reform Marginal cost model spot market base 45.0 40.0 35.0

/MWh

30.0 25.0 20.0 15.0 10.0 5.0

Ja

nu ar y Se M 19 pt ay 99 em 1 Ja ber 999 nu 1 ar 99 9 Se M y 20 pt ay 00 em 2 Ja ber 000 nu 2 ar 00 0 Se M y 20 pt ay 01 em 2 Ja ber 001 nu 2 ar 00 1 Se M y 20 pt ay 02 em 2 Ja ber 002 nu 2 ar 00 2 Se M y 20 pt ay 03 em 2 Ja ber 003 nu 2 ar 00 3 Se M y 20 pt ay 04 em 2 be 00 r2 4 00 4

0.0

Spot market price base EEX /MWh Marginal costs base model TU Wien /MWh Marginal costs base model EWI Köln /MWh

Fig. 9.18. Price evolution in Germany and marginal cost models. Source: EEX, Müsgens (2004), own investigations. Transmission invoices at EHV level NL Direct Indirect *Preliminary

ES FR AT CZ* CH* SK* DE BE SL PO IT PL 0

2

4

6

8

10

12

14

/MWh Fig. 9.19. Comparison of transmission prices for producers and consumers connected at EHV. Source: ETSO (2004) national regulators; ETSO (2005) national regulates.

295

Competition in the Continental European Electricity Market Household electricity prices (excluding taxes) 16 14 12

c/kWh

10 8 6 4 2 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 AT PL

BE PO

CZ SL

DE SK

FR ES

IT CH

NL HU

Fig. 9.20. Evolution of households’ electricity prices in CE excluding taxes. Source: CEC (2004); CEC (2005); IEA (2005) based on EUROSTAT Dc, average electricity consumption: 3500 kWh. Note that the situation for Italy is specific. Average consumption is lower than 3500 kWh/ year and electricity prices for lower consumption are significantly lower (about 40%).

Industry electricity prices (excluding taxes) 10 9 8 7 c/kWh

6 5 4 3 2 1 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 AT PL

BE PO

CZ SL

DE SK

FR ES

IT CH

NL HU

Fig. 9.21. Evolution of large industrial customers’ electricity prices in CE excl. taxes. Source: CEC (2004); CEC (2005); IEA (2005), based on EUROSTAT Ig, average electricity consumption: 24 GWh.

296

Electricity Market Reform Electricity prices households 2004 CH ES SK SL PO PL LU NL IT HU FR DE CZ BE AT 0

5

Energy price Energy tax

10 c/kWh

15

20

Fees for RES, CHP, Stranded investments VAT ⫹ other taxes

Fig. 9.22. Household electricity prices in CE in 2004/2005 (based on EUROSTAT Dc, average electricity consumption: 3500 kWh). Source: CEC (2005), EUROSTAT (cited from Power in Europe), IEA (2005).

Electricity prices industry 2004 CH ES SK SL PO PL LU NL IT HU FR DE CZ BE AT

10.3

0

2

4

c/kWh

6

8

10

Energy price Energy tax Fees for RES, CHP, Stranded investments Fig. 9.23. Industry electricity prices in CE in 2004/2005 (based on EUROSTAT Ig, average electricity consumption: 24 GWh). Source: CEC (2005); EUROSTAT (cited from Power in Europe), IEA (2005).

297

Competition in the Continental European Electricity Market 1.5 1.4 1.3 1.2 AT CH DE FR IT ES

Index

1.1 1 0.9 0.8 0.7 0.6 0.5 1990

1992

1994

1996

1998

2000

2002

2004

Fig. 9.24. Annual variation of hydro power availability in CE hydro power countries. Source: UCTE and own Investigations.

competing utilities receive for electricity will be equal to their short-term marginal cost. Under perfect competition conditions without any remarkable excess capacities, the price should be equal to the long-run marginal costs (LRMC). But if there is no competition or a too tight capacity the price can be substantially higher than both marginal costs, especially when demand is not true to price. As Figure 9.27 shows in recent years excess capacity decreased continuously in CE submarkets (spare capacity: net capacity minus maximum load). Another important issue regarding the availability of adequate generation capacity is the volatility of hydro power. As Figure 9.24 shows for the major CE hydro countries, hydro power availability varies tremendously over time. Moreover, in countries like Austria, Switzerland and France the differences are very similar. Furthermore, in winter months the minimum production in the long run is only half of the maximum. 9.5.7. Cross-border transmission issues The share of cross-border exchanges in European electricity consumption reached about 13% or total sum of 300 TWh in 2004. However, the volume of exchange was limited by the transmission capacity between neighboring grids. As Glachant (2005) notes, many interconnections are managed by administrative rules without market economic bases. Roughly half of interconnections between countries of Continental Europe were being managed that way in late 2004. To manage this limited transmission capacity, different approaches are applied. The most important ones in EU countries are: ● ●

●

priority list (first come, first served; then “grandfathering”); pro-rata rationing (capacity is allocated in proportion to request if they exceed available capacity); auctions (the TSO accepts bids from potential buyers and allocates the capacity to the ones that value it most).

298

Electricity Market Reform Table 9.11. Type of access to cross-border transmission capacity. Connection

Type

FR–DE/DE–FR AT–DE/DE–AT NL–DE DE–NL FR–IT/IT–FR FR–BE/BE–FR NL–BE BE–NL FR–ES ES–PO/PO–ES SK–HU SL–IT CZ–AT CZ–DE AT–IT

Priority list/pro-rata/partly auctions Priority list Auction Auction Pro-rata Priority list/pro-rata Auction Auction Priority list New method foreseen Auction Auction (short term), pro-rata (long term) Auction Auction Pro-rata

Source: ETSO (2005). 100%

100%

110 100

90%

90

80%

80 70%

70% 57%

60 50 40

y 11% an

lic

m er

G

ep

ub

m

C

ze

ch

R

lg iu

Be

Au

st

ria

0

60% 50% 40% 30% 20% 10% 0%

an

10

Fr

20

22%

24%

29%

30

ce H 13% un ga ry 39% Lu Ita x em ly 8% Th e bo N u et he rg rla nd Po s 24% rtu ga Po l 10% la nd Sl 8% ov ak ia Sl 38% ov en ia S Sw pa itz in 4% er la nd

[GW]

70

NTC (GW) Installed generation capacity GW gross Import capacity as percent of installed capacity Fig. 9.25. Installed gross generation capacity, net transfer capacity (NTC) for transmission and import capacity as percentage of installed generation capacity in CE countries. Source: CEC (2005), UCTE (2005).

Table 9.11 depicts the type of congestion management methods for the most crowded EU borders. As can be seen from Table 9.12 currently there is a wide variety in methods. However, as the European association of TSOs–ETSO (2004) – states, “The Regulation 1228/2003 … on ‘Conditions for access to the network for cross-border exchanges in electricity’ clearly states that the implementation of market-based congestion management methods are preferred”. Figure 9.25 shows installed gross generation capacity, NTC for transmission and import capacity as percentage of installed generation capacity in EU countries. It clearly shows that


299

Average wholesale electricity price 2005 [ /MWh] Bottlenecks Market separation

29

58

52

28

46

32 47

47

32

64 58

Fig. 9.26. Transmission grid bottlenecks and wholesale electricity prices in EU 2005.

the import capacity as percentage of installed generation capacity is highest in the smaller countries. This figure also reveals the strategic relevance of Switzerland as a transit country. In absolute terms Switzerland has the highest NTC aside from the largest countries Germany and France.11 In Eastern countries many international transmission lines are currently being12 congested by long-term contracts which are taking up much of the potentially available capacity and can reduce the potentially competitive impact of market opening. Figure 9.26 exhibits the major cross-border bottlenecks and documents the corresponding wholesale spot market prices in 2004. 9.6. Future Outlook: Priorities for Improvements Today, the European Union has successfully initiated the most extensive and ambitious project for building a new electricity market. But there are no guarantees that the dynamics of this construction will not, as in the USA, dissipate, or that the internal market will not remain fractured in “national or local blocks” which may persist for a long time (Glachant and Lévêque, 2005; Glachant and Finon, 2005). Moreover, as has been argued by (Haas et al., 1997; Haas and Auer, 2001) the expectation of lasting competition in a “free” market is based on very simplified assumptions of the strategic behavior of electricity generators and network operators. The caveats described by Banks are similar (1996) (“the market is a wonderful thing and it should be exploited as far as possible but it also has its limits”) and Newbery (2002) that are based on the experience in the UK and the Nordic market (Norway, Sweden …).

11 ”Cross-border trade” is not necessarily correlated with “enhanced competition”: when the oligopolies own plants in several European countries, and exchange power between their subsidiaries, it might not contribute to more intensified competition. 12 Most of the transmission lines between countries are frequently congested, particularly those towards the importing countries of Austria, Germany and Hungary.

300


Currently, the major obstacle for a European common market is the general lack of competition in virtually all local and national wholesale, as well as retail electricity markets, because the number of competitors is too low, or because barriers to entry and incentives to collude remain too high.13 These aspects are further reinforced by (at least) two others: (1) insufficient transmission capacity availability between the sub-markets and (2) the increasing horizontal integration with natural gas supply. Hence, the paramount objective is still to construct competitive markets, while – at the same time – ensuring a reasonable level of grid reliability and adequacy of supply. 9.6.1. Access to the market One major priority for improving access to the EU grid is the regulation of the TPA in Germany. Pfaffenberger and Hille (2004) emphasizes this issue, especially for Germany, which has so far not regulated access to the transmission grid. The next priority is to obtain non-discriminatory, open and competitive balancing arrangements. Balancing arrangements may not handicap the arrival of new entrants or existing operators which are not vertically integrated; and they should be open to all potential competitive sources of supply (Glachant and Lévêque, 2005). Another important issue is transmission pricing. A harmonization of national access pricing schemes and cross-border pricing would contribute to lower transaction costs in international competition. This problem could be alleviated by reinforced regional cooperation agreements between TSOs (creating “virtual RTOs”). TSOs should not be authorized to stand as “national guards” protecting only the activities and interests of their historical zones of operation. Therefore, all TSOs wishing to play an active role on the regional level should be encouraged to engage in strengthened cooperation in order to smooth the functioning of the internal market. In order to do so, criterion for evaluating Europe’s economic interest in grid interconnections would be a crucial tool. It is untrue that only bilateral national interests form a legitimate basis for identifying and evaluating interconnection projects useful for expanding the EU internal market. Thus, we must seek criteria for evaluating a pan-European interest in these interconnections (Glachant and Lévêque, 2005).14 9.6.2. Remedies in restructuring utilities Of course, an easy solution with respect to the number of generators in each relevant market would be to have more generators and some divestment. Whatever the theoretical difficulty in designing these structural remedies (Smeers, 2005), there are currently no signs in any country which point in this direction. Most European governments like playing national champions and national mergers. Another issue is that privatization is often seen as being more important than carefully designed competition mechanisms. However, as Newbery (1998) asserted for England,

13

For example, the European heavy industry association writes in its 2004 electricity market design report: “Competition between European power generators and suppliers has virtually disappeared depriving industrial customers of any negotiating power when seeking new supply contracts” (PiE). 14 We can also think on secondary paths for improvement of TSOs work by extending the independence of TSOs; encouraging the harmonization of grid access and connection fees; and last by encouraging TSOs to develop joint forecasts and planning.


301

“competition rather than privatization is the source of the benefits”. And, under competitive pressure public utilities performed reasonably in the Nordic countries. Of particular relevance in this context is the ownership future of EdF. The privatization of EdF has been under discussion for years, and it could start in the Fall of 2005. However, given the limited number of generators engaged in this market, it is unlikely that a partial privatization of EdF would add much to the French “fringe competition” (Glachant and Finon, 2005). 9.6.3. Refining the regulatory or the market design In some countries, changes of the current design are under way. The most important changes appear to be the introduction of a regulator in Germany and a possible structural change of the Spanish market. A further one could be the creation of an organized Western European market which would couple The Netherlands, Belgium and France. Germany has been installing a regulatory authority in 2005. On June 17th, the German parliament passed extensive amendments to the existing Energy Industry Act. The government also outlined detailed guidelines on electricity grid access and grid tariff calculation methods. The new act provides rules on legal, operational and accounting unbundling mirroring the provision of the EU Directive (PiE N455, July 2005). For further details on this issue see Brunekreeft and Bauknecht (2006). The second interesting development took place in Spain. In the summer of 2005, the long awaited White Paper on the electricity sector was presented by Perez-Arrriaga. Although, this document is officially non-binding, it is seen as the ‘road map’ for the future with radical solutions to reduce market concentration. The White Paper is likely to propose a series of measures to limit market domination by the incumbent utilities with the target of increasing tariff transparency, competition and, straightforward, consumer benefit. Furthermore, it suggests virtual divestment of generation capacities. It remains to be seen as to how it will be applied, considering the current take over of Endesa by Gas Natural. Major changes are expected in Italy and France. This will intensify the use of auctionning in allocating cross-border capacity. For the future of the Italian market, the price difference between its neighbor countries will be a key element for the integration of its market into the EU. Many operators are announcing to get into the construction of merchant lines for the direct import of electricity. Could this really be an alternative to the construction of new power plants, which is quite difficult in Italy because of the strong local oppositions? France itself could embed the allocation of its northern interconnection capacity in a market mechanism shared with Belgium and the Netherlands. This would couple the three national PXs (APX, Belpex and Powernext) as well as the three TSOs (Tennet, Elia and RTE). If this mechanism were to work well and be cost effective, it could pave the way to further reinforced regional cooperation within the EU, while maintaining – at least for a while all the existing national PXs and TSOs (Belmans et al., 2005). To enhance competition in Eastern Europe a deeper regional integration could be a way out of this world of currently small, segmented and distorted local markets. A regional approach to market design and restructuring would be an improved solution compared to the national individual approach taken by most countries. Companies that are large on a national basis would be small, or at most medium sized, on a regional scale. Effective regional markets could offset the limited competition within national markets, but require suitable cross-border and balancing arrangements. The limitations of this approach are, that by increasing the relevant market size all indicators would look better, without any change in competitive settings. Kaderjak (2005) argues, that “… the finding that local competitive fringes are massive importers even in large exporting countries indicates the outstanding importance cross-border

302


trade may play in further market integration in the region”. Most important for such a competitive fringe are the low transaction costs for access to the grid and to the market place. An important future impact on the CE electricity markets might result from a further extension into other countries in Eastern and South-Eastern Europe. Slovenia is operating in a joint control block with Croatia and parts of Bosnia and Herzegovina. Bosnia and Herzegovina, as well as Serbia and Montenegro and Kosovo, are recovering from war damages and form again an electricity exporting area. Currently, many transmission lines are congested, see Table 9A.3 (Appendix). Moreover, the situation with respect to competition in the Eastern border countries is not promising. While in Poland and Hungary there are about three large generators, there is only one in the Czech Republic and Slovakia. Yet, it has to be considered that excess generation capacity exist in Poland, Czech Republic (Fig. 9.6) Bulgaria, Romania and Ukraine (see Auer et al., 2005). Hence, if the transmission system is extended in and between the crucial countries along the former EU-15 countries easterly border, there could be the chance for an extension of the current Central European market to the East. Therefore a core issue in improving the “EU Internal Market Design” is congestion management at interconnections. In practice, European interconnections are always treated like borders, and their congestions result from domestic decisions and priorities decreed separately in each member state. There is no comprehensive operational cooperation to minimize congestions at the borders, or to maximize the capacities available at the interconnections (Glachant and Pignon, 2005). A possible engine to do so is voluntary regional agreements. Since the institutional framework makes it difficult to quickly establish a fully operational regulation covering the entire European Union, pursuing comprehensive voluntary regional agreements could constitute an excellent auxiliary engine for the current phase.15 Moreover, during the current phase, let’s say 2005–2009, construction of the internal market could continue to advance in a decentralized framework in which national regulators could play a key role. We may consider that the problems of the internal market can best be addressed where they actually arise, which is what regulators already know how to do within their “national blocks”. Problems of unification and convergence between member states are most pertinent where trade is greater, interconnections most sought after, and wholesale market prices already tend to converge. Voluntary regionalization of convergence between some pioneering “national blocks” thus appears to be a promising step in the right direction during the current phase (see Glachant and Lévêque, 2005). Finally, it must be emphasized that the minimum requirement for more competition in the EU electricity markets is increased transparency concerning power plants and amounts generated. This relevant market information must be made available to all market participants in a simple way at low costs. As Jamasb and Pollitt (2005) state, “In the post-liberalisation era, some types of data have been deemed commercially sensitive and are not made available even to regulators. There is a need for adequate disclosure, more transparency, and the collection and publication of new types of data. … Improving the quality of data requires joint efforts and agreement on types of data needed, collection methods, and standard reporting formats”.16 Moreover, the issuing of licenses for 15

Harmonization – especially regional – is here a crucial issue: Harmonization of transactions so as to open a European bilateral market (“European purchases and sales passport”); Harmonization of transactions for reciprocated opening of organized markets (“virtual EuroPX”); Harmonization of rules for reciprocated opening of balancing mechanisms (“Balancing club”); Harmonization domestic mechanisms for fostering priority energy sources(see Glachant and Lévêque, 2005). 16 This provides a comprehensive discussion and a detailed overview on transparency in European electricity markets. And shows that especially the information on the availability of power plants is very poor in CE.


303

generators (Newbery, 2002) could ensure more transparent capacity availability and avoid sudden bottlenecks in generation. 9.6.4. Perspectives for adequacy, reliability and security of supply: generation and transmission capacity As argued in Section 9.4, the development of adequate capacity in generation is most important in this context, and the question as to how capacity margin is distributed among generators Figure 9.27 depicts the currently looming developments of load and generation capacity.17 This picture is not the same in different countries. In Italy, load has already surpassed available net capacity. In Spain and Portugal the danger of shortages exists (Alba, 2003; Crampes and Fabra, 2005): “With no plant entering into operation from 1998 to 2002, and a steep increase in demand … the system has indeed been operating below acceptable adequacy since 2000”. In Western Europe (FR, DE, CH, AT) the current trend implies generation capacity needs by 2007 or 2008. Eastern Europe (CZ, HU, PL, SK, SL) has adequate generation capacity for the foreseeable future, and will continue to be heavily inclined toward coal and nuclear power. Furthermore, domestic production of coal would make it politically awkward to reduce the share of coal-fired generation, despite environmental concerns. It seems that necessary environmental investments will be covered by customers through electricity price increases. As most of the power plants, like the coal companies, are still state owned, they are not much interested in switching to other fuels like gas. Additionally, higher gas prices are not attracting many investors. Nuclear power is controversial, but does have environmental benefits in a carbon-constrained world. Concerning the limited support for renewable generation, regulatory and political uncertainty has also prevented more than small-scale deployment of renewable technologies. The one remaining major uncertainty in Eastern European countries is the magnitude of demand growth. Another important prerequisite for a sufficiently wide market is the sufficiency of transmission capacity for neighbor regions, and the increasingly number of potential competitor generators. Currently, transmission constraints are having a substantial impact on the separation of sub-markets in Continental Europe. Hence, the basic conditions to bring about a European-wide electricity market are an extension of the grid at its bottlenecks,18 and a non-discriminating, open and comparable access to the transmission grid at reasonable non-pancaked rates. Or, as Newbery (2003) puts it “… to rapidly increase transmission capacity offered at efficient prices”. 9.6.5. The future of regulatory governance As Newbery (2002) states “(so far) Appropriate regulation has been largely ignored by the Commission and many EU countries, but without it, there are serious risks that the benefits of liberalization may be lost, and the political costs of flawed outcomes may undermine support for reform.”

17

The figures for load forecast are taken from UCTE (2005(a\b\c)). The figures for the trend in generation capacities are based on existing capacities, approved new capacities, decommissioning of nuclear due to IAEA and a limited lifetime of fossil plants of 40 years. 18 Yet, extending CB TM capacities faces – aside from potential acceptance problems – two other conditions: (i) Who will invest? (ii) How can the recovery of investments be ensured?

304

Electricity Market Reform DE⫹FR⫹AT⫹CH: Trends in load versus generation capacity

90

Gross capacity Load

240

80

Load

Gross capacity

70

200

60

160

Net capacity

GW

GW

Italy: Trends in load versus generation capacity

120

50

Net capacity

40 30

80

20 40

10

0 1995 (a)

2000

2005

2010

2015

2020

0 1995 (b)

Spain and Portugal: Trends in load versus generation capacity 70

70

Gross capacity

30

60

Net capacity Load

20 10 2005

2010

40

2015

2020

Net capacity

2020

Load

30

10 2000

2015

Gross capacity

40

20

0 1995 (c)

2010

50 GW

GW

40

2005

Eastern Europe: Trends in load versus generation capacity

60 50

2000

0 1995 (d)

2000

2005

2010

2015

2020

Benelux: Trends in load versus generation capacities Gross capacity Load

35 30 GW

25

Net capacity

20 15 10 5

(e)

0 1995

2000

2005

2010

2015

2020

Fig. 9.27. (a) Current and future trend of generation capacity and load in Western Europe. (b) Current and future trend of generation capacity and load in Italy. (c) Trends of generation capacity and load on the Iberian peninsula. (d) Trends of generation capacity and load in Eastern Europe. (e) Trends of generation capacity and load in Benelux.


305

The current European regulatory governance is a decentralized framework that is essentially national and in an incomplete process of convergence. If there were to be only one way of implementing competitive reforms, or only one way of transposing the First and Second EU Directives, then this decentralization would simply capture economies in transaction costs. Referees would be better and more cost effective if they were to remain on the field until the end of the game, rather than changing country at halftime. However, there are several legitimate paths to competitive reform. And there are yet other paths, if less legitimate, but still legally possible, owing to this very institutional decentralization and the “flexible” compromise that is characteristic of European directives. Nor is there featured, within the European context, a centralized regulator who could create additional complementary rules to steer national rules toward convergence, or a federal regulator with the power to legitimize national rules ex-ante or launch ex-post reviews for decisions taken on the ground (see Glachant and Lévêque, 2005). Thus, pan-European convergence between national blocks is being sought by other means. The best known is the process of voluntary agreements between the stakeholders: the Florence Forum, the Madrid Forum. This is a self-regulatory process, but different from the German one since it integrates national regulators. Relevant authorities and stakeholders voluntarily meet to establish principles or rules that, though not binding, delimit a “code of good conduct”. Nevertheless, when the underlying dynamics appeared to have lagged, the Commission sought to reboot it with a Second Directive (and regulation; in 2003) which contained national divergence and bolster convergence. Table 9.12 depicts the competences and the power of the regulators in CE countries in 2004 based on Green et al. (2005). It can be seen that no regulator scores with the maximum of power (6). The major power curtailments are the limited power to enforce competition and the still widely prevailing ministerial involvement.19 The European Union still appears to be in its infancy in matters of detecting and remedying market power in the field of energy. The sector enquiry started by the EU Competition Authority in Summer 2005 was then the very first step in a more systematic approach to the many particularities of market power in the electricity industry. A more or less permanent arrangement exists for detecting market power in some of these markets in a few countries of the European Union – but not in all of them – and even more rarely, is an array of organized remedies (Glachant and Littlechild, 2004). The implicit assumptions appear to be that either: ●

●

existing markets function sufficiently well and ongoing monitoring would be a waste of valuable time on a non-priority activity; or detecting and correcting eventual anomalies is not very difficult, so that any problem will reveal itself spontaneously in a timely fashion.

To go further in monitoring and mitigating market power, it would be useful to ask the European Competition Authority to create with national Competition Authorities and regulators having authority (many of them have no power in the competition area) a European market surveillance network sharing the data, the tools and the knowledge. Finally, European regulatory could also be enhanced. We think that a good workable solution would be to encourage bilateral and regional harmonization agreements between regulators.20 National regulators currently hold the institutional keys for the resumption of

19

In Germany and Switzerland no regulator existed in 2004. Rules for reserves and balancing, access to interconnections and congestion management, joint approval of investments in the grid, etc. 20

306


Table 9.12. Competences and power of the regulators. Power to enforce competition (Yes ⫽ Strong)

Power to regulate Ex-ante/ Ministerial the network Ex-post involvement access (Ex-ante ⫽ (No ⫽ (Yes ⫽ Strong) Strong) Strong)

Power to settle disputes (Yes ⫽ Strong)

Power to acquire information (Yes ⫽ Strong)

Summary: Number of strong (maximum: 6)

AT

Advisory

Ex-ante

General guidelines

Yes

Yes

Yes

5

BE

Advisory

Ex-ante

No

Yes

Yes

Yes

5.5

CZ

No

Ex-ante

No

Yes

Yes

Yes

5

DE

N.A.

N.A.

N.A.

N.A.

N.A.

N.A.

0

FR

Yes

Ex-ante

Tariff approval

Yes

Yes

Yes

5

HU

Advisory

Ex-ante

Tariff approval

No

Yes

Yes

4

IT

Advisory

Ex-ante

General guidelines

Yes

Yes

Yes

5

LU

No

Ex-ante

Yes

No

Yes

Yes

3

NL

Yes

Ex-ante

Instructions

Yes

No

Yes

4.5

PL

Advisory

Ex-ante

Supervision

Yes

Yes

Yes

5

PO

Advisory

Ex-ante

No

Yes

Yes

Yes

5.5

SK

No

Ex-ante

No

Yes

Yes

Limited

4.5

SL

Advisory

Ex-ante

Non-eligible

Yes

Yes

Yes

5

SP

Advisory

Ex-ante

Yes

No

Yes

Yes

3.5

CH

N.A.

N.A.

N.A.

N.A.

N.A.

N.A.

0

Source: CEC (2004); CEC (2005) personal national information.

regional construction in the internal market. Notably, they have the competencies to combine significant advances on the regional scale with the true constraints of the national level. The Commission’s challenge will be to motivate them in the mobilization of these competencies for the benefit of the Community rather than for the status quo. One of way of doing this is to develop pan-European regulatory knowledge and training in the European Union. The time has come to organize the expertise of regulatory personnel on the scale of the European Union, whether it be to disseminate existing knowledge to improve the yield to investments in human capital or to increase the efficiency of regulation (see Glachant and Lévêque, 2005).21 9.6.6. The future of environmental issues There are three fundamental environmental issues with respect to electricity markets in CE: energy taxes, emission trading and the promotion of RES. 21

The European University Institute in Florence is already a reference point for European regulation and a meeting place for the EU regulators, and could provide the basis for an ambitious project.

307

Competition in the Continental European Electricity Market CO2-index of EEX (Germany)

25 20 15 10

04.10.2005

04.09.2005

05.08.2005

06.07.2005

06.06.2005

07.05.2005

07.04.2005

08.03.2005

06.02.2005

0

07.01.2005

5

08.12.2004

Price of certificate ( /tonneCO2)

30

Fig. 9.28. Prices of CO2 emission certificates in CE 2004–2005. Source: www.eex.de

Energy taxes on electricity consumption for households exist in most CE countries. There are no signs of significant changes of this instrument. Emission trading in CE countries started in 2005. The first major response was quite a steep increase in CO2 certificate prices, and later on, electricity prices, Figures 9.28 and 9.17. Of course, electricity prices will be influenced by the market price of CO2 certificates. (For further details on CO2 matters see Brunekreeft and Bauknecht, 2006). It is still too early to say to which extent. Two major open questions still exist with respect to RES electricity generation. Firstly, will the EU finally come to the conclusion that a harmonization of different promotion schemes is preferable? So far, countries with feed-in tariffs (Germany, Spain, Austria) have been performing much better in adding more capacity than countries which have been relying on TGC-based quota systems (UK, Belgium, Italy). Secondly, will the EU stick to ambitious targets for the period after 2010, despite the nonattainment of certain countries concerning targets for 2010? 9.6.7. Summing up: perspectives for delusion or for competition? To bring about the EU’s goal of effective competition in a single integrated European electricity market and to avoid market power, the following structural conditions have to be fulfilled: ●

●

●

Complete ownership separation of the grid from generation and supply in all EU countries. Adequate transmission capacity for connecting the single sub-markets thus creating a larger market with more potential competing players. This is also important for a continuing integration of Eastern and Southern-Eastern European countries. Of course, most ideally, would be a harmonized system which provides the same conditions for national and cross-border electricity exchanges. Adequate spare capacity in generation: now, it is of importance that incentives for investments in sufficient capacity are provided by the markets and encouraged by the regulatory authorities. Note that adequate capacity can also be brought about by proper

308

●

●


demand-side load management, in particular by smart metering which would allow for new forms of demand responsiveness. Adequate spare capacity in generation is not enough if concentration is too high; therefore actual disinvestment as well as temporary “Virtual Disinvestment” (Virtual Power Plants by auctioning rights to dispatchable energy) could restructure the generation sector, so as to be sure that more generators and more suppliers can compete, even at the fringe, in every single sub-market. Full market opening in all countries (notably by ending uncompetitive vested contracts, and by establishing a sensible temporary price control for incumbent monopolies).

We know that these conditions look like a wishful Christmas list, but none of us prefer nightmares.

9.7. Conclusions While the liberalized CE electricity market is still under construction, some conclusions regarding developments so far can already be drawn. ●

●

●

●

Firstly, liberalization in CE started about a decade after the advances made in the UK and Norway. However, it seems that the CE countries did not learn much from their experience regarding conditions for competition. Instead of divesting generation capacity and increasing the number of competitors (as recommended by Newbery and Pollitt, 1997) most countries pursued mergers (DE, NL), retained oligopolies (NL, ES, AT, CH), private monopoly (BE), or supported the concept of national champions (PO, FR). Only Italy has chosen a quite different strategy of divestment of the former national champion ENEL. Secondly, the CE electricity market is the largest regional market in Europe, and its geographical position implies that further progress toward an integrated electricity market in Europe will depend strongly on the development of this market (Jamasb and Politt, 2005). France and Germany play a key role within this market because of their size and geographically central positions. Thirdly, the major obstacle for a common market that works reasonably, is currently, a general lack of competition in virtually all local and national wholesale as well as retail electricity markets. The number of competitors is to low, or barriers to entry are too high, or incentives to collude are too high. This aspect is being reinforced by two others: insufficient transmission capacity is available between the sub-markets; an increasing horizontal integration with natural gas supply. Fourthly, the EC itself is in an ambiguous position. On the one hand, it still advocates the goal of a European-wide common electricity market, by the year, it is said, 2012. On the other hand, only very weak light-handed measures are being implemented on the European scale. One of the major problems is still, and will be, that the market power of the large – and still growing – incumbent generators cannot be tackled by the EC alone because it cannot ask for deep structural or regulatory remedies. The second one is the behavior of TSOs being not unbundled from generation or from the interests of their national block of stakeholders. The EC acts weakly because it would require severe interferences in the Member States’ institutions and policies. Only the European Competition Authority (DG COMP) and the European Court of Justice have some power to pull national governments and national entities out of their retrenchments. How it can be done, is still to be seen.


309

So currently it is not likely that the measures described above will be implemented. As Newbery (2002) argued “the EU lacks the necessary legislative and regulatory power to mitigate generator market power. Unless markets are made more contestable, transmission capacity expanded and adequate generation capacity ensured, liberalization may lead to higher prices”. The national governments pursue quite different objectives. In some countries it is obvious that, so far, governments support their national utilities, and are not eager to introduce effective competition. Hence, it cannot be taken for granted that one integrated European electricity market will ever emerge. A second-best solution would be to foster competition in regional sub-markets by incremental reduction of barriers to cross-border trade as well as to the inclusion of generators and suppliers, thus paving the way to more confrontation between electricity and gas companies in the dual fuel markets. A very ambiguous role is played by privatization. On the one hand, there is currently a strong majority in Europe who see privatization as the politically correct solution regarding ownership.22 On the other hand, privatization frequently means only the maximize of the market value of the shares sold to the buyers, being … the large incumbent players (the “seven brothers” depicted by (Thomas, 2003). This problem partially applies to EdF, the most important looming privatization case. Of course, the French government is not looking to reducing the potential value of its EDF shares (50–60 billions of euros). Therefore, it has no economic incentive in strengthening competition at home and it prefers, instead, strengthening the position of its own champion in France as well as in the EU markets. Finally, it is stated that currently in most regions there are still sufficient spare capacities in generation and transmission available. The definitive litmus test for liberalization will come in every sub-market in CE at the point-of-time when the bulk of excess capacities has disappeared and demand has come close to available capacities. That is to say, the most important problem is to provide long-term incentives for investments in the upgrade and in new generation and transmission capacities, as well as in demand-side efficiency and demand responsive measures. This issue is especially relevant in the context of decentralized – versus – further centralized development of the electricity supply system.

●

●

Acknowledgments Findings or work presented in this chapter has been financed within projects of the EC (SESSA, a 6th FP R&D project) and the Austrian National Bank (ÖNB). Of course views expressed here are under the sole responsibility of the authors and absolutely not of these institutions. The authors are grateful for valuable contributions and discussions to: Hans Auer, Julian Barquin, Ronnie Belmans, Paul Joskow, Peter Kaderjak, Jacek Kaminski, Jaroslav Knapek, Stephen Littlechild, Arturo Lorenzoni, Miroslav Maly, David Newbery, Ignacio PerezArriaga, Wolfgang Pfaffenberger, Pippo Ranci, Gustav Resch, Alena Salamonova, Fereidoon P. Sioshansi, Yves Smeers, Maria Isabel Soares, Miha Tomsic, Aviel Verbruggen, Twan Vollebregt. However, the authors are solely responsible for the view expressed and any remaining error.

22

Note, that this is not the opinion of all authors of this chapter.

310


References Alba, J. (2003). Investment in gas-fired generation, IEA/NEA Workshop on Power generation Investment in liberalized electricity markets, March 25, 2003, Paris. Auer, J., Keseric, N., Haas, R., Resch, G., Huber, C. and Faber, T. (2005). Medium- and long-term effects of EU-electricity enlargement. In C. von Hirschhausen and J.-M. Glachant (eds.), Proceedings of the 3rd SESSA conference Perspectives and Challenges of EU Electricity Enlargement, Berlin. Banks, F. Economics of electricity deregulation and privatisation – an introductory survey. Energy – The International Journal 1996, 21(1), 173–187. Belmans, R., Glachant, J.-M. and Meeus, L. Regional electricity market integration. KU Leuven and University Paris XI, France, Belgium and Netherlands, September 2005. Bergman, L., Brunekreeft, G., Doyle, C., von der Fehr, N.-H., Newbery, D.M, Pollitt, M. and Regibeau, P. (1999). A European Market for Electricity? Center for Economic Policy Research, London, UK. Brunekreeft, G. and Bauknecht, D. (2006). Energy policy and investment in the German power market (This book). CEC (Commission of the European Communities) (2004). DG TREN Working Paper: Third Benchmarking Report on the Implementation of the Internal Electricity and Gas Market, Brussels. CEC (Commission of the European Communities) (2005). Annual Report on the Implementation of the Internal Electricity and Gas Market, Brussels. Codognet, M.K., El Khodary, M., Glachant, J.M., Hiroux, C., Mollard, M., Leveque, F., Perez, Y. and Plagnet, M.A. (2005). Mergers and Acquisitions in the European Electricity Sector. CERNA and GRJM, Paris. Crampes, C. and Fabra, N. (2005). The Spanish electricity industry: plus ca change…. The Energy Journal Special Issue on European Electricity Liberalisation, 127–146. De Paoli, L. (ed.) (2001). The Electricity Industry in Transition. FrancoAngeli, Milano. ETSO (2004). An Overview of Current Cross-border Congestion Management Methods in Europe, http:// www.etso-net.org ETSO (2005). Benchmarking on Transmission Pricing in Europe: Synthesis 2003, http://www.etso-net.org European Commission (1997). Directive 96/92EC of the European Parliament and of the Council Concerning the Common Rules for the Internal Electricity Market. Official Journal L27 of the 1/30/1997. European Commission, Luxemburg. Finon, D. (2003). Introducing competition in the French electricity supply industry: erosion of the public hierarchy by the European institutional integration. In J.-M. Glachant and D. Finon (eds.), Competition in European Electricity Markets. Edward Elgar, Northhampton, Massachusetts, pp. 257–286. Finon, D. and Perez, Y. (2006). Transactional efficiency and public promotion of environmental technologies: the case of renewable energies in the electric industry. Ecological Economic Review (forthcoming). Glachant, J.-M. (2005). Implementing the European internal energy market in 2005–2009, a proposal from academia. Working Paper European Energy Institute and Groupe Réseaux Jean Monnet – ADIS at University Paris XI, www.grjm.net Glachant, J.-M. and Finon, D. (ed.) (2003). Competition in European Electricity Markets. Edward Elgar, Northhampton, Massachusetts. Glachant, J.-M. and Finon, D. (2005). A competitive fringe in the shadow of a state monopoly: the case of France. The Energy Journal, Special Issue on European Electricity Liberalisation, 181–204. Glachant, J.-M. and Lévêque, F. (2005). Electricity Internal Market in the European Union: What to Do Next? Contribution to SESSA Research Project, September, www.sessa.eu.com Glachant, J.-M. and Littlechild, S. (2004). Etude comparative des pratiques de surveillance des marchés électriques étrangers, Rapport pour la Commission de Régulation de l’Energie, Paris. Glachant, J.-M. and Pignon, V. (2005). Nordic congestion’s arrangement as a model for Europe? Physical constraints vs. economic incentives. Utilities Policy, 13, 153–162. Green, R., Lorenzoni, A., Perez, Y. and Pollitt, M. (2005). Chapter 2 in Glachant, J.M., and Léveˆque, F. (eds), Electricity Reform in Europe: Towards a Single Energy Market, Edward Elgar, forthcoming 2006. Haas, R. and Auer, H. (2001). How to ensure effective competition in Western European electricity markets. IAEE-Newsletter, 3, 16–20.


311

Haas, R., Orasch, W., Huber, C. and Auer, H. (1997). Competition versus regulation in European electricity markets. Proceedings of the 2nd European IAEE Conference, July 2–4, Vienna, Austria. Haas, R., Auer, H. and Stadler, M. (2002). Introducing Competition in the Western European electricity market: A Critical Review, ENER-Bulletin 24.01 on ENER Forum 2: Monitoring the progress of the implementation of the EU Gas and Electricity Directives: Are European markets becoming competitive? Pitney Bowes Management Services Denmark. Huber, C., Faber, T., Haas, R., Resch, G., Green, J., Ölz, S., White, S., Cleijne, H., Ruijgrok, W., Morthorst, P.E., SkyHe, K., Gual, M., Del Rio, P., Hernandez, F., Tacsir, A., Ragwitz, M., Schleich, J., Orasch, W., Bokemann, M. and Lins, C. (2004). Action plan for a dynamic RES_E policy, EEG wien. IEA (2005). Energy Prices and Taxes, 2nd quarter 2005. International Energy Agency, Paris. Jamasb, T. and Pollitt, M. (2005). Electricity market reform in the European Union: review of progress toward liberalization & integration. The Energy Journal Special Issue on European Electricity Liberalisation, 11–49. Kaderjak, P. (2005). A comparison of electricity market models of CEE new member states, mimeo, Corvinus University of Budapest, 1–37. Lorenzoni, A. (2003). Institutional and organisational reform of the Italian electricity supply industry: reconciling competition with the single tariff. In J.-M. Glachant and D. Finon (eds.), Competition in European Electricity Markets. Edward Elgar, Northhampton, Massachusetts, pp. 311–326. Mez, L. (2003). New corporate strategies in the German electricity supply industry. In J.-M. Glachant and D. Finon (eds.), Competition in European Electricity Markets. Edward Elgar, Northhampton, Massachusetts, pp. 193–216. Muesgens, F. (2004). Market power in the German wholesale electricity market – an analysis of marginal costs and prices. Proceedings of the 6th IAEE European Conference, Zurich. Newbery, D. (1998). Freer electricity markets in the UK -– a progress report. Energy Policy, 26(10), 743–749. Newbery, D. (2002). European deregulation: problems of liberalising the electricity industry. European Economic Review, 46, 919–927. Newbery, D. (2005). Electricity liberalisation in Britain: the quest for a satisfactory wholesale market design. The Energy Journal Special Issue on European Electricity Liberalisation, 43–70. Newbery, D. (2006). Electricity liberalisation in Britain: and the evolution of market design, This book. Newbery, D. and Pollitt, M. (1997). The restructuring and privatisation of Britain’s CEGB: Was it worth it? Journal of Industrial Economics, 45(3), 269–304. OECD (2005). National Accounts of OECD countries, OECD, Paris. Parthenay, C.D. and Perez, Y. (2005). Environnement institutionnel et trajectoire des entreprises, une analyse northienne de l’industrie électrique. Numéro Spécial de la Revue de Management International, Mai. Pfaffenberger, W. and Hille, M. (2004). Investitionen im liberalisierten Energiemarkt: Optionen, Marktmechanismen. Bremer Energieinstitut, Rahmenbedingungen. Smeers, Y. (2005). How can one measure market power in restructured electricity systems? Contribution to SESSA Research Project, May, www.sessa.eu.com Soares, M.I. (2003). The Iberian electricity market: towards a common market? In J.-M. Glachant and D. Finon (eds.), Competition in European Electricity Markets. Edward Elgar, Northhampton, Massachusetts, pp. 327–350. Thomas, S. (2003). The seven brothers. Energy Policy, 31(5), 393–403. UCTE (2005a). Online Database, www.ucte.org UCTE (2005b). Monthly Reports, www.ucte.org UCTE (2005c). System Adequacy Forecast Report. van Damme, E. (2005). Liberalising the Dutch electricity market 1998–2002. The Energy Journal Special Issue on European Electricity Liberalisation, 155–180. Verbruggen, A. and Vanderstappen, E. (1999). Electricity sector restructuring in Belgium during the 90s. Utilities Policy, 8, 159–171.

312

Appendix Table 9A.1. Population, GDP, electricity consumption and generation (by source) in CE countries in 2004.

Country

8.1 10.4 10.2 61.5 82.5 10.1 58.1 0.5 16.2 38.2 10.4 5.38 2.0 40.8 7.4

GDP (billion USD 2003) 255.2 304.2 90.4 1758 2402 82.8 1468.3 27 512.7 209.5 146.8 32.7 19.0 840.5 321.8

Net consumption (TWh) 51.8 87.5 61.4 445.1 554 38.2 322 6.3 110.9 144.8 45.5 26 12.3 234.5 60.4

*Numbers on hydro include the pumped storage generation. Source: OECD (2005); UCTE (2005), national homepages.

Total generation (TWh) 56.5 81.4 77.9 548.2 570.1 31.0 300.4 4.0 94.1 154.1 39.4 28.9 13.2 257.1 64.5

Hydro* (TWh)

Other renewables (TWh)

Thermal (TWh)

Nuclear (TWh)

37.6 1.6 2.5 64.5 26.7 0.2 48.7 0.9 0.1 3.2 9.9 3.5 2.7 33.3 35.1

0.9 1.2 0 1.5 25 0.2 7.27 0 4 1.2 1.7 0 0 15 1

17.9 33.7 50.6 55.4 360 19.4 244.4 3.1 86.4 149.7 27.8 9 4.5 147.9 3

0 44.9 24.8 426.8 158.4 11.2 0 0 3.6 0 0 16.4 5.2 60.9 25.4

Imports (TWh) 19 14.6 9.8 7 45.8 11.4 51.5 6.1 20.8 5.3 3.1 1.8 2.9 9.5 27

Exports (TWh) 13.4 8.3 25.5 73.1 53.8 4.5 0.5 2.8 3.8 14.6 0.3 4.1 3.8 8.2 30.2


AT BE CZ FR DE HU IT LU NL PL PT SK SL ES CH

Population (Mio)

Country

Hydro (MW)

Other renewables (MW)

Nuclear (MW)

Thermal (MW)

Total gross capacity (MW)

Available net capacity (MW)

Peak load (MW)

AT BE CZ FR DE HU IT LU NL PL PO SK SL ES CH

11,700 1,413 2,128 25,110 9,895 48 20,499 1,128 37 2,192 4,512 2,429 840 18,241 13,200

670 248 11 950 16,460 28 2,747 43 1,228 73 572 2 0 6,899 300

0 5,801 3,760 63,400 20,643 1,755 0 0 449 0 0 2,640 670 7,694 3,200

5,900 8,206 10,526 26,920 79,533 5,657 55,112 474 19,251 29,451 6,571 2,900 1,262 31,098 600

18,270 15,668 16,425 116,380 126,531 7,998 78,358 1,645 20,965 31,716 11,655 8,059 2,772 63,932 17,300

13,446 12,700 11,716 84, 016 79,989 5,811 48,148 1205 16,408 25,511 8,137 5,227 2,185 40,961 12,278

8,962 13,708 10,157 81,400 77,200 6,012 53,606 994 15,601 21,146 8,261 4,319 2,006 37,724 9,656

Source: UCTE (2005).


Table 9A.2. Electricity generation capacity (by source) in CE countries in 2004.

313

314

Table 9A.3. Transmission capacity and physical flows 2004 for connections/directions with more than 35% use. Summer power flow maximum (MW)

Percent used Winter

Percent used Summer

(GWh) 2004

(GWh) maximum

2,300 3,000 1,400 1,600 3,800 2,650 2,250 2,800 1,200 220 1,650 700 1,700 1,100 650 1,720 1,100 380 1,400 700

767 1,194 585 1,295 3,664 1,758 466 3,341 642 144 1,733 383 1,387 1,095 220 795 1,161 889 1,246 1,263

2,393 1,296 876 1,156 1,874 1,944 1,701 2,232 571 271 1,233 254 1,731 728 304 480 1,010 640 849 1,114

33.3 39.8 41.8 80.9 96.4 66.3 20.7 119.3 53.5 65.5 105.0 54.7 81.6 99.5 33.8 46.2 105.5 233.9 89.0 180.4

104.0 43.2 62.6 72.3 49.3 73.4 75.6 79.7 47.6 123.2 74.7 36.3 101.8 66.2 46.8 28.0 91.8 168.4 60.6 159.1

15,482 11,830 4,465 8,922 17,357 17,125 7,597 19,915 4,419 1,621 9,154 3,158 13,116 6,248 2,002 6,045 8,546 6,180 6,034 8,523

20,148 26,280 12,264 14,016 33,288 23,214 19,710 24,528 10,512 1,927.2 14,454 6,132 14,892 9,636 5,694 15,067.2 9,636 3,328.8 12,264 6,132

Source: ETSO, UCTE (2005).

Percent used 2004 76.8 45.0 36.4 63.7 52.1 73.8 38.5 81 42.0 84.1 63.3 51.5 88.1 64.8 35.2 40.1 88.7 185.7 49.2 139.0


FR – DE DE–CH AT–DE DE–AT DE–NL FR–IT FR–BE CH–IT AT–CH AT–IT PL–CZ DE–PL CZ–DE CZ–AT AT–SL CZ–SK SK–HU SL–IT FR–ES ES–PO

NTC (MW) 2005

Winter power flow maximum (MW)

Competition in the Continental European Electricity Market Table 9A.4. Countries’ acronyms. Acronym country

Country

AT BE CZ FR DE HU IT LU NL PL PO SK SL ES CH

Austria Belgium Czech Republic France Germany Hungary Italy Luxemburg The Netherlands Poland Portugal Slovakia Slovenia Spain Switzerland

315


PART IV North America, New World, New Challenges


Chapter 10 California Electricity Restructuring, The Crisis, and Its Aftermath1 JAMES L. SWEENEY Stanford University, Terman Engineering Center, CA, USA

Abstract In Chapter 10, James Sweeney provides an overview of California’s electricity restructuring in mid-1990s. California’s goal was to deregulate both the wholesale and the retail electricity market, after a transition period during which utilities would be able to recover stranded costs. Three regulatory rules were in place during the transition period: (a) regulators forced investor-owned utilities (IOUs) to divest at least 50% of their generating assets, (b) regulators precluded the utilities from entering forward contracts to acquire electricity, thus moving bulk power acquisition dominantly to spot markets, and (c) regulators imposed retail price caps on the utilities. In the midst of the transition period, the western electricity crisis, driven primarily by a sharp reduction in available hydropower and limitations on natural gas supplies, drove wholesale electricity prices throughout the western US to triple-digit or quadruple-digit levels. The three regulatory rules together proved disastrous, driving one utility into bankruptcy and another near bankruptcy, and left the state with billions of dollars of losses. Flaws in the California market design encouraged market gaming or exercise of market power by generators, traders, utilities, and the state itself. With the crisis over, California has now eliminated the over-reliance on spot markets, encourages long-term contracts, has established resource adequacy rules for utility electricity acquisition, and has eliminated retail price caps. An ISO-led market redesign project is underway to restructure transmission congestion management. Nevertheless, the future 1 A significant portion of this chapter includes material reprinted from The California Electricity Crisis, by James L. Sweeney, with permission of the publisher, Hoover Institution Press, copyright 2002 by the Board of Trustees of the Leland Stanford Junior University. My research was made possible through the Hoover Institution on War, Revolution and Peace, the Stanford Institute for Economic Policy Research, and ExxonMobil Foundation. I would like to thank Ziad Alaywan, Peter Arth, Severin Borenstein, Tom Casten, Ralph Cavanagh, John Cogan, Linda Cohen, Robert Crow, Elizabeth Farrow, Jim Harding, William Hogan, Hill Huntington, Jeffrey Jones, Paul Joskow, Lester Lave, Scott Lowe, Robert Naylor, Stephen Peck, Dmitri Perekhodstev, John Raisian, Gregory Rosston, John Shoven, Perry Sioshansi, George Shultz, Robert Weisenmiller and James Wilson for their helpful suggestions for The California Electricity Crisis and/or this chapter.

319

320


regulatory direction remains uncertain, with some advocates trying to force a return to a vertically integrated monopoly structure and others striving to improve regulatory rules within a partially regulated, partially deregulated system. 10.1. California’s Restructuring 10.1.1. California electricity system before restructuring Prior to 1998, California’s electricity system traditionally operated similarly to electricity systems throughout the US. It included three large IOUs, collectively selling most of the electricity in California. Each IOU had a franchise in one of three separate parts of the state – Pacific Gas and Electric Company (PG&E) in Northern and Central California, Southern California Edison (SCE) in Coastal, Central, and Southern California, and San Diego Gas and Electric (SDG&E) in San Diego. In addition were several2 much smaller IOUs , several electric co-ops, and numerous municipal utility systems, the largest of which were the Los Angeles Department of Water and Power (LADWP) and the Sacramento Municipal Utility District (SMUD). The IOUs served 78% of the California customers, the municipal utilities served 22%. Electric co-ops and the federal agencies collectively served less than 0.1% of the customers. In terms of total MWh of electricity, the IOUs sell 72%, the municipal utilities, 24%, and the federal agencies 3%. Each utility operated as a local monopoly, selling electricity in its own exclusive franchise area, with no direct retail competition from other electricity sellers. The three large IOUs, as well as some of the municipal utilities, were vertically integrated to include generation, transmission, and local distribution. Some municipal utilities operated as only local distribution companies; some included generation and transmission. Figure 10.1 shows the service areas of the various utilities as of 1996, while the restructuring was being debated. For IOUs, almost all significant financial decisions involving any of the three functions were subject to the jurisdiction and control of the statewide regulatory body, the California Public Utilities Commission (CPUC). Customers paid retail prices for electricity based on operating costs plus a regulated rate of return on the prudently incurred “used and useful” invested capital. The CPUC would review whether costs were prudent and would determine the “fair” rate of return on invested capital. Pricing was based primarily on cost of service and only secondarily on market conditions. The significant decisions made by the publicly owned municipal utilities were subject to the jurisdiction and control of their appointed or elected governing bodies. Thus, their strategies could be based on local decision-making, rather than on statewide regulations. However, they typically were operated so that over a span of several years their revenues roughly equaled their total costs of operation. Thus, for municipal utilities as well as for IOUs, pricing was based primarily on cost of service and only secondarily on market conditions. Figure 10.2 shows the mix of resource types used to generate electricity in California.3 The three largest sources of electricity are natural gas, nuclear, and hydroelectric. In 1998, at the 2

Smaller IOUs were PacifiCorp, Sierra Pacific Power Co., and Southern California Water Co. Source: California Energy Commission (CEC). The Commission provides this important explanation of a data change. “Prior to 2001, utility-owned shares of coal, nuclear plants and some firm contract generation outside California were considered part of utility-owned generation category. Since 2001 most of this data is included in the energy imports category. However, two coal-fired plants, Intermountain and Mohave, which are located outside of California, but inside LADWP and CAISO control areas, respectively, are still treated as utility-owned generation.” 3

California Electricity Restructuring, The Crisis, and Its Aftermath

321

Fig. 10.1. California utility service areas, 1996. Source: CEC.

time of the restructuring, 36% of electricity was generated using natural gas, 21% from hydropower, and 18% from nuclear. The 15% generated using coal is dominantly generated outside of California, but owned by California utilities. California has been a modest user of electricity, when measured on a per-capita basis. The industrial structure is not a heavy electricity consumer, relative to industry in other states. The State is blessed with a temperate climate and thus air conditioning loads are relatively low. Aggressive energy efficiency programs, coupled with high prices of electricity have kept consumption relatively low. California has had a long tradition of efficiency incentives and services that have helped to reduce California’s electricity intensity in recent decades. In 1999, right after the restructuring, the residential electricity consumption per customer4 was 37% below the US average during that year. 4

Data from US Department of Energy. http://www.eia.doe.gov/cneaf/electricity/esr/esrt14p4.html. Data are based on number of customers in the residential sector. This figure corresponds closely to the number of households.

322

Electricity Market Reform 30 28 26

Average MW (Thousands)

24 22 20 18 16 14 12

All Other Wind Organic Waste Geothermal Coal Hydroelectric Nuclear Gas

10 8 6 4 2 03

02

20

01

20

00

20

99

20

98

19

97

19

96

19

94

95

19

19

93

19

92

19

91

19

19

19

90

0

32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 01 20

00 20

99 19

98 19

97 19

96 19

95 19

94 19

93 19

92 19

19

19

91

Residential Commercial Industrial Agricultural & water pumping Other 90


Fig. 10.2. Electricity generation in California.

Fig. 10.3. Electricity consumption in California.

Figure 10.3 shows the electricity consumption over time for the various major consuming sectors. From 1990 to 2000, use of electricity increased from 26,000 MW average consumption rate to just above 30,000 MW, a growth of 16% over 10 years, or 1.4% per year. Growth in energy consumption, however, was somewhat faster from the 1997 through 2000 period, increasing by almost 2000 MW during the 3 years, or an average growth rate of 2.3% per year. Peak loads were growing at roughly the same rates.


Existing System

British Columbia

WA

Alberta

MidColumbia

Total Capacity in MW as of 1/1/00 Total MWs ⫽ 158,889 On Existing Transmission System

E. WA/ W. MT

MT

Colstrip

Snake

Oregon

SE ID

ID

JB

WY

N. NV

I PDC

Northern California

Utah

C

CHB

IPP D

Colorado

Central CA Nav Glen

S. NV

SP15

Coal Hydro Gas Wind Other

323

Southern California

Marketplace

Palo Verde

AZ

Four Corners

NM

Mexico

Fig. 10.4. Installed capacity by generation type and transmission in the western Grid, as of January 2000. Source: Western Governors’ Association.

Although the IOUs in California operated as vertically integrated monopolies, they did purchase some electricity from external sources. California utilities had long-term contracts to purchase hydroelectric power from the Bonneville Power Administration (BPA), a federal power-marketing agency, which sells power generated primarily from hydroelectric projects in the Columbia River. Both Munis and IOUs also had other contracts to purchase electricity from Federal projects. California traditionally sold electricity to Pacific Northwest entities in the winter, when Pacific Northwest demands peak and purchased electricity from the Pacific Northwest during the summer, when California demands peak. This trade was part of a larger pattern of interconnections throughout the west. A grid connects the western states, plus the Canadian provinces of British Columbia, and Alberta. Figure 10.4, created by the western Governors’ Association5 gives an indication of the transmission capacity between the various states and portions of states. The widths of the lines connecting the circles are proportional to the relative transmission capacities. There is much transmission capacity between California and the Pacific Northwest and between California, Nevada, and Arizona. These parts of the western region are linked particularly closely and the greatest flows of electricity occur among these areas. California represents 42% of the

5

“Conceptual Plans for Electricity Transmission in the West.” Western Governors’ Association, August 2001.

324


generation and 51% of the consumption in this tightly linked portion of WECC. The capacity to transmit electricity between the mountain states and the far west is more limited.

10.1.2. Changing federal regulations Several US regulatory changes had important impacts on California and set the stage for California’s restructuring. One of the most important was the Public Utility Regulatory Policies Act of 1978 (PURPA), enacted primarily to promote development of small-scale renewable sources of energy for electricity generation. Cogeneration was included as well, as a means of more efficiently converting primary energy into electricity and usable heat. PURPA mandated state regulatory commissions to establish procedures to require electric utilities to interconnect with and buy capacity and energy offered by any non-utility facility that qualified under PURPA. These so-called “qualifying facilities” or “QFs” were typically small generating facilities based on renewable energy, waste products, or natural-gas-fired cogeneration units. PURPA required utilities to pay a price for electricity from QFs equal to the “avoided cost” of electricity generation. This avoided cost was meant to be the totality of costs that a utility would avoid by purchasing electricity from these small alternative sources. PURPA left it to the state regulatory commissions to interpret what dollar price corresponded to “avoided cost” and the precise conditions under which the electricity and capacity must be purchased. In California, the CPUC took a very aggressive stance implementing PURPA. It set high prices for electricity purchased by the IOUs6 under PURPA, requiring them to sign contracts based on standard offers with guaranteed prices that rose sharply over time.7 The financial incentives and guaranteed market for QF electricity, coupled with tax incentives established by the Federal government, created a significant industry of renewable electricity generation in California, including wind farms and electricity generation from wood wastes. These policy changes also led to large increases in cogeneration capacity,8 which was largely natural gas fired. Figure 10.5 shows the composition of the new generation capacity in California coming on line between 1978 and the beginning of January 2000. The black component of the bars shows the plants with nameplate capacity 50 MW or greater; the grey component shows plants with nameplate capacity smaller than 50 MW. In the time between 1978 and the beginning of 2000, new nameplate capacity for electricity from cogeneration was as large as new conventional generation. The total new QF capacity built during this time exceeded the new

6

Since the CPUC did not regulate the municipal utilities, these high prices were not relevant to these entities. 7 Under Interim Standard Offer No. 4 (ISO4), a QF based on renewable energy could sign a contract based on a fixed forecast of future electricity price. Such a QF entering a contract would be guaranteed $57/MWh in 1985, $81/MWh in 1990 and $109/MWh in 1994. After 10 years the contract price reverted to the short run avoided cost, which typically would be far lower than the fixed price guarantee. Gasfired cogeneration units were not treated nearly as generously but were generally paid an annual average of about $25/MWh for capacity and about $25–$30/MWh for energy. 8 Cogeneration now is the single biggest source of PURPA electric generation capacity in California. Of the roughly 10,200 MW of total QF nameplate capacity in California in 2001, about 5700 MW is cogeneration, 4500 is generation from renewables such as wind or organic wastes. (Data from CEC database of electric generating plants on line in California.)


7000

MW nameplate capacity

6000

325

Greater than 50 MW Smaller than 50 MW

5000 4000 3000 2000 1000 0 QF cogen

QF renewables Cogen, not QF

Conventional

Nuclear

Fig. 10.5. New plant construction 1978 through January 2000.

capacity from conventional plus nuclear generation.9 The majority of the QF capacity was from small units (nameplate capacity smaller than 50 MW). However, with long-term contractual obligations to purchase electricity from QFs at a high cost, by the early-1990s the utilities were facing a high average cost of electricity generation. In addition, California utilities had invested in nuclear power plants, units whose construction costs turned out to be greater than initially predicted, further increasing the average cost of electricity generation. These factors together helped make California’s retail prices among the highest in the nation. Only in the states of California, Alaska, Hawaii, and the Northeastern states did average retail prices to residential customers exceed 8 cents/KWh ($80/ MWh). California’s average revenue in 1998 was 9 cents/ KWh ($90/ MWh). Another important effect of PURPA was some separation of ownership of generation from ownership of local retail sales. However, utilities still controlled all electricity transmission lines. A utility that wished to stifle competition in electricity generation could do so by refusing to allow its competitors to transmit electricity along its transmission lines. Creating a competitive market for electricity generation required federal officials to deal with issues of utility control of transmission lines. The Energy Policy Act of 1992 (EPACT) was an attempt to address this problem. Among its many provisions, EPACT opened access by non-utilities to the transmission networks. And in 1996, the Federal Energy Regulatory Commission (FERC) issued Order 888, which much more generally opened transmission access to non-utilities. These regulatory changes together started to transform the electricity transmission system into a common carrier system.10 With EPACT and Order 888, it became much more difficult to control electricity generation markets by restricting access to transmission lines. Utilities still made the investment decisions

9

Delivered electricity from many renewables, particularly wind and solar, are significantly smaller than suggested by the nameplate capacity. 10 These issues are discussed in detail in Chapter 14.

326


for transmission facilities and thus could still exercise some control over generation markets, but this form of control was less effective than direct control over access to transmission lines. These developments, happening while California was designing its electricity market changes, were important for establishing the opportunity for wholesale electricity competition in California. 10.1.3. The context for California electricity restructuring 10.1.3.1. Generation/wholesale markets The high retail price of electricity in California was one basis for arguments that California’s electric system should be deregulated to create a more competitive and presumably lower cost, electricity system. The major contributors to the high retail price were high average cost of generating electricity and high prices embedded in contracts for purchasing electricity under PURPA contracts. Some advocates of electricity generation deregulation expected deregulation to reduce retail prices of electricity quickly. But as described in this chapter, this expectation was based on a fallacy: it implicitly assumed that deregulation could correct the historical problems that had led to the high generation costs and high costs of purchasing bulk power under contracts. But the costly investments in nuclear power plants and the long-term contracts for QFs could not be reversed; sunk costs cannot be changed by restructuring. Regulators can reallocate who will bear those costs, but sunk costs will continue to be borne by some combination of companies and individuals. And, as became apparent over time in California, laws and legal precedents, basic notions of fairness, and political forces can all sharply limit the ability of regulators to reallocate many of the historical sunk costs. Changing regulations, however, can halt practices that had led to high costs. But in California, many of the practices that had increased generation costs had already been halted. At the time of the restructuring debate, the state was no longer investing in new nuclear power plants. New cogeneration plants and renewable energy investments were being made on a market basis when such investments were expected to be economically attractive. The high price standard offers under PURPA were no longer required for new contracts. There was a more subtle argument, however, that deregulation would change some practices that had elevated costs and that could be expected, absent deregulation, to continue elevating costs. Halting such practices would reduce costs, although the cost reductions would be gradual. There was a widespread belief that the regulatory system did not provide strong enough incentives for utility owned electricity generators to minimize costs, so it was doubtful that the regulated system led to the lowest cost mix of energy generation technologies. Utilities seemed to be favoring their own generation over generation by independent power producers (IPPs) and thus not cost minimizing. There remained incentives and opportunities for utilities to block distributed generation and to rely instead on centralstation power, even if distributed generation had lower overall costs. Economists and other industry analysts were arguing that creating competition could change economic incentives facing the utilities. The change in incentives would gradually reduce costs of electricity generation, which would gradually reduce retail prices. This argument, although not proven,11 was probably valid, even though the hope of rapid cost savings was never realistic. 11

Those arguments were discussed in detail in the Yellow Book and the Blue Book. These documents were developed by the CPUC in the process of the restructuring.


327

In addition, it was expected that expansion of wholesale markets would encourage more California investment in new generating capacity by IPPs. Most analysts anticipated that the healthy California economy would continue to need more electricity. Many doubted whether the old regulated system would be responsive enough to those economic needs. And many argued that the old regulated system would lead utilities to discourage new investment by IPPs. Thus, there was the opportunity in California to deregulate electricity generation to expand the scope of the existing competitive portion of the electricity generation industry. Expanding competition in electricity generation was expected to create incentives for cost cutting, to encourage investments in new generating capacity by IPPs, and to provide a flexible system for a dynamic California economy. 10.1.3.2. Stranded costs A problem of wholesale market deregulation was the issue of stranded costs. If future costs would be low for new generation, then future wholesale prices would be low, as well. This implied that the existing high cost-generating units would no longer be economically viable in a competitive environment. Costs incurred by the utilities in constructing these plants would be “stranded”; utilities would incur losses, without policy intervention. The issue of stranded costs was not a problem of “going forward” costs – future total costs of electricity generation ignoring sunk costs – even if those costs might be very high for some of the units. If future wholesale electricity prices turned out to be lower than the operating costs of these plants, the plants would shut down in a competitive environment and their entire remaining book value would be a loss to its owner. However, such plants should be shut down for economic efficiency reasons. On the other hand, if the future wholesale electricity prices turned out to be higher than the operating costs of these plants, these plants could sell electricity at the wholesale price and would find generation to be more profitable than shutting down. These plants could compete in a market environment, as would be desirable for economic efficiency. However, there would still be a capital loss: the owner would not be able to recover all of the remaining book value. These fixed losses were sunk costs, and therefore not expected to influence the marketclearing price (MCP). But someone would have to bear the losses. Who should bear these losses – the utilities or their customers – was a politically important issue, a legal issue, and an issue of fairness. There were several possible regulatory alternatives consistent with deregulation of generation. One alternative would be to allow the utility to include those costs in retail prices, thereby keeping retail prices high. That solution would provide motivation for customers to bypass utilities with large stranded costs, by purchasing directly from generators or generating electricity themselves, say in cogeneration units. Customers most able to do so would be the large users of electricity. If enough large customers bypassed utilities, utilities would primarily sell to residential and small commercial consumers; small users would thus pay most of the stranded costs. Retail competition would be very difficult to design under that alternative. Another option would be to require the utility to write off the assets as losses, requiring the stockholders to bear the consequences of stranded costs. Retail competition could more easily develop. But utilities argued persuasively that it would have been inequitable for their investors to bear the stranded costs of long-term contracts and generating investments that had already been approved by the CPUC as prudent under the old “regulatory compact”. It had generally been understood that under the “regulatory compact” utilities would make investments to serve the needs of ratepayers and ratepayers would pay back the costs of

328


those investments plus a fair rate of return on the investments, over the life of the equipment. The utilities argued that eliminating the obligation of the ratepayers to pay back the costs, after the utilities had made the investment, would be unfair. And because some additional costs, such as some of the QF contracts, had been forced on them by the CPUC implementation of PURPA, forcing the utilities to bear the costs would have been particularly unfair and may well have been illegal. The debate over who should be liable for the stranded costs became central to the regulatory hearings and legislation. And the size of stranded costs was difficult to predict, since they depended on the future sales prices of wholesale electricity and future natural gas prices, factors that were highly uncertain. This debate was not limited to California. While California debated its restructuring, how to deal with stranded costs was a national debate. The debate was particularly intense because the stakes were large. Some analysts estimated the nationwide stranded costs as exceeding $100 Billion. 10.1.3.3. Retail markets There was also a possibility of creating competitive markets for retail electricity, since many recognized that the commodity, electricity, could be unbundled from its distribution services. Retail competition offered the possibility that competing retailers would provide differentiated energy services that would be attractive to consumers. Some retailers could provide “green power” to environmentally conscious consumers. Others could bundle energy efficiency measures with electricity to help consumers reduce the overall cost of obtaining energy services. Some retailers could provide higher reliability of electricity for customers for whom reliability was essential or interruptible service for those customers willing to accept service interruptions in exchange for a lower overall bill. Some could sell electricity at real-time prices for those customers that wished lowest average cost but did not mind price variability and others could sell electricity at guaranteed prices, essentially selling risk management services bundled with electricity. A competitive retail market, in principle, could enhance consumer options and create a more flexible system. 10.1.4. The restructuring process The CPUC took leadership to reduce economic regulation in electricity markets, spearheading the move toward electricity deregulation in California. CPUC members had been aware of the UK restructuring; the UK experience suggested the possibility of introducing more competition to California’s electricity markets. The deregulation that they envisioned would rely on competitive market forces in both wholesale and retail electricity markets. The process was initiated in 1992, culminated with a final CPUC restructuring order in December 1995, and legislation – Assembly Bill 1890 (AB 1890) – signed in September 1996. Table 10.1 provides a timeline of the process and the subsequent events. Each element of the table is discussed at a later point in this chapter. 10.1.4.1. CPUC regulatory proceedings In April 1992, the CPUC initiated a review of trends in the electric industry, which initially resulted in a staff report12 published in February 1993. This report, commonly referred to as 12

“California’s electric services industry: perspectives on the past, strategies for the future,” California Public Utilities Commission, February 3, 1993.


329

Table 10.1. Timeline of California’s restructuring, electricity crisis, and subsequent events. PURPA passed EPACT opened access to transmission networks CPUC initiated review of trends in electricity markets “Yellow Book” published “Blue Book” published CPUC Decision D. 95-05-045 Memo of understanding presented Final CPUC Restructuring Order Request filed with FERC to create PX and ISO AB 1890 passed and signed into Law FERC Order 888 FERC authorized ISO and PX First generating plant divestiture Proposition 9 defeated Last generating plant divestiture Wholesale prices spike above $400/MWh Wholesale prices spike above $2000/MWh California DWR begins buying electricity FERC Price Mitigation Order Wholesale prices fall sharply FERC Final Price Mitigation Order Scheduled end of transition period Governor Davis recalled in special election Arnold Schwarzenegger become Governor Nunyez Bill passed legislature; vetoed Proposition 80 qualified as an initiative

1978 1992 1992 February 1993 April 1994 May 1995 September 1995 December 1995 April 1996 September 1996 1996 1997 June 1998 November 1998 April 1999 June 2000 December 2000 January 2001 April 2001 May–June 2001 June 19, 2001 March 2002 November 2003 January 2004 2004 2005

the “Yellow Book,” laid out a well reasoned and valid intellectual framework for the subsequent debate on deregulation of California’s utilities. It clearly articulated the fundamental reasons why California should reform its regulatory structure: “First, California’s current regulatory framework, significant portions of which were developed under circumstances which no longer persist, is ill suited to govern today’s electric services industry.” “Second, the state’s current regulatory approach is incompatible with the industry structure likely to emerge in the ensuing decades.” It expanded on these conclusions, describing problems of the then current regulatory system: “The regulatory program blunts incentives for efficient utility operations.” “The current regulatory program increases the potential for inefficient investment due to unbalanced incentives governing utility investment options.” “The current regulatory approach requires many complex proceedings, which increase administrative costs and threaten the quality of public participation and Commission decisions.” “The current regulatory approach offers utility management limited incentives and flexibility to respond to competitive pressures.” “The current regulatory approach conflicts with the Commission’s policy of encouraging competition in the electric services industry.”

330


The Yellow Book laid out four broad options, ranging from limited structural change through a restructuring the utility industry, including divestiture of utility generation assets and creation of a competitive retail electricity industry. One option, recommended by the CPUC staff, included creation of a “core/non-core” retail market in which a limited segment of the largest consumers could enter direct-access contracts with IPPs, bypassing the utilities. As will be discussed at a later point in this chapter, a more aggressive approach was actually incorporated into the California restructuring. However, at the time of this writing, 12 years after completion of the “Yellow Book”, Governor Schwarzenegger and the CPUC are promoting just such a “core/non-core” retail market. Building on the conclusions in the Yellow Book, in April 1994, the CPUC issued an Order proposing a process of restructuring California’s electricity industry. Often referred to as the “Blue Book”, this order envisioned competitive retail and wholesale electricity markets. This and all subsequent restructuring Orders maintained the utilities as monopoly providers of delivery services. The Blue Book laid the foundation for California’s subsequent electricity restructuring, proposing several fundamental changes, including replacing cost-of-service regulation with performance-based regulation, wherever regulation was needed. The Blue Book proposed to grant all purchasers of electricity voluntary and direct access to electricity suppliers, in a time-phased manner. It envisioned that both the regulated utility and unregulated retail purchases would coexist. Electricity generation would be deregulated. Wholesale prices would be disciplined by competitive wholesale markets. The Blue Book addressed the issue of stranded costs by proposing a financial transfer from the utility customers to the generation side of the utilities, in the form of a limitedtime-duration non-bypassable “competition transition charge” (CTC), imposed on each retail customer. The utility would receive all money collected through the CTC, allowing it to recover stranded costs. The magnitude of stranded costs itself depended on the market-determined wholesale price of electricity. The higher the wholesale price, the smaller would be the stranded costs, and the smaller would be the CTC. Thus the CTC would be determined only over time, as cost and price data was revealed in the marketplace. Although the “Blue Book” laid the foundation for the restructuring, many steps were required to complete the process. Each step seemed to add more complexity to the restructured system. And many of the restructuring design elements changed greatly as the process continued. In May 1995, the CPUC issued a Decision13 that laid out two broad policy alternatives for organization of restructured wholesale markets and transmission management: a preferred (majority) and alternative proposed policy. These alternatives were strongly influenced by observation of the UK electricity market structure. The preferred structure was a wholesale power pool, managed by a newly created California Independent System Operator (CAISO). The CAISO would dispatch generation based on a day-ahead auction and would arrange transmission access for generators that successfully sold in that auction. Under this proposal, management of the grid, dispatch of generators, and wholesale trading would be the integrated functions. Wholesale prices for electricity could vary sharply with supply and demand conditions, with risk for both generators and consumers. Risk management would be available only through financial instruments to hedge prices. The CPUC was to take no responsibility for

13

CPUC Decision D.95-05-045.


331

establishing such hedge instruments, but the Decision envisioned private firms creating such markets. Physical, bilateral contracts would be allowed after 2 years. The alternative policy envisioned consumer choice through physical, bilateral contracts, separate from any pool bidding, to be available immediately. In September 1995, four major participants, including a utility, a group of generators, and two electricity user groups, presented a Memorandum of Understanding (MOU) with their joint recommendations. Although it addressed virtually all elements of the proposed restructuring, the MOU focused on market structure and stranded cost issues. The proposed market structure combined features of the preferred and of the alternative proposals from May. It proposed a system more complex and less coordinated than either of those envisioned in the May Order. The MOU proposed creation of a Power Exchange (PX), creation of an ISO, and early phase-in of direct bilateral contracts between generators and individual customers or distribution companies. The ISO and PX would be separate entities, operating independently of one another. The PX would develop a visible electricity spot market with transparent electricity prices, open to all suppliers, both within and outside of California. The ISO would manage the grid. Normally these functions, integral parts of a smoothly functioning system, would be tightly integrated into one organization. This structure, ultimately adopted by California, almost assured that the functions would not be well coordinated. It would provide opportunities for energy traders to operate profitably; market inefficiencies could create profit opportunities through arbitrage and through selling financial instruments for managing the increased risks.14 A continued set of hearings and public submissions led to a final CPUC restructuring Order,15 issued in December 1995, often referred to as “The Preferred Policy Decision”. The CPUC Order followed the MOU recommendation to separate the ISO (to manage the grid) and PX (to create wholesale markets.) The organizational separation of the two closely connected functions promised to create an extremely complex and untested system. As shall be explained later in this chapter, the resulting inefficiencies and unusual incentives associated with this structure would help make the subsequent energy crisis worse than it would have otherwise been. This structure, combined with the non-storable nature of electricity, virtually assured that firms would take advantage of the market complexities. Like the Blue Book, the restructuring Order proposed to deal with stranded costs through a CTC. The CTC would be designed to allow utilities to recoup virtually all stranded costs16 by the year 2005. This CTC will be discussed more fully at a later point. At the retail level, the CPUC envisioned a system in which consumers would have many options for electricity purchases. Consumers could continue complete reliance on a local distribution company to purchase and deliver electricity or could opt for direct access through bilateral contracts to deliver electricity. Those relying on a local distribution company could agree to either pay the average cost of electricity throughout the year or pay a real-time price, a price that varied on an hour-by-hour basis with changing wholesale market conditions. 14

This seems to be a remarkable public policy concept: create market inefficiencies to make the system profitable for arbitrageurs, traders, and financial markets. 15 Order instituting rulemaking on the commission’s proposed policies governing restructuring California’s electric services industry and reforming regulation. Decision D.95-12-063 (December 20, 1995) as modified by D.96-01-009 (January 10, 1996). 16 In addition, long-term contractual obligations entered into before January 1, 1996 would be recovered over a longer time period.

332


Those paying a real-time price could choose hedging contracts with third parties to reduce the risk. The CPUC restructuring Order included one provision that could have been interpreted as creating a price cap on retail electricity prices. The provision would have limited the size of the CTC, assuring that if the CTC otherwise would have increased electricity prices above the January 1996 levels, then the CTC would be reduced or eliminated to limit retail price increases.17 Electricity distribution functions would remain with the utilities and would be regulated by the CPUC. The regulated distribution costs would include a separate non-bypassable component of retail rates that would have provided funds for other social goals: a Public Interest Energy Research Program and demand-side management programs to promote energy efficiency. This restructuring Order left open many implementation details. Implementation rules were developed through a sequence of CPUC decisions.18 10.1.4.2. State legislation: AB 1890 Although the CPUC had issued the restructuring Order, such a fundamental reform would be politically more viable if it were the product of legislation, not simply regulatory rulemaking. Soon after the CPUC restructuring Order, the state legislature embraced this role. The legislative process culminated in AB 1890, passed by the California Legislature and signed into law in September 1996. It was originally to take effect on January 1 1998, but was delayed to March 1998 due to implementation delays. The bipartisan nature of the restructuring legislation was striking. Primary leadership for the entire legislative package came from a Democratic member of the State Senate and a Republican member of the State Assembly. The bill passed with no dissenting votes from legislators of either party, a rarity in contentious California politics. A Republican Governor, Pete Wilson, signed the bill. Moreover, the bipartisan legislation built upon an open public process led by the CPUC. The legislative process started from the CPUC restructuring Order of 1995, but modified several central provisions and added its own features. AB 1890, like the CPUC restructuring Order, allowed a utility to include generation, transmission, and local distribution, but would not allow coordinated decision-making among these functions. Decision-making and control of the transmission function would reside with the ISO, not the utility owning the transmission lines. The market structure separated local distribution decisions from fossil-fuel-fired19 electricity generation decisions: a utility that both generated electricity from fossil-fuel-fired plants and sold electricity at retail would face a strong economic incentive to operate as if two separate companies owned these two functions. This separation was accomplished by requiring the utility to sell all electricity it generated from fossil-fuel-fired plants through the PX or the ISO and to acquire all electricity for retail sale through these markets. It could not keep its own-generated electricity for its retail sales. Market-clearing conditions operated independently of the identity of buyers or sellers. Thus, the utility would be unable to treat 17

This was not a rules precluding retail prices from rising above the January 1996 levels, but rather precluding the CTC from pushing retail prices above those levels. 18 Particularly important was the 1996, Roadmap decision. CPUC Decision D.96-03-022. 19 Hydroelectric generation could still be coordinated with retail sales and would thus provide the utility some opportunity of changing production with changes in load.


333

itself preferentially either as a buyer or seller of electricity. In that way, market designers hoped to eliminate, or at least sharply reduce, the market power of the utilities operating as producers of wholesale electricity and simultaneously as dominant users of wholesale electricity, typically transferring the electricity it generated to itself. In addition, the utility was not allowed to recover costs of purchasing electricity through new power purchase agreements other than through the PX or CAISO. Unfortunately, this critical provision created a fundamental problem of risk management. Initially the PX-operated markets that cleared only 1 day ahead of the electricity use. The ISO included only day-of markets. Thus this rule precluded acquisition through forward contracts, dangerously exposing the IOUs to all short-term volatilities in wholesale prices. In addition, as discussed later, the reliance on spot markets, rather than long-term contract markets, increased opportunities for exercise of market power by generators. Only in 1999 did the PX create any markets that operated more than a day before the time of electricity use. And these markets never became a significant venue for power acquisition. AB 1890 partially separated distribution services from retail sales of electricity and confirmed that distribution services would continue to be subject to CPUC regulatory authority. The incumbent utility would supply both distribution services and electricity, while competitors would supply only electricity. The Act promised to create competition for retail electricity sales by authorizing direct transactions between electricity suppliers and end use customers and by allowing electricity aggregators, while the IOUs would be the default seller of electricity. One option would have been to require a complete separation of the two functions, by allowing companies to either provide distribution services or sell retail electricity. Such a system, which has been adopted in Texas, is discussed more fully in Chapter 11. Under such a system, the incumbent utility might start with a competitive advantage if it chose to sell retail electricity rather than distribution services. But since the utility would have no lasting structural advantage, over time any competitive advantage could be expected to dissipate. But such a system was never seriously considered in the restructuring, perhaps because the utilities themselves played such a large role in the system design. Direct access was to start simultaneously with the initiation of the PX and ISO. Aggregation would be allowed by private sector marketers, as long as individual customers could freely choose remaining with the local utility or purchasing electricity from the aggregator. Aggregation by cities or other public agencies would be generally allowed. AB 1890 transformed the cap on the CTC into a retail price cap, a critical change. The law required that bundled electricity plus distribution services prices for residential and small commercial customers be reduced immediately by at least 10% below their June 10, 1996 levels. The transition period, during which the stranded costs would be recovered, was made much shorter than proposed under the CPUC restructuring Order. This period would end no later than March 31, 2002, or whenever the stranded costs had been fully recovered, whichever came sooner. The IOUs were not guaranteed that they would recover their entire stranded costs; they would be at risk for costs not recovered by March 31, 2002. AB 1890, by shortening the transition period and reducing the retail electricity rates, increased the difficulty of recovering the stranded costs. To recover its stranded costs, each utility would propose to the CPUC a cost recovery plan that included the capped retail prices. It was anticipated that wholesale costs would decline. Under this assumption, the differential between capped retail prices and costs20

20

The CTC would be equal to this difference. The CTC was a non-bypassable charge that would be imposed on all retail sellers in addition to the utilities.

334


would increase, leading to financial accumulations each year and stranded costs would be recovered over a limited transition period. But one major risk seemed to be ignored: wholesale prices could not be predicted well. AB 1890 included no statement of what might happen if the wholesale price exceeded the capped retail price. Perhaps if that issue had been clearly resolved, various parties would have more fully understood the financial risks and would have been able to take appropriate steps to manage risk once the law had been passed. And perhaps the retail price control would have been conditioned on wholesale prices remaining below some historical level, with provisions for abandoning price control if wholesale prices rose above that level. A safeguard, at least in theory, was that AB 1890 imposed no restrictions to stop the CPUC from modifying or abandoning the stranded cost recovery plan, if so requested by the utility. The CPUC could agree to reduce the amount of stranded costs to be recovered to just the amount already recovered, thereby terminating the requirement that retail rates for that utility remain at their price-capped level. AB 1890, like the CPUC restructuring Order, kept organizationally separate the PX and the ISO. It directed the CPUC to work with the utilities to develop a PX, but it included no further guidance about the competitive auction, the bidding structure, or length of the advance period during electricity could be purchased. AB 1890 described the technical functions of the ISO but gave no guidance to its market functions. Implementation issues were left to the CPUC, the ISO Board, and to FERC. The stage was set for the subsequent state–federal jurisdictional disputes. AB 1890 directed CPUC to work with the utilities to obtain authorization of FERC for the creation of ISO and the PX. Even before its passage, in April 1996, the three IOUs submitted requests to FERC21 asking approval of those restructuring elements subject to FERC jurisdiction. FERC largely approved these proposals and in 1997, FERC authorized the first limited operation of ISO and the PX. The set of regulatory changes, culminating in AB 1890, promised to fundamentally change the electricity system from one strictly regulated from “cradle to grave”, into one in which market forces would play the primary role, at least once each utility passed its transition period. Wholesale markets were intended to allow competition to determine supply, demand, and prices of electricity in wholesale transactions. Although analysts envisioned that most small retail customers would continue obtaining their electricity bundled with distribution services, sold by regulated utilities, the seeds for a competitive retail market were planted. Given the anticipation that wholesale prices would fall, the transition period was expected to be just that, a smooth transition to a competitive market, a transition in which the problem of stranded costs would be solved. Unfortunately, when massive system changes are implemented, reality can diverge greatly from expectations. 10.1.4.3. After passage of AB 1890 But the process was not yet completed with passage of AB 1890. In the next year a group of consumer advocates gathered signatures to qualify what became Proposition 9, voted in the 1998 California ballot. Proposition 9 would have turned back the restructuring and would have reinforced a system of tightly regulated electricity transactions. Proposition 9 created

21

FERC Dockets Nos. ER96-1663-000 and EC96-19-000. The applications were filed after the CPUC restructuring order but before passage of AB 1890. These applications were approved only after AB 1890 was signed into law. FERC took due note of the passage of AB 1890 during its proceedings.


335

much uncertainty and reduced the incentives for private companies to invest in new electric generating facilities. By creating significant delays in the capital investment associated with investing in new generating plants, Proposition 9 increased the risks that were inherent in this newly restructured system. It was not until the defeat of Proposition 9 in November 1998 that assured AB 1890 would be allowed to operate. Like the restructuring order, AB 1890 was simply a framework, not detailed designs for system implementation. The structure of the PX and ISO, as well as of the markets they were to operate, was left to stakeholder committees and the CUPC in California to design and left to FERC to approve or disapprove the designs. Careful delineation of the jurisdictional split between the Federal regulators and the state regulators was left to the various parties to work out. Working out jurisdictional conflicts within the short time frame was close to hopeless. That jurisdictional split has never been fully resolved. AB 1890 came into effect (still under the cloud of Proposition 9) only 18 months after it became law. With such a tight implementation schedule, the original applications to FERC were filed while the California legislature was still considering AB 1890. Once FERC approved the applications, the large size and diversity of the part-time stakeholder boards made it difficult, if not impossible, to seriously rethink or revise the original structure. The following examines the various system components, in more depth. 10.1.5. Wholesale markets under the restructured system 10.1.5.1. The California power exchange (PX) The IOUs and the CPUC developed, and FERC approved, plans for the PX and for the wholesale markets that the PX would manage. The PX organized a set of competitive auctions, open on a non-discriminatory basis to all suppliers.22 The PX initially established oneday-ahead and day-of wholesale markets for electricity. Only much later did it establish markets that allowed contractual agreements extending longer than one day in advance. For both the one-day-ahead and day-of wholesale markets, the PX accepted bids to sell electricity hour-by-hour and bids to purchase electricity hour-by-hour. Hourly prices were determined on a market-clearing basis with all buyers paying the same MCP and all sellers receiving the same MCP. All bids to sell with offer prices lower than or equal to the MCP and all bids to purchase with offer prices greater than or equal to the MCP would be accepted; all others would be rejected. This one-price auction system was designed to simulate a perfectly competitive commodity market. There were apparent alternatives to such a one-price auction system. One alternative, in principle, would have been to set up a single commodities futures market. Agents would enter bids to buy and to sell, prices would adjust, and MCP would be reached. Prices for a given delivery time would be free to adjust as the delivery time grew near. But adjustment process would take time; electricity markets would have to adjust on a much faster time scale than normal commodity markets. All adjustments would have to be completed, starting at most one day before the day of delivery. And there were 24 separate hourly markets for each day. This can be compared with the future market for natural gas, which sets a price for delivery for an entire month, not 720 different prices (24 ⫻ 30) for each month. Although development of such a structure would have fully integrated the various markets and

22

The three large IOUs were required to sell all their remaining generation through the PX. For all other entities, use of the PX was optional.

336


would have reduced the opportunities for market gaming, it appeared (and still appears) too complex to be successfully implemented. A second alternative would be to design the system to pay bidders what they bid, rather than to pay them the MCP. The total cost of all purchases would be averaged, and the buyers would each pay the average bid price. Some have argued that a system of paying on an as-bid basis rather than on a marketclearing basis would result in smaller total payments by the buyers of electricity. After all, those bidding to sell at prices below the cut-off price would not receive the cut-off price but would receive only their bid prices. The fallacy of that reasoning is that it assumes that the sellers of electricity would offer the same bids under an as-bid system as they would under a market-clearing system. But under an as-bid system, each firm makes the most profit by guessing the cut-off price and bidding at or just below that price, as long as the cut-off price is at least as high as its marginal cost. Thus, even in a competitive market, suppliers would not bid at their marginal costs under an as-bid system. There was an additional difficulty with an hourly auction system. Many generating plants, typically operating as base-load plants, have long and costly periods for ramping up from no production to full capacity. These plants might be profitable to operate if they received at least a particular price, say $30/MWh, for a large fraction of the day or for all of the peak period during a day. However, if they were to operate for only a few hours, even at a higher price, say $40/MWh, they might not be profitable to operate, since the fixed costs of ramping up could be greater than the profit earned during those limited hours. For such plants, their offer price at any hour must depend on whether they would be generating electricity at the other hours during the day. For such plants, bidding based on unit commitments, commitments of the unit to operate for long blocks of time, would be more appropriate. But that solution was not chosen. Other plants, for example, nuclear, had no option to cycle or reduce production, but rather had to be run as base-load. But for these plants the problem could be easily solved by bidding in their production at zero price, thereby assuring that they would sell their output at the MCP. Once trades had been arranged through the PX, the PX served as the scheduling coordinator for scheduling generation plans and loads to the CAISO. 10.1.5.2. The California independent system operator The CAISO was given the responsibility for managing the transmission grid and for assuring that resources were available to assure safe operations of the grid, including assuring that there would be sufficient quantities of electricity available at all times. ISO real-time markets. To perform these functions, the CAISO collected the schedules of electricity to be generated and of the electric loads to be served by the IOUs, by the Munis, and by a host of other entities,23 on an hour-by-hour basis. Forty-some “scheduling coordinators” or SCs reported their schedules to CAISO. CAISO integrated these schedules and assured that these schedules collectively did not overload any parts of the transmission grid or were in any other ways not feasible. If they were not feasible, then the CAISO could order adjustments to dispatch generation feasibly.

23

AB 1890 applied only to the IOUs and not the Munis. Under the new system, CAISO does not control all load, only the load of the IOUs and others who wish to join. Due to high costs of the CAISO, manyof the Munis have declined to use the services of the CAISO.


337

Each scheduling coordinator was required to submit a balanced schedule in which the total loads and resources were equal to one another.24 That is, the total projected use of electricity at each hour and the total generating resources to provide that electricity were required to be equal to one another for each submission by every SC. Thus, in theory, the sum total of loads and resources would be balanced for the system, absent CAISO-ordered adjustments. However the participants in the market could not perfectly project the electricity needs and the loads and resources could become unbalanced. To correct such imbalances, the CAISO would run a real-time energy imbalance market, buying and selling electricity after the PX day-of market had closed. Deviations between predicted and actual supplies and demand would be corrected on this imbalance market. System designers expected these differences to be small. The CAISO tariffs did not include penalties for imbalances, even for extreme imbalances. As discussed later, there was an incentive to mis-schedule. Here again, reality diverged sharply from expectations: imbalances turned out to be large fractions of scheduled power. The CAISO was responsible for purchasing and selling real-time electricity on behalf of the SCs. Since it was engaging in these balancing transactions, its decision rules were central to the real-time market for electricity. One decision rule was very different from that in the PX. The CAISO would acquire sufficient electricity to meet the loads it predicted, not simply the loads that were bid into the market. These predicted loads did not depend at all on price. CAISO would not reduce electricity acquisitions, even if the acquisition prices became very high. In contrast, bids to purchase electricity on the PX could be price dependant, in fact could include far more price elasticity than would characterize the actual loads. Possible demand-side adjustments were excluded from the market price determination. The mentality of reliability at any cost at a later time led to exceeding high costs for all electricity users in California, costs often higher than the reliability was worth to them. A second decision rule was that CAISO would reject bids above some wholesale price cap level. This wholesale price cap (discussed later ) was during different times $250/MWh, $500/MWh, or $750/MWh. Together, these two rules implied that the demand function in the CAISO real-time market was completely or almost completely independent of prices up to the cap but perfectly elastic at the cap level. These rules together increased the volatility of prices on the real-time market, up to the capped price, but kept MCPs from exceeding the cap. They also implied that any demand response that might appear in the PX bids would disappear for the final CAISO real-time power acquisition. CAISO ancillary services markets. The CAISO paid sellers of ancillary services to make their generating units available, should they be needed. If these units were needed, the sellers were paid for the electricity generated.25 Every load-serving entity or LSE, typically an electric utility, was responsible for its proportional share of ancillary services. Each scheduling coordinator could choose whether to

24

More precisely, the sum of all loads and the sum of all generation submitted by a scheduling coordinator must be within 2 MW of each other. 25 Initially this system was set up so that these units would be paid for the electricity plus paid for the ancillary services. However, FERC later required ISO to change the rules so that generators would be paid for either generating electricity or providing ancillary services but not both.

338


provide its share of ancillary services or to have these services purchased on its behalf by CAISO. For those ancillary services not self-supplied by the scheduling coordinator, CAISO managed a single-price bidding system that operated in day-ahead and hour-ahead periods. The CAISO could obtain additional ancillary services through supplemental bids offered during the hour the reserves were needed. 10.1.5.3. Divestiture of IOU generating assets AB 1890 and subsequent CPUC rulings sharply reduced the degree of vertical integration in the industry. However, there was still a concern that common ownership of generation and retail functions would give the utilities wholesale market power,26 would make it difficult to operate a competitive wholesale market, and would lead to high wholesale prices. To address that concern, several options were considered. One option was to allow utilities to continue acquiring electricity directly from their own generators, as well as buying it from non-utility generators, either through organized markets or through bilateral contracts, keeping some vertical integration. A second option was to require the utilities to divest themselves of their generating assets, at least their thermal assets. The CPUC ultimately implemented a two-fold solution. First, the CPUC required the utilities to divest themselves of 50% of their generation assets and provided financial incentives to divest the remainder. Second, all remaining thermal generation owned by the IOU could be sold only through the PX or the CAISO.27 Together, those rules would assure that the PX and the CAISO markets would include large volumes of transactions and that IOUs would be precluded from any meaningful self-dealing between their wholesale and retail operations. The incentives for divestiture were successful. As of 2000, only 29% of the electricity sold in the state was generated by the IOUs versus 44% generated through plants that had been divested and owned by non-utility generators. 10.1.5.4. Forward wholesale contracts for electricity (and lack thereof) Divestiture of plants, however successful, created new problems. As the IOUs divested their generation assets, the CPUC feared that if there were linked agreement both to sell the generator and to purchase electricity under a long-term contract from that generator, there would be financial incentives to distort the selling price and the long-term sales price to increase profits. Guarding against this potential would require more regulatory oversight. In addition, there was a fear that long-term contracts could simply substitute for a utility ownership of the generators, hence preventing the emergence of a competitive market. Moreover, potential new entrants into the wholesale market might be discouraged just as much as they would be absent divestiture of the assets. In order to assure that the PX markets would not be too thin, there was a desire to limit the long-term contracts at the wholesale level. As indicted above, the CPUC required utilities to acquire all electricity, not already under long-term contract, through the PX or CAISO. Since the PX and the CAISO

26

In general, vertical integration itself does not create market power. However, as long as the incumbent utility remains dominant in the retail market, if the utility’s acquisition of bulk power is not perfectly regulated, vertical integration allows it to extend its market power in the retail market into the otherwise competitive wholesale market. 27 More precisely, no costs could be recovered for this generation unless the electricity were sold through the PX or the ISO.


339

markets operated no more than 1-day ahead of the electricity delivery, this requirement effectively prohibited the utilities from entering any forward contracts. Perhaps intentionally or unintentionally, these rules together made the California IOUs overly dependant on a spot market, a market that turned out to be very volatile and subject to market gaming. The IOUs, realizing their potential vulnerability, unsuccessfully tried as early as 1999 to gain the right to procure electricity on a longer-term basis. In March 1999, SCE filed an application for a pilot program under which it could enter traditional power purchase agreements for electricity and for capacity. The CPUC denied the application. In mid-1999 the PX applied to organize a block forward market, FERC approved the application, and the CPUC approved the request by SCE and PG&E to participate in that market. But the block forward market allowed contracts for no more than 1 year. Until August 2000 the utilities had no right to enter bilateral contracts. This, along with the retail price controls, during the energy crisis proved to be fatal flaws in the system. Forward contracts could have substantially reduced the risk of large changes – up or down – in the acquisition cost of electricity. Utilities could have guarded against or limited the high risk of large fluctuations in the wholesale electricity price. But, that road was not taken by the CPUC. 10.1.5.5. Wholesale markets: in summary AB1890 and CPUC rules created a complicated set of wholesale markets poorly coordinated with one another. These markets were given monopoly or near monopoly status and thus utilities could not escape problems associated directly with these markets. Risk management options were taken from the IOUs through divestiture and through reliance on wholesale spot markets. Therefore, volatility in the wholesale markets was nearly assured. The interplay of these various markets, the resulting bidding strategies of utility-buyers, private generators, marketers, and municipal utilities, and the responses of the CAISO and the PX personnel were all untested at the time of the restructuring. In addition, there remained opportunities for exercise of market power by even those generators with small market share. 10.1.6. Retail markets Creating competitive retail markets was seen to be even more of a challenge, even though there had been extensive experience in other nations. At least two factors stood in the way of creating a competitive retail market for electricity: retail market power of the incumbent utilities and risk management. Local distribution companies had a natural monopoly for the delivery services, the wires, transformers, and control systems.28 In addition, in the short term they could be expected to have a significant degree of market power for the supply of electricity itself, since electricity had always been bundled with the delivery services. Unless retail sales of electricity were fully unbundled from monopoly delivery services, the issue of retail electricity market power would remain. The new system partially decoupled delivery services from retail sales of electricity. Delivery services would still be provided by utilities, as monopoly franchises, earning a regulated fee. Unbundled electricity could be sold by aggregators or generators and delivered by

28

The CPUC had jurisdiction over only IOUs, and thus they were treated differently from the Munis. This discussion is based on the changes for the IOUs.

340


the local utility. Under the new system, in principle, any customers could enter bilateral contracts with electricity suppliers and therefore to bypass the incumbent IOUs for electricity, even though the CTC could not be bypassed. However, the decoupling was only partial. The local utility could sell electricity bundled with distribution services. IOUs would operate as regulated retail sellers of electricity, subject to review and control by the CPUC. Any customer who wished to pay only one monthly bill or wished to have only one company with which to deal, had no option but to remain with the incumbent utility. This gave the utility a great competitive advantage as seller of electricity. As discussed above, during the transition period, AB 1890 imposed price caps for retail electricity sales by the incumbent utilities during a transition period. This price cap created a dilemma. On the one hand, the IOUs were required by AB 1890 to reduce electricity prices for residential and small commercial customers by 10%, as discussed above. On the other hand, the CTC magnitude was to be chosen so that all stranded costs could be recovered over a small number of years, although recovery was not guaranteed. Thus with the CTC, the retail price of electricity alone would be approximately equal to the recent historical electricity component of the bundled retail price, not below that level. The dilemma was resolved through a financial instrument. The utilities were authorized during the transition period to issue “rate reduction bonds” to finance the difference between their cost for electricity (wholesale price plus CTC) and the price-capped retail price as well as to refinance some of their existing capital equipment. These bonds would be repaid once all stranded costs had been recovered and the CTC was no longer in operation. This plan implied that a significant share of the price reduction the customers thought they were enjoying would be repaid in later years. The IOUs would be default sellers of electricity, available for everyone who wished to purchase retail electricity from these utilities. Their price-capped rates would be available for all customers, even those that switched to other retail suppliers but subsequently chose to return. This implied an asymmetrical relationship between the utility and the new competitors: the new competitors could choose whether to take new customers, but the utility had no choice whether to take new or returning customers. New entrants could create distinctions among electricity delivered from different sources. For example, they could sell “green” electricity, advertised to be generated entirely or primarily by renewable sources. But this component of the market would necessarily be small, because most of the renewable electricity was being sold under contract to the large electric utilities. For residential customers, entrants could bundle energy efficiency measures with electricity to help consumers reduce the overall cost of obtaining energy services (e.g., warmth, lighting, cooking, clothes drying, refrigeration). But the IOUs themselves were offering these services using public benefit charges included in non-bypassable surcharges within the delivery fees. This limited the ability of new entrants to attract customers through provision of energy efficiency services. New entrants would find it very difficult, if not impossible, to compete on the basis of price, given that their customers would be required to pay the CTC and given that the utilities were selling electricity at retail rates 10% smaller than the pre-restructuring retail rates, with the reduction financed using the rate reduction bonds. New entrants did not have the same competitive advantage. They could compete on the basis of price if they marketed electricity primarily to those customers whose loads were less time variant than typical loads, for example, some industrial customers. For such customers, a new entrant could save money on the wholesale purchases of electricity and might be able to sell electricity at a lower retail price than did the incumbent utilities. But, new entrants would have to pick customers carefully. In principle, some retailers could provide higher reliability or interruptible service for particular customers with differing needs. However, because the utility would deliver the


341

electricity, it was not clear that an individual electricity retailer could offer such services without full cooperation from the utility providing the delivery services. Residential retail customers typically would need a reason to switch their purchases of electricity from the well-know incumbent utility, particularly if they would remain dependent on that utility. But the retail restructuring gave very little room for new entrants to offer their potential customers any good reasons to switch. Thus, although the ability to bundle some services with the electricity or to offer an enhanced quality of electricity (e.g., more green, more dependable) provided some opportunities for new entrants, at least during the transition period, these opportunities were quite minimal. Finally, the IOUs during the transition period were not simply passive default sellers of electricity, indifferent to how much they sold. The more electricity they sold, the more rapidly they would collect funds paying their stranded costs. And because they were at risk for any stranded costs not recovered by March 31, 2001, the more electricity they sold, the higher would be the probability that they would recover all of their stranded costs. Thus the IOUs had a financial incentive during the transition to keep as many customers as they could themselves and to limit the role of new retail competitors. Thus, unsurprisingly, the success of retail competition was relatively limited, particularly for sales to residential and small commercial customers. The only significant success of retail competition was for large industrial users of electricity. For these large industrial and commercial customers, the advantages of more flexible, creative contract structures could be great. Direct-access contracts between these large users and independent generators of electricity represented the only significant amount of non-utility-sold electricity. 10.1.7. Municipal utilities California’s many municipal utilities, serving 22% of California’s customers, continued operating as they had prior to the restructuring. Each municipal utility had a governing board, either appointed or elected within the municipality, responsible for managing the utility to benefit residents. Typically, municipal utilities were expected to cover their costs through sales of electricity. The governing boards retained the ability to increase retail prices at which the municipal utility sold electricity, if the need arose. These utilities typically purchased electricity using a mix of short-, medium-, and long-term contracts so they were hedged from rapidly changing wholesale prices. Municipal utilities, therefore, differed sharply from the IOUs in that they retained all capabilities to manage their risks. This difference, as will be seen at a later point, allowed the munis to go through the electricity crisis relatively unharmed, while the IOUs were brought to their knees.

10.2. Market Issues in the Restructured System 10.2.1. Bidding strategies for electricity generators The PX market and the CAISO markets, as discussed above, were one-price auction systems. The theory of bidding in such a market, for a supplier of electricity operating purely competitively, is that the optimal bid must equal the incremental cost of the electricity.29 29

If the bid is above the marginal cost and the market clearing price is below the bid, but above the incremental cost, then the supplier would lose a profit opportunity. (See Carlton and Perloff, 1994.)

342


The incremental cost of an electricity sale on the PX includes the opportunity cost of the PX sale. Since electricity sold in the PX could not be sold in the subsequent CAISO real-time or ancillary services markets, the opportunity cost of PX sales would be the expected price for selling in one of the CAISO markets. Thus, the correctly calculated increment cost for selling electricity in the PX is the higher of three amounts: a) the incremental cost of generating the electricity; b) the expected value of the MCP in the subsequent CAISO real-time market; or c) the expected value of returns30 from selling in the subsequent ancillary services market. Thus, because the subsequent CAISO real-time market, for a perfectly competitive supplier of electricity, optimal bidding into the PX implied that the optimal bid would be equal to the higher of the three amounts.31 For a particular firm, the optimal bid would not be the marginal cost of generating electricity unless it expected the CAISO markets to clear at or below its marginal cost of generating electricity. Whenever, the CAISO’s MCP was expected to exceed significantly the generation cost for an individual supplier, competitive optimal bidding into the PX would significantly exceed the generation cost.32 A similar analysis holds for bidding in the ancillary services market and all CAISO markets that clear before the CAISO real-time market clears. Only in the last of the markets to clear will optimal competitive bids not include an opportunity cost associated with the existence of other markets,33 since at that time, the opportunity to bid in the previous market would have passed. However, the system did not preclude the exercise of market power. A firm with market power could submit bids to the PX and the CAISO above the correctly calculated incremental cost, under the expectation it could increase the MCP by so doing. A high price may imply that not all of the electricity from the firm’s generators could be sold. The sure-to-berejected bid on a portion of the available power could sacrifice profits for that portion but might increase prices on all the electricity it did sell. If the gain on the rest of the electricity were great enough, then bidding a high price only some of the available generation could be a profitable strategy. Such bidding strategies could be profitable as long as the firm had a large enough market share. 30

These returns include the payment for supply the services plus the expected value of revenue from supplying the electricity, if the electricity were needed. 31 The optimal bidding for a competitive firm is even more complex than this suggests. In addition to the PX day-ahead market, there is a day-of market, five different ancillary services markets that clear sequentially, and the ISO real-time market. After the day-ahead market closes and the real-time market closes there is a post-close quantity match that allows additional sales. The opportunity cost would include the price expected from the highest of these markets. But in addition, if each of these price are uncertain, but not perfectly correlated, then the opportunity cost will be greater than the maximum of the expected values. 32 A simple numerical example will illustrate this point. Assume that the cost of natural gas and operating expenses for producing electricity were $50/MWh, but a firm expected the price to clear at $200/MWh in the ISO real-time market. If the firm bid $50 in the PX, and the PX MCP turned out to be $180, the firm would be paid $180. This amount would be $20 less than the $200 it could have expected from the ISO real-time market. Thus the optimal bid in the PX of a firm operating perfectly competitively (as a price taker) would not be $50, but would be $200. However, if that firm were not able to sell in the PX, then the optimal bid in the ISO real-time market would be $50. 33 This comment is not meant to preclude the possibility that there could be an opportunity cost associated with environmental constraints on generation from the plant.


343

Exercise of market power might be possible for some firms, but whether it would be generally be possible was not known at the time of the restructuring. What was known, however, was that those firms that did have market power were likely to exercise that power, since actions to maximize profits are normally to be expected for private sector firms. It was less likely that firms would have market power when the system was well below capacity or when additional electricity could easily be imported into California; it was more likely when capacity was constrained, as proved the case in 2000–2001. Low elasticity of demand for electricity at the wholesale level makes the existence of and the exercise of market power more likely. During the 2000–2001 period, California’s particular regulatory rules for the IOUs34 assured a near-zero electricity demand elasticity. Both opportunity costs and market power lead to bid prices in the PX that can be above the marginal generation costs. Thus the simple observation that PX bid prices significantly exceed marginal generation costs does not tell whether the firm is operating competitively and including opportunity costs or is exercising market power.35 Since both opportunity costs and market power will lead to PX bid prices higher than marginal generation costs, perhaps very much larger, it has been particularly difficult to use PX bidding patterns to infer whether firms were exercising market power in any particular instances. 10.2.2. PX-bidding strategies for utilities Utilities would bid to purchase bulk power in the PX, submitting bids that indicated the quantities of electricity they wished to acquire at various prices. The theory of bidding in such a market, for a purchaser of electricity operating purely competitively, is that the optimal bid must equal the incremental value of the electricity to its buyer. The utility faces a load that is, independent of the MCP because the wholesale price is not passed through to the customers. Thus it must acquire bulk power in either the PX or the CAISO markets. Electricity that it does not acquire in the PX must be acquired in the CAISO at whatever is the MCP in the real-time market. The incremental value to a utility for power acquired in the PX is the lower of two amounts: 1. the incremental value of the electricity, or 2. the expected value of the MCP in the subsequent CAISO real-time market. However, the incremental value of the electricity is virtually unlimited, since the utility is obligated to sell to its customers. Thus, the value to incremental value to a utility for power acquired in the PX is simply the expected value of the MCP in the subsequent CAISO real-time market. A utility, operating perfectly competitively, would be expected to place a purchase bid equal to its expected load at a price equal to the expected price in the CAISO real-time market. The CAISO tariffs, however, if followed in letter and in spirit, limited the ability to substitute between the markets. In particular, as discussed above, each SC was required to submit 34

As discussed at another point in this chapter, three factors led to almost zero demand elasticity: the ISO mentality of reliability at all cost, the normal regulatory lag between increases in wholesale price and retail prices, and the retail price control imposed in California. 35 In the example of Footnote 32, one might observe the firm bidding $200 in the PX and conclude that a bid well above $50 would be evidence that the firm was attempting to exercise market power. Such a conclusion would be incorrect. Unfortunately, that logic mistakenly underlaid public statements and litigations after the electricity crisis, asserting that such bids were clear evidence of market power or of market gaming.

344


a balanced schedule in which the total loads and resources were equal to one another. The balanced-schedule requirement implied that a utility that needed to acquire electricity would be required to purchase in the PX even if it expected prices to be significantly lower in the CAISO real-time market. Thus this balanced-schedule requirement would be in conflict with the theoretical competitive bidding strategy. Utility managers, operating competitively, could be expected to understand that it would be most profitable to acquire electricity from the market with the lower expected price and thus to limit their bids in the PX to the price they expected to face in the CAISO. There could be a strong financial incentive to thereby risk violating the balanced-schedule requirement, submitting an unbalanced schedule if the CAISO expected price were lower than the PX MCP. This conflict between the incentives for unbalanced schedules and the requirements for balanced schedules could be resolved through systematically biasing load projections. By systematically underestimating expected load, a utility could submit a schedule that would in fact be unbalanced but that appeared on paper to be balanced. It could thereby meet the letter, but not the spirit, of the requirement for balanced loads. This became a routine practice for utilities, thus implying that prices on the PX and the CAISO would tend to equilibrate, at least statistically. The price equilibration between the PX and the CAISO implied that wholesale price caps in the CAISO real-time markets (described above) would limit market prices in the PX, even though there were no formal price caps in the PX markets. However, price caps can create shortages. When price caps were binding, utilities might not be able to satisfy all remaining electricity demands through real-time CAISO purchases. Normally the likelihood of a shortage in the real-time market would provide motivation for the utility to purchase in the PX, assuring acquisition of adequate electricity. However, under the CAISO rules, all utilities would equally share any shortage, independently of whether they had purchased enough electricity on the PX. Thus, bidding in the PX above the CAISO price cap would cost the utility more but would provide no additional protection. Therefore, there was no incentive for any utilities to bid above the price cap on the PX and the CAISO real-time price caps effectively controlled the maximum prices on the PX. The system did not preclude the exercise of market power by utilities, the dominant buyers of bulk power. A utility with market power could submit bids to the PX below the expected MCP in the CAISO, under the expectation it could decrease the PX MCP by so doing. A low price may imply that not all of the needed electricity could be acquired on the PX and some must be acquired from the CAISO real-time market, through under-scheduling its loads. The utility could thereby reduce the prices on all the electricity it did acquire on the PX, especially if it could estimate the bids to be submitted by the generators. Such a strategy would require the utility to increase the quantity of electricity it purchased on the CAISO, thus increasing demand for CAISO real-time purchases. However, more electricity would be available for sale on the CAISO since the generators would have sold less electricity on the PX as a result of the utility bidding strategy. Whether the MCP would increase on the CAISO real-time market significantly because of the utilities’ bidding strategy is unclear. However, if any such increase were small enough and the price reduction on the PX were large enough, then such a bidding strategy could reduce the overall cost to the utility of its bulk power acquisitions, at least initially.36

36

However, since generators, in submitting their bids to the PX, could be expected to take into account their expectations of the MCP on the CAISO real-time markets, increases in the CAISO real-time price would increase the MCP in the PX. Thus, it is not clear whether the net impact was to increase or decrease the overall electricity price level once the generators understood the utility bidding strategy.


345

Such a bidding strategy would lead to a pattern of prices that were statistically higher on the CAISO real-time market than on the PX day-ahead market, a pattern that was, in fact observed. It is unlikely that such as possible utility strategy was understood by CPUC officials at the time of the restructuring. However, testimony of PG&E personnel to the CPUC would have informed CPUC members of this strategy well before the greatest price increases occurred.

10.2.3. Market risk 10.2.3.1. Wholesale market risk The restructuring of markets created significant economic risks for the IOUs because the wholesale market for electricity could be very volatile, long-term electricity supply contracts were precluded by regulation, and the utilities faced retail price caps. Wholesale price volatility could result from either energy limitations (limitations on the availability or cost of the energy used to generate electricity) or generation capacity limitations (limitations on the capital equipment to generate electricity). Energy limitations could increase the price of electricity at all times, peak and off-peak. Generation capacity limitations could increase the price of electricity when the system was operating at or near peak, but would have little impact on prices when the system was operating in off-peak periods. Either limitation, if binding, could drive the incremental cost of generating electricity up sharply and thus increase the MCPs. The short-run elasticity of demand for electricity is typically low for end-users. In addition, utilities do not charge customers real-time prices, but rather average-cost-based prices, thus reducing consumer response to wholesale price changes. The wholesale price elasticity is further reduced because of retail price control; wholesale price increases would not be passed on to retail customers, even over an extended time. Measured at the wholesale level, the elasticity of demand for electricity would be near zero. Thus, small reductions in supply or increases in demand could sharply increase prices. The role of demand elasticity is illustrated in Figure 10.6. The supply function is reduced, from the dashed gray line to the dashed black line. If the demand elasticity is very low, as represented by the dashed black line, then the new price will be much higher than the initial price. If, on the other hand, the demand elasticity is higher, the new price will increase by only a much smaller amount. Under the restructured system, all spot sales of electricity would sell at a price equal to the MCP, a price that would be at least as costly as the marginal cost of generating electricity from the most expensive unit being used. Moreover, the utilities were buying most of their electricity on these spot markets, because they had divested at least half of their generating assets and had not entered into long-term electricity supply contracts. Thus, total expenditure for acquiring electricity was very uncertain; risks were large. The mix of energy sources for generating electricity increased the risk that the system could approach energy limitations. During peak demand periods, much of the primary energy used for electricity generation in California was natural gas (see Fig. 2). And the infrastructure of pipelines to move natural gas in California was extremely limited, as was the capacity of pipelines to bring natural gas into California. Therefore, the risk stemming from limitations on natural gas availability or from volatile natural gas prices was great. The second largest source of electricity supply was through hydropower, which is dependent on unpredictable rainfall during the previous year. And much of the electricity imports into California were derived from hydroelectric power in the Pacific Northwest, also subject

346


Electricity price

New price, very low demand elasticity

New price, higher demand elasticity Original price

Quantity of electricity Fig. 10.6. Price increases in response to supply reductions, two demand elasticities.

to wide variations from year to year. The risks were highly correlated: low rainfall in California could be expected during those years of low rainfall in the Pacific Northwest. Divestiture, absent long-term power purchase agreements, increased the allocation of the risk to IOUs. If the utilities had not divested generating capacity, they would have faced cost variations that changed with the average generation cost. But, because they had divested the assets and were purchasing electricity on the spot market, they would face cost variations that changed with the marginal cost of electricity, because the MCP changed with the marginal cost of electricity generation. With divestiture, then, if the marginal cost of generation increased while the average cost of generation increased by only a smaller amount, the IPPs that owned the divested units would enjoy increased profits. At the same time the utilities would see reductions in their profits or would face losses. Absent divestiture, those profits would be earned by the generation side of the utilities while the losses would be incurred by the distribution side. The overall variation in profit would be reduced because losses by one side of the utility would be partially balanced by gains to the other side.37 In general, the marginal cost is much more volatile than the average cost. Thus divestiture exposed the IOUs to increased profit risk whenever there was risk of wholesale market price variations. The absence of long-term power purchase agreements (beyond the pre-existing contracts) increased the risk that firms would possess and exercise market power. Generators could exercise market power by bidding above their (correctly calculated) costs in the PX and CAISO markets or, equivalently, not bidding at all into these markets. By reducing the

37

This discussion assumes that the retail price control remained in effect. Without retail price control, cost variations could be passed on to the consumers through changing rates and the utility would face very little risk in either case.


347

capacity that was made available,38 the MCP would be increased for all other electricity being sold on the spot market. If the MCP increased enough, it would be more profitable to reduce supply of electricity. However, if a large share of the electricity were sold under either long- or medium-term contracts, then this incentive would be sharply reduced. Consider two situations for a given generator, assuming that all other generators are similarly situated. In the first situation, the generator sells 100% of its electricity on the PX and/or CAISO spot markets; in the second situation, it sells 90% of its electricity on term markets and only 10% on the PX and/or CAISO spot markets. Now for the two situations, examine the economics for a generator deciding whether to reduce the supply it offered to the market at the competitive price, either by bidding well above its costs or by physically withholding supply. In the first situation, the firm could influence the price for all of its power, since it would have no term contracts. For illustration, assume it reduced its production by 50% of its total capacity. Assume that this reduction in supply to the spot markets would increase price and would thereby increase the margin of price over cost by a percentage X. If X were larger than 50%, then it would be profitable to reduce supply. The firm would have market power. In the second situation, as the firm estimated its changes in profits, it would recognize that the price of the contractually power would not be influenced by the MCP in the PX or CAISO markets. It could influence the price for only the 10% it sold on the PX/CAISO markets. If it reduced its supply by that entire 10% available for these markets, it would make no profit; thus, its maximum profit could be achieved by a reduction in supply of less than 10% of capacity. For illustration, assume it reduced its production by 5% of its total capacity. Assume that this reduction in supply to the spot markets would increase price and thereby increase the margin of price over cost by a percentage Y. If Y were larger than 50%, then it would be profitable to reduce supply. In the first situation, the firm would reduce supply by 10 times as much as would in the second situation. If all generators were similarly situated, the spot markets would be 10 times as large in the first situation. Thus, it might appear that the firm’s reduction in supply would be the same percentage of the market in either situation and therefore that the price change would be the same in both situations, that is, that X and Y would be identical. Thus, it might seem that there would be the same amount of market power in either situation. Such a conclusion would be incorrect. In the second situation, the 90% of the electricity, although under contract, could be offered on the spot market for resale by the contractual purchaser. Thus, the elasticity of supply of electricity on the spot markets would be increased by the existence of the 90% of electricity under contract, even though the overall elasticity of supply and demand for electricity would be unchanged. Thus, the Y would be substantially smaller than X and it is substantially less likely that firms would have market power in the second situation. The difference between X and Y can be evaluated more completely, by adding an assumption that all contractual buyers act as profit maximizing entities willing to resell on the spot markets whenever that would be profitable. In that case, the amount of electricity available on spot markets in the second situation (although not necessarily actually bid), would be the same as in the first situation. However, the supply reduction from a firm deliberately 38

Note that such bidding strategies for natural-gas fired generation made the capacity constraint tighter, but would have only a small impact on the energy constraint. The natural gas that would otherwise be used by the generator would still remain in the system, either to be used by the particular generator at another time or sold to another generator. The capacity not used at that time would not be available any other way.

348


withholding supply would be 10 times as much in the first situation. Thus, a rough approximation is that the price increase in the first situation (X) would be 10 times as large as the price increase in the second situation (Y). For example, if Y were 10%, if a firm reduced its output it would decrease profit. There would be no incentive to reduce output. However, under this example, X would be 100%. If such a firm with no long-term contracts were to decrease its output by 50%, its profit would increase substantially. There would be a strong incentive to reduce output so as to increase price. Thus with or without term contracts, a firm could bid above its cost, but the incentive to do so would be much smaller with term contracts. In fact, even if it were very likely that a firm without term contracts would have market power, that same firm with term contracts is very unlikely to have market power. The great reliance on the spot market sharply increased the risk that generators would possess market power and thus the risk that generators would exercise market power. The reliance on spot markets thus led to an important risk that wholesale prices would be sharply increased through exercise of market power. 10.2.3.2. Retail market risk The deregulated retail market created incentives for consumers to shift back and forth between regulated utilities and the unregulated retailers if market conditions changed, at least once the transition period were over. Retail customers could choose to buy from competitors of the IOUs whenever those competitors offered electricity at a more attractive price and buy from the utilities when the opposite was the case. The IOUs were obligated to serve all customers, including those that switched back from an unregulated competitor. The retail price that they would charge would be based on the average cost of their acquisition of electricity, once the transition period were over so that there no longer was a retail price cap. If spot prices were lower than the average prices of electricity acquired by utilities, unregulated firms could sell electricity at the lower price and customers would shift purchases away from regulated utilities. With fewer customers, regulated utilities would purchase less electricity on spot markets, thereby increasing their average cost and the prices charged to remaining customers. The price increase would cause more customers to leave the regulated utility, thereby further increasing the price for the remaining ones – the so-called death spiral. If spot prices turned out to be much higher than the average acquisition cost, customers would abandon the unregulated competitors and purchase electricity from utilities, leading to a death spiral for the unregulated competitors. The increased retail sales would lead to increased spot market purchases, thereby increasing average cost and regulated price for the utility. Thus, both the IOUs and the unregulated competitors faced high risk because of the asymmetrical role of the IOUs as the default providers of electricity.

10.3. Western Electricity Crisis and California Financial Crisis Until early 2000, in terms of publicly visible goals, the system seemed to be operating as intended. Wholesale prices remained below historical average costs and there was sufficient headroom for the IOUs to collect the CTC, which they could apply to cover stranded costs. By 1999 SDG&E had collected sufficient CTC to cover all of their stranded costs and would no longer be subject to a retail price cap. As stipulated under AB 1890, henceforth, the CPUC would normally allow SDG&E to pass on increases in the wholesale price of power to its retail customers. However, that change turned out to be politically unacceptable beginning in 2000 when wholesale spot prices surged.


349

In California, applications to construct many new electricity generating units had been filed and approved; construction had begun for many thousands of megawatts of new capacity. However, very little competitive retail market for electricity had developed for small customers. Although several companies entered the market to sell consumers “green” electricity, generated entirely by renewable sources, this component of the market was small. Large industrial or commercial customers, however, were entering bilateral, direct-access contracts with generators. Such large customers had little or no need for non-utility aggregators. Less publicly visible, were operations of the PX and the CAISO. The market structure continued to prove unwieldy. The congestion management approach, relying on zonal pricing, a complex system of adjustment bidding to relieve anticipated transmission congestion, and the opportunity for gaming of the congestion markets, seemed to be fundamentally flawed. The CAISO submitted to FERC a sequence of proposed amendments – 24 by the year 2000 – to its tariffs. A congestion market redesign process was initiated but never completed. In mid-2000, however, came signs that all were not well. Soon after the electricity crisis was upon the entire western part of the US. 10.3.1. The nature of the crisis The “California electricity crisis” was, in reality, two crises: 1. An electricity crisis associated with very tight wholesale markets for electricity in the west and California’s poorly functioning electricity markets and; 2. A financial crisis affecting California’s IOUs, resulting from rigidities of the state’s regulatory control of these utilities. The California electricity crisis was part of the electricity crisis facing the western states. Demands for electricity throughout the West had grown over the years but supply had not. New generation capacity in California was under construction but construction had not been completed. Short-term supply reductions led to very tight markets and wholesale spot prices increased sharply in California and throughout the West. I will focus attention on the California supply and demand issue. For more complete discussion of the western supply and demand balance, the reader is referred to the paper by Fisher and Duane.39 The financial crisis started primarily as a crisis for the investor owner electric utilities and turned into a crisis for the state budget. That is, the financial challenge turned into a financial crisis was the direct result of state regulatory, administrative, and legislative action or, more precisely, inaction. These two crises will be discussed in turn. 10.3.2. The electricity crisis The first indications of the oncoming electricity crisis appeared in May 2000, with sharp wholesale price increases. By December 2000, the difficulties had grown into a crisis, with massive increases in the wholesale electricity prices and with frequent energy emergencies as the operating reserves were at dangerously low levels. The crisis remained severe all winter. Not until late Spring 2001, did the electricity crisis start to disappear, with wholesale 39

A complete discussion of the long-term trends in the West appear in “Trends in electricity consumption, peak demand, and generating capacity in California and the Western Grid, 1977–2000”. Jolanka V. Fisher and Timothy P. Duane, (2001), Working Paper from the University of California Energy Institute Program on Workable Energy Regulation (POWER), September.

350


prices falling, electricity consumption declining relative to 2000 levels, and decline in the frequency of energy emergencies. By Fall 2001, wholesale prices had declined to historical levels and energy emergencies had disappeared entirely. The electricity crisis had passed. 10.3.2.1. Wholesale price increases In May 2000 and in June of 2000 wholesale electricity prices in all of the regional markets reached40 peaks above $400/ MWh. In July, peak prices soared even higher on all of the western markets, with all markets showing highest prices exceeding $500/ MWh during 1 week. Then prices fell, only to peak again, but not as high, in August. After the August peak, prices started falling again and reached a low during the week of November 6. But the crisis had only begun. Spot prices began rising in November 2000. In December prices skyrocketed, exceeding $1000/MWh for the average of the high and low price peaks in some weeks. Prices were even higher in Washington and Oregon than in California.41 The following three figures illustrate these extraordinarily high prices. These figures show prices, estimated in The Western Price Survey, on a week-by-week basis, from January 3, 2000, through September 30, 2001. Data are present for three market centers – at the California–Oregon Border42 (COB), receipt points along the Columbia River (Mid-Columbia), and at switchyard of the Palo Verde nuclear power plant, Arizona – and for the exchanges in California. California prices shown are the PX MCP until January 29, 2001. After that date, as will be discussed later, the PX became defunct so that PX prices were not available. After January 29 prices are given for Northern California (NP 15) and Southern California (SP 15). The first graph, Figure 10.7, shows the average of the low and the high peak prices for western markets, but the vertical scale is truncated to show detail before and after the December spikes. The second graph, Figure 10.8, includes the entire range in order to show how high the average of low and the high peak prices were in December 2000 in the Pacific Northwest. The third graph, Figure 10.9 shows the average of the low and the high off-peak prices for western markets, with the vertical scale truncated as in Figure 10.7. Particularly striking is the observation that off-peak prices, not simply peak prices, were extraordinarily high during the crisis. The average of the low and the high off-peak prices in several weeks exceeding $250/MWh. These figures suggest that different market forces were operational at different times. During peak periods, high prices could result from generation/transmission capacity limitations, particularly during times of especially high peak demands. In addition, as will be discussed at a later point, these high peak prices may be partially the result of exercise of market power by generators who recognized that most generators are already at or near full capacity. However, at off peak times, generation was not near fully capacity and so generation capacity limitations could not explain the off-peak prices. Nor is it likely that the exercise of market power explains these prices, either in California or in the other western markets, given the large amount of unused capacity during off-peak periods. 40

The Energy NewsData publication, Clearing Up reported on June 16, 2000: “Markets throughout the region were drawn into price spikes. Power prices at Palo Verde/Four Corners and Mid-Columbia/COB moved upward in huge increments all week, with almost no overlap from day to day. At the peak of trading on Wednesday Palo Verde peak power was fetching 450–500 mills/KWh. COB and Mid-C had been over 400 mills/KWh.” 41 Although California demand is typically lower in the winter than in the summer, the reverse is true in the Pacific Northwest. And the hydropower reduction was greatest in the Pacific Northwest. 42 The interconnection point at the California/Oregon border of the Pacific Northwest/Pacific Southwest AC Intertie.

$

2800 2600 2400 2200 2000 1800 1600 1400 1200 1000 800 600 400 200 0

31 January 2000

10 September 2001

13 August 2001

16 July 2001

18 June 2001

21 May 2001

23 April 2001

26 March 2001

26 February 2001

29 January 2001

1 January 2001

4 December 2000

6 November 2000

9 October 2000

11 September 2000

14 August 2000

17 July 2000

19 June 2000

22 May 2000

24 April 2000

10 September 2001

13 August 2001

16 July 2001

18 June 2001

21 May 2001

23 April 2001

26 March 2001

26 February 2001

29 January 2001

1 January 2001

4 December 2000

6 November 2000

9 October 2000

11 September 2000

14 August 2000

17 July 2000

19 June 2000

22 May 2000

24 April 2000

300

27 March 2000

350

27 March 2000

400

28 February 2000

450

28 February 2000

31 January 2000

3 January 2000

0

3 January 2000

$

California Electricity Restructuring, The Crisis, and Its Aftermath 351

Mid-Columbia COB California PX California NP 15 Palo verde California SP 15

250

200

150

100

50

Week beginning date

Fig. 10.7. Spot power prices: average of high and low peak prices ($/MWh). Source: Western price survey, www.newsdata.com

Mid-columbia COB California PX California NP 15 Palo verde California SP 15

Week beginning date

Fig. 10.8. Spot power prices: average of high and low peak prices ($/MWh). Source: Western price survey, www.newsdata.com

352

Electricity Market Reform 450 Mid-Columbia COB California PX California NP 15 Palo verde California SP 15

400 350 300

$583, $678

$

250 200 150 100

10 September 2001

16 July 2001

13 August 2001

21 May 2001

18 June 2001

23 April 2001

26 March 2001

29 January 2001

26 February 2001

1 January 2001

4 December 2000

9 October 2000

6 November 2000

11 September 2000

17 July 2000

14 August 2000

22 May 2000

19 June 2000

24 April 2000

27 March 2000

31 January 2000

28 February 2000

0

3 January 2000

50

Week beginning date Fig. 10.9. Spot power prices: average of high and low off-peak prices ($/MWh). Source: Western price survey, www.newsdata.com

However, limitations in energy inputs to generation – natural gas and stored water for hydropower – were relevant in both peak and off-peak periods. Since both natural gas and water were storable inputs, in sharp contrast to electricity which is not storable, their market prices or opportunity costs did not change greatly, if at all, during daily cycles. As will be discussed more fully later, natural gas and stored water (for hydropower) were in short supply throughout the electricity crisis. In addition, within some air quality districts, particularly in southern California, environmental constraints limited cumulative generation and thus also were relevant in off-peak periods, as will also be discussed at a later point. To the extent that data are available, prices were similar among the western markets for most of the time, with the exception of 2 weeks in December 2000 (the week of December 11 and 18). During those weeks, the maximum prices43 during the peak period were about $5000/MWh at Mid-Columbia and $4000/MWh at COB, while the maximum prices on the PX were $1400 and $950. On the other hand, the maximum prices during the off-peak period were about $700 and $1100, respectively, on the PX, but did not exceed $400 during the week44 of December 18 at Mid-Columbia and COB. Throughout January 2001 off-peak prices at the PX substantially exceeded the other western prices. Transmission capacity constraints between the Pacific Northwest and California allowed these price differentials.

43

The average of the high and low peak prices during these 2 weeks were $2700 and $2687 at MidColumbia and $2200 and $2188 at COB. 44 I do not have data for off-peak prices in the week beginning December 11 for COB or Mid-Columbia.


353

4

3

2

1

0

5 June 00 19 June 00 3 July 00 17 July 00 31 July 00 14 August 00 28 August 00 11 September 00 25 September 00 9 October 00 23 October 00 6 November 00 20 November 00 4 December 00 18 December 00 1 January 01 15 January 01 29 January 01 12 February 01 26 February 01 12 March 01 26 March 01 9 April 01 23 April 01 7 May 01 21 May 01 4 June 01 18 June 01 2 July 01

Emergency stage

Rolling blackouts

Fig. 10.10. Energy emergencies and blackouts in California. Source: CAISO.

Note also that prices in SP 15 tracked Palo Verde price most closely; prices in NP 15 tracked COB and Mid-Columbia most closely. Transmission constraints between northern and southern California, along Path 15 allowed the difference between these regions to persist. Figures 10.7–10.9 make it clear that the electricity crisis was not simply a California phenomenon. The high prices – both peak and off-peak – were felt similarly, although not identically, throughout the western electricity markets. 10.3.2.2. Energy emergencies and rolling blackouts The peak of the electricity crisis was marked by energy emergencies in California and the profound fear of rolling blackouts that occurred when some customers were “blacked out” to avoid instability in the entire grid. Figure 10.10 plots the energy emergencies declared by CAISO. Data are on a daily basis; each major division on the horizontal axis represents 2 weeks. Stage 1, 2, and 3 emergencies are indicated respectively, by 1, 2, and 3 on the vertical axis. Rolling blackouts are indicated by the number 4 on the axis. Blackouts came to symbolize the electricity crisis in California. Nevertheless, although the threat of blackouts was frequent, actual blackouts were very rare. The CAISO ordered blackouts in fact on only six separate days, as shown in Figure 10.10. Moreover, blackouts were called for only a small fraction of the load at any time. The most severe was on January 18 in which 1000 MW of load was curtailed, accounting for 3.2% of the peak demand that day. Other rolling blackouts ranged from 300 MW to 500 MW, or 0.9% of the peak load to 1.7% of the peak load. Data on blackouts are shown in Table 10.2. 10.3.2.3. Economic forces during the electricity crisis Supply and demand conditions throughout the western US are tightly linked, as shown by the degree that the wholesale spot prices moved together. Transmission lines, across which electricity can be transmitted, connect adjacent states and adjacent regions within a state, as shown in Figure 10.4. Since electricity generation in any part of the West could serve the electrical loads in

354

Electricity Market Reform Table 10.2. Rolling blackouts in California. Date (all 2001) January 17 January 18 March 19 March 20 May 7 May 8

Curtailment ordered (MW)

Peak (%)

500 1000 500 500 300 400

1.6 3.2 1.7 1.7 0.9 1.1

Source: CAISO.

any other part of the West, electricity supply and demand changes in any part of the West could have price impacts on the entire region. A tight market in the Pacific Northwest could easily translate to a tight market and high wholesale prices in any western state, including California. The following nine economic forces together led the prices to skyrocket and/or placed limitations on market-induced downward pressures on prices. The tenth item was the only one pushing prices downward. ● ●

●

●

●

●

●

●

●

Firstly, limitations on generation capacity created an initial tight market situation. Secondly, extremely low rainfall in the Pacific Northwest and California reduced energy available to generate hydroelectricity. Thirdly, limitations on natural gas supply and increases in natural gas demand led to sharp increases in price in California. Fourthly, generation capacity was reduced as a historically large number of generation plants were offline and not generating power. Fifthly, the cost of marginal generation was increased because of the sharp rise in the price of NOx emissions permits. Sixthly, limitations on demand response implied that small changes in supply or demand could lead to large changes in price. Seventhly, California’s energy market, often described as “dysfunctional”, facilitated market gaming, particular for ancillary services and transmission congestion. Eighthly, increased financial risk reduced electricity supply and led to risk premia for selling to California utilities. Ninthly, exercise of market power and market gaming exploited the tight market situation and the California energy market, further increasing prices throughout the west.

Each of these factors will be briefly discussed below, with a primary focus on California markets and the impacts of the regional markets on California, and only a secondary focus on the entire regional supply and demand balances. Limitations on generation capacity. Each generator is constrained by an upper limit on the physical capacity to produce electricity; the operator cannot generate electricity from a plant at rates beyond physical capacity. This physical capacity constraint leads to a “hockey stick” shape of the electricity supply function: for operation well below capacity, small changes in price can lead to large changes in electricity supply; for operation near capacity, the same changes in prices motivate very small changes in electricity supply. Looked at the other way, for operation well below capacity, small changes in available capacity or demand will have only very small impacts on price; for operation near capacity, small changes in available capacity or demand can have large impacts on price.


355

Total MW capacity (Thousands)

8 7 6 5

On hold Under review Under construction Operational ⬍50 MW Operational ⬎50 MW

4 3 2 1 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

0

Application year Fig. 10.11. New generation capacity in California as of 2004.

Throughout the West there had been very little growth in generating capacity in the decade prior to the crisis, although there had been continuing growth in electricity use. Thus the peak period electricity demand had gradually become closer and closer to available supply.45 After California’s restructuring, permits were granted for many new generation plants and construction had started on many. In the years 1997 through 2001, applications averaged about 4500 MW per year, in comparison to an average of well less than 1000 MW per year between 1980 and 1996. But construction of new generation capacity is always slow, taking several years from application for a permit to generation of electricity. Figure 10.11 shows the California new generation capacity that was initiated in the various years from 1980 through 2004. Data are grouped by the year in which the approval application was first submitted.46 However, none of the plants that had been proposed in 1997 or later were on line at the beginning of the crisis. From May through October of 2000, only 1250 MW of new generation capacity came on line in the west (including western Canada). Only during the last part of the crisis did new generation capacity place any significant downward pressure on prices. Pacific Northwest drought. In 2000 the Pacific Northwest and Northern California experienced extraordinarily low rainfall, following years of particularly high hydroelectric supply. Low rainfall reduced the hydropower that could be generated in those locations.47 The drought can be measured by the runoff at the Dalles, the downstream dam of the Columbia River system. Figure 10.12 shows this runoff48 by year, from 1981 through 2004, 45

Since there were only few electrical interconnections between the West and the rest of the United States, electricity supply was not available from the Eastern United States. 46 Plants under 50 MW capacity did not need CEC approval and thus data is available when they first went on line but not when they were started. For these plants it was assumed that the process was started 2 years before the plant went on line. 47 Similar issues are discussed in Chapter 5. 48 Data from the National Oceanographic and Atmospheric Administration, available on the Web site: http://www.nwrfc.noaa.gov/water_supply/ws_runoff_display.cgi? TDAO3

356 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Percentage of average runoff


Water year (January–July) Fig. 10.12. Water runoff for the Columbia river, at the Dalles.

25


Total California 20

15

10

5

0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Note: Total includes California, Oregon, and Washington.

Fig. 10.13. Hydroelectric generation in electric power industry CA, OR, WA. Source: EIA.

as a percentage of the runoff averaged from 1961 through 2004. The average runoff for a water year (January through July) is 104 Million Acre Feet (MAF). The runoff for January through July 2001 was 58 MAF, just 58% of the average runoff. Not only was 2001 a year with particularly low water, but 2001 had been preceded by 6 years near or above average. Low stream runoff in the Pacific Northwest lead to a 14% decline in hydroelectric generation from 1999 to 2000 and a further 31% decrease from 2000 to 2001. These data are shown in Figure 10.13.


357

8


7 6 5 4 3 2 1 0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Fig. 10.14. Imports of electricity into California.

The reduced generation in the Pacific Northwest implied that there was less electricity available for importation into California during the crisis. Figure 10.14 shows the total imports of electricity into California.49 This figure shows how sharply energy imports into California were reduced in 2000 and 2001, putting great pressure on production within California, and driving up California prices. It should be noted that the 2000 and 2001 data in Figure 10.13 and in Figure 10.14 are for the entire year, while the energy crisis including only portions of each of these years. Thus, generation and import reductions during those times were probably greater than indicated here. Natural gas limitations. The reduction of electricity imports into California implied that the state needed to increase its electricity generation. In California, total electric generation increased 10.5% from 1999 to 2000, decreasing somewhat from 2000 to 2001. Less hydropower was available from California sources. The state therefore had to turn to other sources to generate the needed electricity. The only source available and not otherwise fully utilized was natural gas. Thus the amount of gas-fired electricity generation increased significantly. Many older- and less-efficient plants were kept online. Since these older units have higher heat rates than the newer units, using older units led to an increase in natural gas demand per MWh generated. The two effects – an increase in megawatt hours of electricity generated using natural gas and an increase in the natural gas used per megawatt hour of electricity generation – amplified each other in increasing the demand for natural gas in California. 49

In 2001 California changed its categorization of production, leading to changes in the official data series for imports. Prior to that year utility-owned shares of coal, nuclear, and some firm contract generation from outside California were include as internal to California. Since 2001 most of this data is included in the energy imports category. The data presented here for 2001 and beyond are energy imports estimated using the same methods as in 2000.

358

Electricity Market Reform 250 95–96 97–98 99–00 01–02

Billions of cubic feet

225

96–97 98–98 00–01 02–03

200 175 150 125

ril

ay M

Ap

Ju n

e Ju ly Au Se gus t pt em be r O ct ob N ov er em b D ec er em be r Ja nu a Fe ry br ua ry M ar ch

100

Fig. 10.15. Delivery of natural gas to California, all users. Source: EIA natural gas monthly.

As a result, natural gas demand increased sharply. Figure 10.15 graphs deliveries of natural gas to all users in California, by month, for various years. Each line begins in June of 1 year and extends through May of the next year, showing the difference between the time of the electricity crisis in California and other years, before and after the crisis. Natural gas deliveries to all California users in the period December 2000 through April 2001 was 19% higher than in the corresponding period, December 1999 through April 2000.50 Natural gas used in California comes primarily from Canada and secondarily from California sources, the Rocky Mountains, the US Southwest, and storage withdrawals. These supplies were all limited. Natural gas in storage in California and throughout the West at the beginning of the 2000–2001 heating season was well below normal levels. Beginningheating-season storage was 22% below the average of the previous 5 years in California and 30–40% below the average in most of the West.51 An explosion in the El Paso pipeline plus subsequent reductions in capacity of that pipeline reduced the available gas carried into California from the Southwest. Capacity limits on the pipelines from Canada limited the increases of Canadian gas. The resulting increase in demand for natural gas tightened gas markets throughout the western US, especially in California. As a result natural gas prices increased sharply. Part of the reason for the large increase in the natural gas prices was the low elasticity of demand for natural gas. A large fraction of the demand for natural gas was derived from the demand for electricity. But since electricity wholesale demand was virtually inelastic in California, 50

Energy Information Administration (EIA) “Natural Gas Monthly”. A similar pattern can be seen from data on deliveries Washington, Oregon, plus California. These delivery figures, it should be remembered, represent the equilibrium increase in demand for natural gas; the increase in the demand functions for natural gas were greater than the increases shown here, since natural gas prices increased sharply, reducing actual consumption. 51 Source: Energy Information Administration, “Electricity shortage in California: issues for petroleum and natural gas supply”.

359

8/1/01

7/1/01

6/1/01

5/1/01

4/1/01

3/1/01

2/1/01

1/1/01

12/1/00

11/1/00

10/1/00

9/1/00

8/1/00

7/1/00

6/1/00

5/1/00

4/1/00

3/1/00

SoCal gas, large pkgs price PG&E citygate price

2/1/00

60 55 50 45 40 35 30 25 20 15 10 5 0

1/1/00

$/MMBtu


Fig. 10.16. Natural gas spot market prices. Source: Gas daily annual price guide 2001.

that component of demand was also virtually inelastic. Natural gas demand for residential uses and commercial/industrial uses in California could be expected to have a small, but non-zero, elasticity of demand. Therefore, the overall short-run elasticity of demand for natural gas was very small. For that reason demand increases or supply decreases led to disproportionately high increases in the natural gas price. Figure 10.16, based on data from Gas Daily Annual Price Guide, shows increases in the California natural gas spot prices from the beginning of January 2000 through August, 2002. Prices are shown for two delivery points to utilities in California. Figure 10.16 shows that natural gas price at both California locations more than tripled from January 2000 through mid-November 2000. Then in late November and December, the California natural gas price increased even more sharply, to about $50 per million Btu in December. Although this price peak lasted for only 2 weeks, the spot natural gas prices in California remained for most days above $10 per million Btu until May 2001. This increase in the natural gas price greatly increased the cost of electric generation. The lowest cost impacts were on new generation plants. With heat rates for a new combined cycle power plant of 6.8 million Btu per MWh, every $1/MMBtu increase in the natural gas price increased generation cost by $6.8/MWh. For elders, less efficient plants with twice the heat rate, every $1/MMBtu increase in the natural gas price increased generation cost by $13.6/MWh. Thus a $40/MMBtu increase in the price of natural gas would increase the cost of operating such older plants by $544/MWh. Increases in the price of natural gas led to increases in the price of electricity in both peak and off-peak periods, since natural gas generating units generally set the MCP. However, since the higher heat-rate units were likely to be the marginal units at peak times and somewhat lower heat-rate units were likely to be the marginal units at the off-peak times, the increase in natural gas price had a greater impact on electricity prices during the peak hours. Generation plants offline. In California, large numbers of generators went offline in during the late Fall 2000 and winter 2001. Historically, between 1000 and 6000 MW average daily

360


generating capacity would normally be off-line in a month. However in the period between October 2000 and May 2001, a monthly average of 12,000 MW generating capacity was offline, reaching a peak out of 14,000 MW in April 2001. The generating capacity off-line resulted from a combination of causes. However, the greatest number was off-line reportedly for repairs or maintenance. But whether the maintenance and repairs were forced upon the generators, were part of a competitive cost minimizing solution, or were designed to increase wholesale price has not been fully resolved. Outages of the older natural-gas-fired units would have increased prices at the peak period, but not significantly in the off-peak periods. During off-peak times there was sufficient electric generation capacity available so that the binding constraint on production was the availability of natural gas, or for an individual plant, the price of natural gas. A gas-fired plant off-line did not use any natural gas; natural gas would be released for other generation. Thus there would be little or no impact on the off-peak electricity price. A new plant with a low heat rate would have had some impact on the off-peak prices, as well as on the peak prices. The plant next in the merit order was likely to have a higher heat rate. Thus a newer plan going off-line would result in increases in natural gas use for a given amount of generation, would increase natural gas price, and would thus increase the off-peak electricity price. At peak or near-peak times, generators would earn rents on the capital equipment. The less capacity available, the higher the MCP of electricity. Thus, in peak times plants going off-line would increase electricity prices. A smaller, but significant number of the offline plants – perhaps up to 3000 MW – were not old gas fired units but rather were the QFs whose operators were not being paid for the electricity they were selling the IOUs. When these plants went off-line they would increase the demand for natural gas, thus increasing both peak and off-peak prices. Thus, some of the capacity reductions, but far from a majority, was the result of uncertainty imposed on the QFs and the reduction in their cash flow. Another outage was an accident at SCE’s San Onofre nuclear power plant in February 2001, taking over 1000 MW offline for several months. Reductions in nuclear power output increased both peak and off-peak prices, since its shut down would not release natural gas for other users. Environmental limitations. In 1993 the California South Coast Air Quality Management District instituted the REgional Clean Air Incentives Market (RECLAIM) to force reductions in emissions of two criteria pollutants, NOx and SOx. The RECLAIM program required emitters of NOx and SOx to acquire enough RECLAIM Trading Credit permits to match their actual emissions each year. Through 1998, there was an excess supply of permits but during 1999, the number of allocated permits diminished to equal total emissions. Through compliance year 1999, the price of NOx trading credits ranged from $1500 to $3000 per ton. In 2000, prices for the first 10 months of the year 2000 increased to an average of about $45,000 per ton.52 This price increase stemmed from both the further reductions in the available permits, thereby reducing their supply, and increases in the amount of electricity being generated using gas-fired units, thereby increasing the demand for RECLAIM credits.53

52

“White Paper on Stabilization of NOx RTC Prices”, South Coast Air Quality Management District, Web site: http://www.aqmd.gov/hb/010123a.html 53 This increase in the increase in electricity generated using natural gas will be discussed more fully at a later point.


361

A new baseload gas fired generating unit, releasing about 0.1 pounds of NOx per MWh of electricity, would pay only $2.25/MWh at a price of $45,000 per ton. But old gas-fired turbines, emitting up to 4 pounds of NOx per MWh, would face a cost of RECLAIM credits of $90/MWh. As the market-clearing prices in both the PX and the CAISO were based on the highest accepted bids, this cost of RECLAIM credits added to the price for all electricity whenever bids based on generation from these old units were accepted. Since the old units were used primarily in peak periods, the cost of RECLAIM credits increased electricity price by the greatest amount during peak periods. Limitations on demand adjustment. The nature of the supply and demand adjustments in this market implied that small variations in supply or demand for electricity could lead to disproportionately large variations in price. In no states did retail prices change simultaneously with wholesale prices, thus demand adjustment did not serve to reduce the very short run (i.e., hourly) price increases. In California, however, retail price controls assured that retail prices did not change at all until 2001, and then by very little, relative to the wholesale price increases. These price controls thus greatly reduced the longer-run demand adjustments and stopped longer-run demand adjustments from limiting longer-run price increases. Thus limitation on retail price adjustment blunted the incentives on users of electricity to reduce their consumption; thus demand reductions, which would otherwise have placed downward pressure on prices in either the very short term or longer term were very limited. California’s dysfunctional energy market structure. As discussed above, California’s energy market structure, typically described as “dysfunctional,” included a complicated set of wholesale markets, not well coordinated with one another, and congestion markets in severe need of redesign. Since these markets were given monopoly or near monopoly status, California utilities were forced to purchase through them. The CAISO was charged with avoiding blackouts, almost at any cost. As such, when shortages seemed likely, the CAISO would search for electricity outside the California electricity markets, which it purchased at very high prices. The anticipation of selling at high prices to the CAISO provided an incentive for electricity traders and brokers to hold back their supplies or demand greatly elevated prices, confident that they would ultimately be able to sell at high prices to the CAISO. The structure forced exceptionally large dependence on the spot market. The interplay of these various markets, the resulting bidding strategies of generators of electricity, electricity marketers, and traders, and the responses of the CAISO personnel together created a very complex market, rife for market power and market gaming. Risk created by the financial crisis. Owing to the financial crisis, discussed subsequently, utilities began delaying payments to electricity generators, promising to pay later. However, it was becoming clear that absent state policy measures, utilities were unlikely to be capable of paying for electricity in a timely manner, if at all. This created significant uncertainty among suppliers. Uncertainty of future payments would lead a rational competitive supplier to increase the price at which it was offering to sell supplies into the California market, further increasing the wholesale prices. This phenomenon was particularly important in the November 2000 through January 2001 period when utilities were not paying for electricity and their credit ratings were declining sharply. In January, the state of California stepped in as the creditworthy buyer, seemingly guaranteeing the payment for all electricity purchased on behalf of the utilities.

362


But California refused to pay its own spot market wholesale power bills. During 2001, continuing non-payment caused some suppliers to drop offline; continuing risk of nonpayment caused others to include a credit premium in their bids. FERC subsequently approved a 10% credit premium to compensate for continuing financial risks for sales to California. Exercise of market power and market gaming. A final element is the incentives for exercise of market power by electricity generators selling into California markets, for gaming of the system by electricity marketers, such as Enron, and for exercise of market power by the highly concentrated set of wholesale buyers of electricity. Exercise of market power and gaming was facilitated by the complex rules of the California electricity markets. As discussed above, a supplier with market power would have an incentive to bid into organized California markets at prices higher than its correctly calculated incremental cost or could refrain from bidding some units. This incentive was strongest when the entire system was near capacity, as was the case during peak periods of the crisis. The degree to which exercise of market power was responsible for increased prices is currently the subject of many investigations and litigations. Serious research work, such as work of Borenstein et al. (2002) and Joskow and Kahn (2002) attempt to quantify market power and conclude that prices were elevated above marginal cost during the crisis as well as before the crisis, that is, that market power led to significant price increases. Work by Hogan and Harvey (2001) questions these conclusions. In what follows I discuss reasons why market power may have been less important during 2001 than is generally concluded in the political discussions, even though it was some part of price increases. The IOUs were dominant buyers of electricity. However, they had virtually no control over their loads, other than through their advertising, in conjunction with California state advertising, urging consumers to conserve electricity and to shift their use of electricity to off-peak periods. But, as discussed above, they could attempt to exercise market power by reducing their price bids in the PX and under-scheduling the load submitted to the CAISO, thus modifying their mix of demands between the PX and the CAISO markets and reducing PX MCP, at least initially. In mid-January 2001, California Governor Davis ordered the state Department of Water Resources (DWR) to start purchasing wholesale electricity on behalf of the electric utilities, who by that time were no longer creditworthy buyers. Other than electricity self-generated or purchased through pre-existing contracts, IOUs acquired all their electricity through the DWR. This change fundamentally altered the market structure in California, shifting market power toward the state. Instead of competing buyers of electricity, buying on the PX or CAISO, the DWR become the dominant buyer of electricity in California. The DWR could choose how much electricity to purchase in forward contracts and how much to acquire on the real-time CAISO market. If DWR chose, it could acquire all electricity through bilateral contracts outside the CAISO. As dominant buyer, the DWR could not change total electricity consumption. However, it could negotiate forward contracts of varying time duration, with various sellers. In addition, it was not required to disclose prices it paid for particular transactions.54 DWR could price discriminate and exercise its market power as a dominant buyer.

54

Data on purchase prices ultimately had to be released, but the releases were enough after the time of transactions that the information would have little or no value to the generators trying to defend against discriminatory pricing.


363

90 Cumulative percentage

80 70 60 50 40 30 20 10 0

Mirant $230 Dynegy $187 Williams $252 Powerex Corp. $425 AES $158 Duke Energy $128a LADWP $276 Sempra Energy $366 DWR $129 BPA $229 TransAlta $298 El Paso $210 BP Energy $244 MIECO, inc. $200 Public Service of New Mexico $223 Constellation Power Source $118 Reliant Energy $236 Merrill Lynch Capital Services $202 Calpine Energy Services $165 SMUD $268 City of Burbank $306 Pinnacle West $195 Coral Power, L.L.C. $336 Eugene Water & Electric bd $381 Enron $181 Morgan Stanley Capital Group $90 Commonwealth Energy Corp. $181 Salt River Project $118 PG&E Corp. $168 Grant County PUD $335 Trans Canada Power Co $307 Nevada Power Company $182 No. California Power Agency $161 Tuscon Electric Power $156

100

Fig. 10.17 DWR electricity purchases, January through June 2001. Source: DWR.

Figure 10.17, based on data released by DWR, illustrates the large number of entities from which DWR was buying electricity and the wide range of prices it paid. Each bar represents an entity selling electricity to DWR during the first 6 months of 2001. Black bars represent public entities and gray bars represent private entities. Entities selling the most electricity are furthest to the left. The bar heights show the cumulative fraction of electricity purchased, including that seller. Mirant supplied 26% of the DWR purchases. Dynegy supplied 8%; the cumulative supply for Dynegy and Mirant is 34%. Williams sold another 8% and thus the cumulative percentage is shown as 42%. The graph shows that 4 sellers accounted for almost 50% of the purchases by DWR, 11 sellers accounted for 75%, and 21 sellers together accounted for 90%. Thus the market structure was one in which a single dominant buyer, DWR, was purchasing electricity from many competing sellers. Average prices of electricity sold to DWR are shown above the names of sellers. The prices negotiated by DWR varied widely across sellers. Of the four largest sellers, DWR paid Mirant an average price of $230/MWh, Dynegy an average price of $187/MWh, Williams, $252/MWh, and Powerex $425/MWh. The highest price of these four was more than twice the lowest. Among the sellers accounting for 90% of the transactions, DWR paid prices ranging from $128/MWh (Duke Energy) to $425/MWh (Powerex Corporation). Monthly data show the same pattern of price discrimination for each month. DWR was willing and able to price discriminate among the sellers. A skillful price-discriminating dominant buyer, purchasing from a number of competing sellers should be able to reduce the total acquisition cost, even without reducing the total amount of electricity purchased, although it could not reduce the acquisition price below the prices at which the generators could sell electricity out of state, where there were competing buyers. And DWR had experienced, capable staff members. Thus, it appears as if DWR was able to exercise its market power as a price-discriminating dominant buyer. Once DWR was purchasing electricity, there could be no exercise of market power on the PX, since purchases were no longer going through the PX auction. Market power could still be significant for the CAISO purchases. But the DWR could assure that few purchases were

364


needed through the CAISO. In addition, for a period of time the CAISO was providing DWR with confidential information on generator bids, giving DWR and additional information advantage, until CAISO was ordered to stop that practice. In its acquisition of electricity, DWR’s ability to price discriminate reduced any residual market power. If it suspected that a seller was trying to manipulate market prices, the DWR could discipline that seller by purchases from other sellers, even those offering electricity at higher prices. Such a strategic purchase option implied that as of late January 2001, a generator could exercise market power only (1) for sales through the CAISO real-time market and ancillary services markets or (2) through actual reductions of generation. Undisputed, however, is the conclusion that in addition to market power to increase the price of electricity, there were other possibilities of market gaming, particularly gaming of the ancillary services markets and of congestion prices. The famous Enron “smoking gun” memo detailed some methods by which Enron gamed markets, in particular the ancillary services55 markets and the markets for transmission services. Although some of the strategies were complex arbitrage schemes, some created congestion in transmission lines and thus increased electricity prices. Energy conservation. One force operated to reduce wholesale energy prices during this “perfect storm”: the incentives and information campaigns encouraging energy conservation and time shifting of peak load. Reductions in the demand function reduced prices in both the peak and off-peak periods. Electricity demand, both measured in terms of peak demands and total megawatt hours of electricity, declined from 2000 to 2001. During the Spring and Summer of 2001, reductions in electricity consumption and peak demands took pressure off the tight electricity market. Figure 10.18 shows the reduction in average electricity consumption and in peak demand, based on California Energy Commission (CEC) data.56 The bars show the peak demand reductions on a month-by-month basis from 2000 to 2001; the lines show the reductions in average electricity use. Monthly peak electricity demand was reduced, on average, by 1900 MW (4.4%) and the monthly average use was reduced by 1200 MW (4.3%). These demand reductions reduce price in both the peak and the off-peak periods, reducing the use of generation capacity and of natural gas needed for generation. In addition, there were reductions in electricity use throughout the west. In the Pacific Northwest, in particular, there were significant numbers of industrial shutdowns, including in the aluminum industry, which together reduced the demand for electricity in the west. These demand reductions stem from a combination of factors – expectations of increased electricity prices, high retail natural gas prices,57 the energy demand-management programs,

55

In order to manage and electric supply system, given the special characteristics of electricity, the system operator needs agreements with generators to provide generation reserves, reserves that could be called upon at short notice to increase or to decrease the total generation of electricity. These reserves plus additional resources or loads that can be controlled to keep the system stable are referred to as “ancillary services”. 56 Data are published by the CEC on its www page. 57 Many customers, particularly in Northern California, received one single bill for natural gas and electricity purchases. When the natural gas prices increased, newspapers carried storied energy bills increasing. Although a reduction in electricity use in response to an increase in natural gas price is not what economists normally predict, it seemed to be occurring in California. Perhaps subsequent empirical work will be able to examine whether this phenomenon in fact occurred in significant amounts.

Reduction in peak MW, 2001 versus 2000 Reduction in average MWh, 2001 versus 2000

December

365

November

October

September

August

July

June

May

April

March

February

5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 (0.5) (1.0) (1.5) (2.0) (2.5) (3.0) (3.5)

January

MW or average MW (Thousands)


Fig. 10.18. Reductions in electricity use, 2001 versus 2000. Source: CEC.

energy efficiency and conservation programs, publicity about electricity problems, and the decline in the California economy. 10.3.3. The financial crisis The high wholesale prices, the lack of long-term contracts, the divestiture of utility generation assets, and the retail price control together created a tremendous financial stress on the utilities. By Spring of 2000, they had divested about 60% of their generating assets. Divestiture itself would have not been particularly harmful if the utilities had replaced these generating assets with long-term fixed-price contracts to acquire the electricity, but they had not been allowed to do so. Both PG&E and SCE made requests of the CPUC and appealed to Governor Davis to allow them to enter such contracts. The initial response was negative. In June 2000 the CPUC voted to allow the utilities to buy power outside the CAISO and the PX markets. However, the California Legislature subsequently overrode the decision. Thus the utilities were forced to continue purchasing and selling all electricity on these two markets. In August 2000, once prices had jumped sharply, the CPUC did file an order allowing the utilities to enter limited amounts of bilateral contracts. Even then, any purchases other than from the PX would either need pre-approval or face an after-the-fact reasonableness review, while all purchases from the PX were considered per se reasonable. This failure to encourage long-term contracts and the resultant over-reliance on spot markets primarily led to the financial crisis. The over-reliance on spot markets assured that when wholesale prices began to soar, the utilities would pay these high prices for almost half of the electricity they would need to serve their loads. This lack of risk management options was central to the financial crisis. The utilities could have avoided a financial crisis if they had been allowed to raise their retail rates to cover the increased wholesale costs. Consumers and businesses would have faced higher prices of electricity but the price increases would not have been of crisis proportions.

366


However, the state of California continued to enforce58 the retail price control established through AB 1890. This continued enforcement assured the financial crisis would become grave. By Fall 2000, the utilities were selling electricity to their customers at an average price of about $65/MWh for the electricity (plus another $60/MWh for transmission, distribution, and other delivery services) yet the wholesale price of electricity ranged from $150/MWh to $1000/MWh for half of the electricity they were selling. Almost one-half of the remainder, from the QFs, was already being purchased at high prices. Generation from their own hydroelectric assets and nuclear power plants accounted for only 30% of their electricity. The accumulations into the transition accounts to pay for stranded costs quickly reversed, with large negative CTC. And as a utility, they could not refuse to sell electricity. Each of the utilities requested increases in retail rates to cover the average cost of the electricity they were selling. Until January 2001, no rate relief was made available to the utilities. In fact, SDG&E, the only of the three major IOUs that under AB 1890 would be no longer was subject to retail price controls, remained free of the controls for only a limited time. The California Legislature voted to re-impose controls. As of December 2000, the financial crisis had come to a head. PG&E ultimately filed for reorganization under Chapter 11 of the bankruptcy code in April 2001; SCE teetered on the brink of bankruptcy for month until that utility negotiated a rate agreement with the CPUC. Once the utilities were no longer creditworthy, the state, through the California DWR, became the buyer of electricity, as noted above. Under Governor Davis’ order, the DWR would begin purchasing electricity on behalf of the utilities, the retail prices would remain low, and the high cost of wholesale power purchases would be borne initially by the state treasury. The state would subsequently issue long-term revenue bonds to reimburse the state treasury. Repayment of these bonds would be a surcharge on retail electricity prices and thus ultimately the ratepayers would pay all of these costs. In essence, the state temporarily put itself in the place of the IOUs, except that the Governor knew that ratepayers would later repay all state losses. After the state became the primary buyer of electricity, there were no longer any transactions on the PX and that institution had no way of raising money to pay its costs. The PX ultimately declared bankruptcy in March. One of the two market institutions established by AB 1890 was thus eliminated.59 The significance of California regulatory controls to the financial crisis can be seen by comparing the financial impacts on the IOUs and Munis within California and those throughout the western states. The electricity crisis was a phenomenon across the West; increases in spot wholesale prices were similar throughout the western states. Yet only the IOUs of California faced a fundamental financial crisis. Some IOUs in other states and some municipal utilities in California faced short-term difficult financial problems but others were not brought to the brink of bankruptcy. Only the California IOUs faced the double whammy: they purchased over 50% of their wholesale electricity on the highly volatile spot market and they were constrained by retail price control. These two issues were the product of California’s unique regulation of the utilities; in no way were they necessary components of deregulation.

58

This provision of AB 1980 could have been overridden during the crisis either by legislation or by emergency decree of the California Governor. 59 The bankruptcy of the PX had no impact on the market because all purchases, other than through the CAISO were through bilateral trades. The PX was no longer being used.


367

10.3.4. State and Federal regulation changes during the crises State policy and Federal policy sharply conflicted during the crisis, making any solution even more difficult. California political leaders emphasized direct government intervention in the market place, reliance on retail and wholesale price controls, plus strong regulatory intervention. The CPUC, by then with changed leadership, stopped trying to improve markets, strongly enforced retail price controls (which the state could control), and lobbied for wholesale price controls (which the state could not control.) The California Governor relied on a public relations campaign of blaming California’s electricity problems on the Federal government, including Federal regulators, on electricity generators, on “deregulation”, all but on his own policy inaction. When the Governor and Legislature did respond to the financial crisis, their response relied on direct governmental wholesale purchases of electricity and negotiations to acquire assets of the utilities, increasing state participation in electricity markets. Federal action, through FERC, though slow, sometimes misdirected, and inconsistent over time, seemed designed to address the underlying flaws in the market design and implementation and to strengthen the role of markets. FERC operated to reduce reliance on direct control of market transactions and market pricing. The state and FERC each had jurisdiction over important parts of the restructured system so neither could fully impose its views on the other and each needed actions of the other to be fully effective. The fundamental ideological differences were never fully resolved. However, ultimately the FERC did develop a set of market-mitigation measures, including limits on wholesale prices.

10.3.4.1. Federal responses: FERC rule making November 1 and December 15, 2000 Orders. In the first major rulemaking, on November 1, 2000, FERC issued a market Order proposing remedies for California’s wholesale electricity markets and on December 15, with an Order directing the remedies. FERC reiterated its earlier conclusion that the wholesale prices were not “just and reasonable” and identified problems of market design as fundamental to the problem.60 A first goal of the Orders was to reduce the reliance in California on spot markets for wholesale electricity. FERC eliminated the requirement that the IOUs buy and sell all power on the PX or CAISO, thereby in theory permitting utilities to enter bilateral forward contracts. Utilities could use their self-generated electricity directly. Although an important long-term change, given the financial crisis, its short-term impact was limited to allowing the utilities to use directly self-generated electricity. And that change probably had little or no impact on either the energy crisis or the financial crisis. FERC did impose strong incentives on utilities and generators to complete almost all scheduling of load and generation with CAISO ahead of time, rather than in real-time transactions. FERC announced a large penalty on all real-time energy transactions greater than 5% of the prescheduled amount. CAISO ancillary service market rules were modified so that a supplier bidding into this market would receive only the energy payment if it were called upon to deliver energy, not the capacity payment in addition. These modifications were designed to make market scheduling more rational, the acquisition process more deliberate, and energy emergencies less frequent. 60

Language from the November 1 Order.

368


The FERC Orders rejected California’s proposal to give the PX authority to impose wholesale price caps. It denied the CAISO request to continue managing price caps, but ordered the CAISO to impose a $250 price cap on its purchases, without modifying the $100 price cap it had imposed for replacement reserves, until the end of December. FERC ordered that all single-price auctions run by the CAISO or the PX be temporarily modified to a hybrid system, often described as a “soft price cap”. A single market-clearing price would be used if it were no greater than $150/MWh. If the market did not clear at $150/MWh or below, suppliers bidding less than $150/MWh would be paid $150/MWh and those bidding above would be paid their actual bid. This rule changed optimal bidding strategies for those times participants expected that the market would not clear at a price at or below $150/MWh. In those cases optimal bidding strategies would be the same as in any as-bid auction system, not like those in the previous single-price system. There would be a strong incentive to increase the bid price. The PX and CAISO were required to report confidentially to FERC all bids in excess of $150 and each seller was required to file bid price, electricity quantity, and marginal generation cost of each such transaction. Bidders above $150/WHh would be required to justify their bids. Lack of valid justification could result in the supplier refunding the higher price. Such a requirement for bid justification could provide strong motivation if generators believed that price bids above costs would be detected and that refunds would include a penalty in addition to the difference between price and marginal cost. However, if generators believed that bids well above costs were unlikely to be detected and that, if detected, the only penalty would be a requirement to refund the difference, then this requirement would provide little or no motivation for limiting bids. Wholesale prices increased after these Orders were issued. April 26, 2001 FERC Price Mitigation Order. By Spring of 2001, FERC agreed to impose more rigid wholesale controls in California, but not in the other western states. The April 26 Order established a “mitigation and monitoring” plan for the California wholesale electric markets, to become effective May 29, 2001. The plan required that all planned outages by units with Participating Generator Agreements61 (PGAs) with CAISO be coordinated with, and approved by, CAISO. Such a requirement would presumably make it easier to judge whether outages of generating units were for manipulative reasons. The price mitigation plan retreated from the hybrid auctions, moving back toward singleprice auctions for the CAISO (the PX was no longer in operation.) All price caps would be eliminated and replaced with “bid caps”, limitations on maximum seller bid prices. Bids could be no higher than an administratively determined estimate of the marginal cost of operating the plant, based on the unit’s heat rate and emissions rate, prices of natural gas and emissions credits, and a $2/MWh fee for operation and maintenance costs. Bids would determine market price through a single-price auction. Bid caps would be imposed only when an energy emergency had been declared. No caps would be imposed in the absence of an energy emergency, in the theory that competitive forces would be sufficient in those times. Exceptions to bid caps were allowed if a firm could establish that its cost was greater than the cap. Such bids would not help determine market-clearing prices. Firms would be 61

A PGA is a legal agreement between a generator and the ISO that establishes “the terms and conditions on which the ISO and the Participating Generator will discharge their respective duties and responsibilities under the ISO Tariff”. The agreement allows the Participating Generator to schedule energy and submit bids to ISO through a Scheduling Coordinator.


369

required to justify their costs or pay refunds. Similarly, resources located outside California could bid, but their bids would not determine MCP during mitigated periods. The plan imposed a “must offer” requirement on fossil-fuel plants,62 a requirement that all sellers with PGAs offer all their available power to the CAISO in real time if not already scheduled. This rule was designed to stop generators from exercising market power by not bidding at all. The Order prohibited other bidding practices if they were seen as potentially allowing anti-competitive market gaming, for example, bids that vary with unit output in a manner not related to its cost characteristics. Finally, the Order made it clear that demand response should be part of the market response. Load- serving entities were required to establish demand-response mechanisms in which they identified the price at which load should be curtailed. FERC June 19 Order broadening scope of price mitigation. The April 26 Order applied only in California, ignoring the fundamental linkages between all of the western markets. This problem was corrected in the June 19 Order. That Order extended the price mitigation plan to all western states and to all times and expanded its scope. The new Order was scheduled to remain in effect until September 30, 2002. The June 19 Order set rules to be implemented in a common manner63 across all of the markets in the western region. The bid caps and the must-offer requirement would remain during energy emergencies.64 But the bid cap would no longer allow costs of emissions credits or other emissions costs.65 These costs and start-up fuel costs could be directly billed to CAISO and thus would have no impact on CAISO MCP. It also imposed price caps on bilateral sales. The MCP in the single-price auction would set an upper limit on the allowable prices for these transactions. Generators and marketers would be treated differently. Generators could justify costs higher than the administratively determined bid caps, but these higher costs would not be used in setting the MCP. Marketers would not be allowed to a price higher than the MCP. That is, they would be required to act as price takers. For non-energy-emergency times the price for spot market sales could be no greater than 85% of the highest CAISO hourly market-clearing price established while the last Stage 1 emergency (not Stage 2 or 3) was in effect. 10.4. California After the Crises Although the electricity crisis itself was a short-term event California still faces risk of inadequate supply. The financial crisis has left a continuing legacy, the financial overhang, as the historical costs are allocated, partly by regulation, partly by litigation. 62

The requirement would not be imposed on hydroelectric plants. In price control times a 10% price premium was allowed for sales to California, to compensate for market risk. 64 FERC: “Order on rehearing of monitoring and mitigation plan for the California wholesale electric markets, establishing west-wide mitigation, and establishing settlement conference,” Issued: June 19, 2001. 65 Such a system would no longer assure that the lowest-cost units would have their bids accepted. In particular, emissions and start-up costs would not count in determining which plants would be dispatched. Unfortunately, this would create a bias toward selecting the plants with less good environmental performance if their costs otherwise were low. 63

370


The experience in California has also led to understanding, albeit painful understanding, of the dynamics of the California electricity markets and of the role of the market/regulatory rules in influencing the dynamics. This understanding has motivated still-continuing “reforms of the reforms,” with changes being made by the legislature, administration, CPUC, CEC, and CAISO. Some of those reforms have been implemented; some issues are still being debated; some are not yet fully understood. For the most part, changes have been motivated by the improved understanding of the California market dynamics. Some groups in California, however, have reached very different conclusions about appropriate further reforms and have initiated a series of proposals to sharply limit competition and to return towards the vertically integrated utility monopolies. And finally is the problem of distortions in public perceptions about the nature and causes of the crises. In what follows, I discuss the current California system. 10.4.1. Current conditions Since the crisis, demand for electricity is California has continued to grow, albeit slowly, while supply growth has been limited. As a result, the overall supply/demand balance has become tighter, particularly in Southern California. The risk of supply shortages in within the next few years is real, absent reductions in growth of electricity use, reductions in peak demands, or increases in supply to California. The financial crisis of the IOUs has been resolved. But consumers currently face high retail prices of electricity, through surcharges design to recover the historical costs. Increasing nationwide natural gas prices can be expected to further increase wholesale costs and retail prices. 10.4.1.1. Electricity generation capacity The restructuring led to increased investments in new generation capacity right after the restructuring and until the end of the electricity crisis. Approximately 23,000 MW of new capacity was approved for construction since AB 1890. Of this total, about 10,000 MW is now on line, and about 5000 MW of new capacity is currently under construction.66 Another 7700 MW has been cancelled or is currently on hold (see Figure 10.11 for year by year data). However, since the crisis, the rate of new capacity applications has declined sharply. Figure 10.11 shows that new applications to construct generating plants declined to 1000 MW in 2002, 500 MW in 2003, and to 250 MW in 2004. With the low rate of new capacity additions, the retirement of some old thermal plants, and growth in electricity usage, markets promise to become progressively tighter over the next few years. The need for additional capacity and/or additional transmission capacity is a particular problem in rapidly growing southern California. At the same time, however, IPPs, such as Calpine Corporation, face severe financial difficulties and are unlikely to invest in new generation capacity absent long-term contracts with retail utilities or well-functioning capacity markets. Utilities, uncertain about their regulatory future, have been hesitant to enter new long-term power purchase agreements. 10.4.1.2. IOU Status As of this writing, the immediate consequences of the financial crisis for the IOUs have been resolved. Pacific Gas and Electric has emerged from bankruptcy. All the IOUs are creditworthy. The state is no longer entering new electricity purchase contracts on behalf of 66

These figures represent about 20% and 10% of California peak demands, respectively.


371

the utilities; utilities are now buying their own electricity. Utilities are managing the longterm electricity purchase contracts negotiated by the state of California. 10.4.1.3. Retail prices of electricity The financial losses during the crisis must be borne by some entities, either generators, utilities, the state, or electricity consumers. The great majority of these costs are being borne by electricity consumers, in the form of increased retail rates or surcharges on the retail rates. The high costs of the DWR-negotiated long-term electricity supply contracts are being paid entirely by consumers. A large portion of the financial losses of the utilities during the crisis is now being recovered in the retail prices, under either court order or CPUC negotiation. The financial losses incurred by the state in buying electricity during the crisis have been financed through over $12 billion in revenue bonds. Debt service of these bonds is being paid by the electricity consumers. The retail price increases were heavily weighted against industrial/commercial users of electricity and against residential customers who for whatever reason were large users of electricity. As currently implemented, the rates (including surcharges) are highly tiered, with households that use low amounts of energy paying prices about equal the cost of delivery services plus acquisition cost of electricity and those using larger amount of electricity paying about three times that marginal price (depending on the particular geographic zone, about $0.24 per KWh). During the last year, the well-head price of natural gas has risen sharply, partly as a result of hurricanes Katrina and Rita. At the time of this writing, the price of natural gas at Henry Hub has increased to over $14/mcf, in comparison to $6.50 as of September 2004. Although natural gas prices can be expected to decline after the 2005–2006 heating season, absent additional damages to the natural gas production facilities, these natural gas prices are expected to remain considerably higher than they have been historically. Such high natural gas prices can be expected to further increase the wholesale prices of electricity and therefore the retail prices. 10.4.1.4. Litigation Some of the historical costs of the crisis may not be borne by consumers, but may be recovered through litigations of various types. Prominently, FERC continues its proceedings to order refunds to California for times when generators and marketers charged electricity prices that were not justified under FERC rules. Many such refunds have already been ordered. Private litigations attempt to reallocate costs from electricity users, back to generators and marketers. Some firms that signed long-term power purchase contracts during the time of the crisis have sued or negotiated for modifications to the price terms in these contracts. In addition, the US Department of Justice has brought criminal indictments against one company and its employees. The legal processes continue. 10.4.2. Changing market structure, market forces The dual crises – the electricity crisis and the financial crisis – motivated many “reforms of the reforms.” Some changes occurred during the crisis, other have been being implemented in the years since the crisis was over. 10.4.2.1. Fundamental market flaws The most fundamental market flaws – the reliance on spot markets and the retail price controls – have been repaired.

372


The inappropriate over-reliance on spot markets has been eliminated. Utilities are no longer relying on spot market purchases for other than small fractions of electricity purchases. Almost all of their electricity is acquired through a combination of self-generation (dominantly, but not entirely, from generating units in existence before the crisis), from the long-term contracts negotiated by the DWR, and from QF contracts from before the crisis. A few new shorter-term contracts have been negotiated by the IOUs. Very little other electricity is being transacted on the organized spot markets. The PX is defunct; no electricity has been sold on that market since 2001. Real-time purchases on the CAISO are sharply reduced from during the crises, since IOUs are no longer systematically under-scheduling on those markets. The FERC has instituted penalties for under-scheduling into the CAISO and the IOUs, with most of their electricity purchased on term markets have no incentive to try exercising market power through under-scheduling. The CPUC has ordered resource adequacy rules for IOUs, including minimum planning reserve margins of 15%. At least 90% of the necessary resources must be acquired at least 1 year in advance.67 This represents a fundamental shift in the regulations, a shift motivated by the experiences of the crisis. Retail price controls no longer are operational; normal ratemaking procedures now govern retail pricing. Of course, normal procedures do not allow wholesale market price fluctuations to be quickly incorporated into retail prices. Thus, short-run demand functions for electricity still are quite unresponsive to wholesale price changes. However, longer-term trends in cost changes can be incorporated into retail rates. This possibility allows demand functions to adjust over the course of a few months to changes in wholesale prices and it helps assure that IOUs will not face a similar financial crisis if wholesale prices do systematically increase.68 10.4.2.2. Retail competition on hold During the crises, the California legislature recognized that the long-term power purchase contracts, then being signed by the DWR, would obligate the state to acquire those contractual quantities of electricity. They further recognized that after the crises were over, the contracts would become obligations of the utilities to purchase the bulk power. In order to protect the state and the utilities from risks that the contractual power would either be a greater quantity than needed by the utilities or at an uncompetitively high price, the legislature passed a law suspending the right for electricity users to enter new direct-access contracts. This suspension of all retail competition would last as long as the state contracts were still operational, until roughly 2015. As originally envisioned, this law would preclude companies and individuals from entering contracts to purchase electricity, except from their local utilities or from municipalities. Subsequently, the CPUC established some limited opportunity for firms to enter direct-access contracts, bypassing the utilities, if those firms paid an “exit fee”, designed to cover that firm’s share of the historical costs. This law and the CPUC implementing rules assure that the costs of the DWR long-term contracts would be recovered from ratepayers.

67

However some of the contracts seem to not be based on identifiable generating assets, but rather liquidated damage contracts. Liquidated damages will protect the utility but will not necessarily assure adequate supplies of electricity. Regulatory rule making in this area is continuing. 68 At the time of this writing, natural gas prices at Henry Hub have increased to $14/mcf, up from $6.50 1 year ago. This will increase the electricity generation costs and therefore can be expected to increase retail rates.


373

The one exception where electric retail competition is now allowed is through community choice aggregation,69 a process under which any California city or county can “aggregate” electric loads of its residents, businesses and municipal facilities. The city or county would acquire bulk power, would sell electricity to customers within the city or county, in competition with the incumbent distribution utility for these sales. The utility would continue to provide distribution services. Governor Schwarzenegger has supported the reopening of direct access to acquire electricity, through a “core/non-core” structure in which large users of electricity would be able to enter direct-access contracts, but smaller “core” customers would rely on the utility to supply electricity.70 The provisions and timing of such a regulatory change are still under debate. In the meanwhile, there now is very little effective retail competition. 10.4.2.3. Electricity demand measures The CPUC, the CEC, and utilities are developing and implementing systems to improve electricity consumption response to changing conditions. Demand-management contracts have been entered with large commercial and industrial users of electricity. Some contractually require those users to curtail use of electricity when conditions are tight; others give the utility the ability to remotely curtail use. Additional demand-response mechanisms are under debate or experimentation. However, much more research and policy analysis is needed to increase the short-run and the longer-run demand response to wholesale prices. This is particularly important because the current market structure so severely blunts the connection between hourly wholesale prices and the retail prices that most consumers face. A second approach would be to improve the connection between hourly wholesale prices and the retail prices, possibly through real-time pricing, or some variant of real-time pricing, such as critical peak pricing. Experiments are underway to test real-time pricing. Interval meters have been installed in facilities of the largest industrial users, allowing the possibility of real-time pricing. Tariffs that include variants of real-time pricing for electricity have been developed and tested, but have not been broadly implemented. PG&E and SDG&E have proposed to deployed advanced meters to all customers over the coming years. The significance of demand response can be seen in Figure 10.19, developed by the CEC, which estimates peak demands for electricity in California for the various usage categories. Almost 30% of the peak demand is for residential or commercial air conditioning. This use, in principle, can be shifted significantly over time within the course of a day. However, absent incentives or controls to do so, there is no reason to expect that a large fraction of this load will shift away from the times of the greatest demands on the system. With appropriate incentives, and with appropriate information technology, such uses can be shifted in time so as to optimize the overall system, reducing the costs at the highest peak loads, and reducing the need for new peaker power plants. California is continuing and expanding its emphasis on longer-term energy efficiency measures. Energy efficiency programs organized through the IOUs have been expanded through a CPUC rulemaking process. The state of California continues its “Flex Your Power” publicity campaign, providing information about and encouragement for energy efficiency. The Governor has ordered all state-owned building to be managed so as to meet minimum energy efficiency standards. New appliance efficiency standards have been established; 69 70

Allowed under AB 117. Such a core/non-core structure was one of the options suggested in the “Yellow Book”.

374


15%

14%

7000 11%

6000

11%

5000 8%

4000

7%

3000 4%

2000

4%

4%

4%

1000

us try er ci al R lig es ht id en tia C lm om is m c er c O i a th lm er is co c Ag m m & bl w ds at . er pu m Pr pi oc ng es s in R du es st re ry fri ge ra to rs

C

om

m

in d

ia l

bl y

nt As se m

es id e

R

C

om

m

er ci al

AC

AC

0

Fig. 10.19. California peak demands by usage category. Source: CEC.

others are under consideration. Energy efficiency is incorporated into California building standards and additional rulemaking can be expected. These actions collectively can be expected to continue the trend of per-capita electricity use in California not growing significantly, in contrast to the rest of the US. 10.4.2.4. IOU wholesale electricity procurement CPUC has recently established rules that now require IOUs to acquire new electricity supplies through a competitive process, intended to level the playing field across different potential suppliers of electricity, in particular between IOU self-supply of electricity and IPP generation. Such a competitive procurement process increases the probability that the utility portfolio will be based on minimum cost, for the given risk, reliability, and environmental performance. But such rules do not eliminate the incentive71 for utilities to purchase from themselves in preference to IPPs; the rules simply make it more difficult for the utilities to act in accordance with that incentive. Moreover, the rules can be applied to allow the utilities to exercise market power over existing generators that do not yet have long-term contracts to sell their power.72 However, the alternatives – simply ban the utilities from any 71

IOUs are not compensated for the credit risk they take on when they enter long-term contracts, although they are compensated for risks when they build their own plants. This provides an incentive to avoid long-term contracts in favor of self-generation. Most of the IPPs are in difficult financial situation, so that there is risk of long-term contracts with these entities. Regulatory uncertainty makes longterm commitments risky, but whether long-term contracts with IPPs are more risky than construction of utility-owned generation is unclear. 72 For example, the ongoing PG&E procurement process excludes already completed IPP plants, such as the Calpine Metcalf plant, from bidding to supply electricity. This restriction is being negotiated at the time of the writing. But it does illustrate the ability of the IOU to adopt procurement rules that do not truly level the playing field, since PG&E does not exclude itself from acquiring electricity from its existing generation assets.


375

self-generation or any new self-generation – would create important dysfunctional incentives and would preclude utility self-generation when such an option was truly the best for consumers. In addition, competition among wholesale buyers of electricity is currently very limited in California. Since California no longer has any meaningful retail competition in electricity markets, the IOUs are virtually the only wholesale electricity buyers in California. This market domination allows the IOUs to exercise market power over the IPPs. IPPs, understanding the dominance of the IOUs as buyers, are unlikely to start construction of new power plants, absent long-term contracts with the IOUs. Moreover, IOUs seem unwilling to enter long-term contracts with IPPs to purchase electricity. Thus while, in principle, wholesale procurement rules promote a level playing field, whether the field will turn out in practice to be truly level is not clear.

10.4.2.5. ISO market redesign The CAISO is developing a new market design in order to correct flaws in California wholesale electricity markets. The current design includes nodal pricing for electricity, rather than the original zonal pricing, and firm marketable transmission rights. Market software to support the new market design is under development. However, the process is very slow and costly. Originally entitled MD 02, for “Market Design 2002”, the process has been renamed, reflecting the reality of the process speed.

10.4.2.6. Capacity incentives It is generally realized that absent payment for providing unused generation capacity, electricity suppliers are unlikely to have economic incentives to invest in enough new capacity to assure that there will always be an adequate reserve margin of unused generation capacity. In order for generators to provide sufficient capacity to assure that they can cover very uncertain peak loads, they need incentives to do so. One option is for generators to enter long-term contracts to sell a bundled combination of electricity and generation capacity, with the generation capacity some agreed-upon fraction greater than the expected electricity generation. Such a bundled combination would be more costly than an electricity-only contract. Absent regulatory rules that require utilities to enter such bundled combinations, it is unlikely that such combinations could compete with electricity-only contracts. A second option is based solely on payments for electricity. Electric generation plants require large capital investments and incur capital costs even when not being used. In periods of adequate electricity supply, the short-term prices for bulk power tend to be too low to repay the capital costs. In periods of short supply, in theory, prices might increase enough that these limited times could allow enough profits to cover the long-term capital costs. In practice, however, as seen during the electricity crisis, wholesale price mitigation can be expected. Such price mitigation measures virtually assure that generators could not repay the long-term capital costs through such short-time periods of short supply. In addition, generators can only wait a limited time to recover capital costs. The need to obtain financing through normal financial markets implies that waiting too long to recover capital costs may be a route to bankruptcy. In response to this recognized problem, the state is examining how to create an electricity generation capacity market that would allow utilities and generators flexibility in meeting

376


resource adequacy requirements. The market could include tradable capacity rights or obligations. However, design is far from complete. Thus it is not clear when and whether such a market will be created. 10.4.3. Other electricity-related policies 10.4.3.1. Environment Environmental protection has been an important component of electricity policy, both before and since the crises. However, since the crises, issues of environment have become more explicit. Broad classes of electric generation/use technologies viewed as environmentally attractive are given priority over those that have greater adverse impacts on the local environment or on the global environment, through a “loading order” among classes of technologies. The highest priority is energy efficiency – energy use technologies and practices that reduce the overall cost of providing energy services, without reducing the available services. Besides the environmental advantages, energy efficiency can reduce overall costs of providing energy services. That advantage is particularly important in times of high electricity costs. Second priority is demand response and conservation measures, as discussed above. Third highest is renewable energy for electricity generation. The renewable portfolio standard in current law requires each IOU to increase its use of renewable sources of electricity to 20% by 2019. Governor Schwarzenegger is supporting acceleration of this standard to 20% by 2010; the CPUC has adopted a policy supporting that acceleration. The California administration is examining whether a 35% standard would be viable at some future year. The CEC is funding research and development to improve use of renewable technologies in California. In addition, policies initiatives, such as the “Million Solar Roofs” initiative, a subsidy program for installation of distributed photovoltaics, are under debate.73 Finally, Governor Schwarzenegger has established greenhouse gas emissions goals for the state and has directed members of his administration to develop strategies to meet those goals. Although many options are being examined, it is likely that some of the strategies will significantly influence electricity generation and use technologies. 10.4.3.2. Transmission During the electricity crisis, transmission capacity significantly limited movement of electricity between northern and southern California. Over a longer term, although existing transmission lines allow electricity movement between California and neighboring states, there is very little transmission capabilities between California and the more distant states. Since the crises, the most important transmission bottleneck in California, Path 15 linking northern and southern California has been upgraded to increase its capacity. Many minor upgrades have been completed or are underway within California. Four western governors74 have proposed a transmission line, the Frontier Line, with a 6000 MW capacity, from Wyoming to California. This would enable large increases of electricity from the Intermountain west to be developed and transmitted across the region, 73

The “Million Solar Roofs” initiative was submitted to the legislature in 2005. It did not pass during that legislative session; further legislative development can be expected during the next session. 74 Wyoming Gov. Dave Freudenthal, Utah Gov. Jon Huntsman Jr., Nevada Gov. Kenny Guinn and California Gov. Arnold Schwarzenegger. The cost is estimated to be $3.3 billion.


377

including to California. Advocates of the program stress the possibility of importing more electricity generated using renewables and clean coal technologies. Yet there is the possibility of importing electricity generated from high carbon dioxide emitting coal plants. Work continues to identify alternative routes, and to examine regulatory, financial, environmental, economic, and engineering issues. Prominent among the regulatory rules will be possible electricity acquisition policies to assure that the electricity imported into California is dominantly based on renewables and clean coal technologies. If the Frontier Line is completed, it would reduce the current risky dependence on natural gas for electricity generation. 10.4.3.3. Natural gas Natural gas limitations were fundamental to the electricity crisis, as discussed above. Since the crisis, natural gas storage facilities have been expanded.75 Many proposals have been made and are under state consideration for siting liquefied natural gas (LNG) facilities in California. There remains an important jurisdictional dispute between the federal government and the state of California as to whether review and approval authority vests with the federal government or state governmental agencies. Although no particular site has been approved, approval of some site is highly likely. Improvements of such natural gas infrastructure could reduce the risk of electricity problems stemming from natural gas restrictions or high natural gas prices. It is likely to be several years until the gas can be delivered from the LNG facilities currently being considered in California, although natural gas will probably be delivered sooner from the Sempra facility being planned in Mexico. 10.4.4. Organizational issues During the crisis, the state legislation created a state power authority,76 with broad authority to construct new electric generating facilities and to acquire existing facilities by use of eminent domain procedures, making the state as a major electricity generator. It initially signed letters of intent to finance over 2000 MW of renewable energy and over 3000 MW of natural-gas-fired peaker units, but all these plans have been cancelled. The authority has been dismantled. Governor Schwarzenegger has proposed a consolidation of the state energy agencies in order to improve the capability of the state to deal with its energy problems. The reorganization plan would create a California Department of Energy and would integrate policy, planning, conservation/efficiency programs, and R&D functions into that agency. Such a plan would make energy more central to decision-making within the Governor’s office and could be expected to improve governmental reactions to further energy problems. The CPUC would remain as then primary regulatory and rate setting agency. However, the legislature has so far rejected that plan twice, once as a reorganization plan and once as a bill. Whether such a reorganization will be approved by the legislature in subsequent sessions remains uncertain. 10.4.5. Proposals to re-regulate The discussion above is based primarily on the policy and regulatory actions being taken by the California administration and legislature, actions generally consistent with “reforming the reforms,” by fixing the problems that exist in the California markets, generally moving in 75 76

The Wild Goose Storage expansion was complete in 2003. Formally, the California Consumer Power and Conservation Financing Authority.

378


a direction of competitive markets where and when possible. However, some California groups do not share the vision underlying these actions but would return to the old system of vertically integrated rate-of-return-regulated utilities. In 2004, the state legislature passed a bill, AB 2006, authored by speaker of the Assembly, Fabian Nunez, that would have moved toward reestablishing vertically integrated utilities. It would have assured that the utility could recover any costs it incurred in building its own generation units and would encourage utilities to build and operate their own generators rather than entering term contracts with IPPs. It would have made rules for direct access so onerous as to preclude any new direct-access contracts, thus permanently virtually eliminating retail competition. The Nunez bill enjoyed strong support, and considerable lobbying, by SCE, one of the two largest utilities. The Governor vetoed that bill. At the time this chapter was drafted, an initiative, Proposition 80, misleadingly77 entitled by its sponsors “the Repeal of Deregulation and Blackout Protection Act”, was scheduled for California vote in November 2005. Sponsored by The Utility Reform Network (TURN), that initiative would have repealed many of the valuable provisions of AB 1890; it would have permanently prohibited new direct-access contracts, except community choice aggregation. It would have subjected all electricity suppliers to CPUC regulation. It would have made real-time pricing – or any electricity pricing that varies with time of day, day of week, or month of the year – very difficult or impossible to implement for residential consumers.78 It would have mandated a renewable portfolio standard of 20% by 2010, but would have not allowed additional increases of that standard. It would have required a 2/3 vote of both houses of the legislature to modify any provisions of the initiative,79 and would have thereby sharply reduced the flexibility to pursue “reforms of the reforms.” This proposition was defeated in November by 66% of the vote, the most negative fraction of all propositions voted on during that election. However, the qualification of Proposition 80 as an initiative provides evidence that there is a strongly felt opposition to market-based reforms within California’s restructuring. 10.4.6. Public perceptions The crises have left a legacy of public perceptions, many of them quite distorted or, at best, incomplete. The electricity system, its operations, and the economics of electricity are hard for even experts to understand fully. For non-specialists these issues are even more difficult and thus non-specialists tend to rely on information from news media. During the energy crisis, political leaders released and the press reported a tremendous amount of politically inspired miscommunication about the nature and causes of the price increases, the financial problems, and the blackouts.80 Official press releases villainized electricity generators and 77

The act would do nothing to prevent blackouts other than confirm actions that are already being undertaken. 78 Although real-time pricing is not now economically attractive for residential customers, with changing technology real-time pricing may well become economically attractive for all customers. 79 In fact it would allow the legislature to “amend this act only to achieve its purposes and intent”. This language assures that virtually no further reforms would be possible without prolonged litigation. 80 For example, during the crisis, Governor Davis’ public communications, carried on television, radio, and newspapers, assign all virtually blame for California’s electricity problems on electricity generators and marketers. Words and phrases such as “profiteering”, “plunder”, “unconscionable”, “price gouging”, “exorbitant profits”, “market marauders”, “pirate generators”, “privateers”, “obscene profits”, and “outrageous wholesale prices” were essential parts of his formal statements.


379

traders, particularly “out-of-state generators”. The communication distortions have now been muted, but they continue. Now that the crises are over, the impressions of many people are based on what they learned during the crisis. Unfortunately, the more distorted are the public perceptions, the more difficult to implement good public policy. An example of politically motivated distortions appears in the 2005 California voters’ pamphlet arguments in favor Proposition 80. Proposition 80 advocates assert: “California’s failed experiment in electric deregulation cost our people and businesses billions of dollars,” implying quite disingenuously that these costs occurred because “Enron and other traders held California hostage ... driving up wholesale prices 1000%.”81 As described above, these advocates promote a ballot initiative which, if passed, would permanently eliminate retail competition, restrict time varying pricing, and mandate particular levels of renewable resources, policy changes quite unrelated to Enron, energy traders, or any of the causes of the electricity and financial crises. Its qualification as an initiative provides evidence that politically inspired distortions of public perceptions remain alive and well in California.82

10.5. Some Reflections Unfortunately, one message repeatedly communicated is that the California experience proves that electricity deregulation has been an utter failure in California, and by extension, is likely to be a failure elsewhere. Yet a fair assessment of the California experience cannot reach such a conclusion. In California, restructuring has set the stage for widespread wholesale market competition and adequate electric generation capacity. California’s restructuring has resulted in increases in new electric generating plants proposed, approved, and under construction. An oft-repeated contention is that deregulation caused the electricity crisis. However, the western electricity crisis was primarily the product of a “perfect storm”, a combination of simultaneous adverse conditions, of flawed market rules, and only secondarily of exercise of market power and market gaming. The financial crisis was the direct result of California

81

The initiative justification in the Voter’s pamphlet begins as follows: “5 years ago, California was devastated by an electricity crisis. “Enron and other energy traders held California hostage, extorting tens of billions of dollars from us. They manipulated the electricity market, driving up wholesale prices 1000%. California faced rolling blackout and untold economic damage. “Audiotapes released by the US Justice Department revealed Enron energy traders boasting of ‘making buckets of money’ by creating power shortages. One trader laughed about ‘all the money you guys stole from those poor grandmothers in California,’ while another ordered a power plant worker to ‘just go ahead and shut her down.” “California’s failed experiment in electric deregulation cost our people and businesses billions of dollars.” “We learned many lessons from that disaster. The state has taken some positive steps to clean up the mess – but not nearly enough. Amazingly, legislation to require sufficient supplies of electricity (AB 2006, the Nunez bill) was vetoed by the Governor last year.” “That’s why Proposition 80 – the Repeal of Deregulation and Blackout Prevention Act – is on the ballot.” The statement was signed by Robert Finkelstein, Executive Director of TURN; Richard Holober, Executive Director, Consumer Federation of California; Nan Brasmer, President, California Alliance of Retired Americans. 82 However, many newspaper editors have been able to see through these distortions and have written editorials urging votes to reject Proposition 80.

380


regulatory actions. However, the financial crisis was not the result of deregulation, but rather of inappropriate regulation. Even though all municipal utilities and IOUs throughout the entire West faced the electricity crisis, only the IOUs located in California, facing the CPUC regulatory rules, experienced the financial crisis. The California experience illustrates that actions economically isolating the supply side of markets from the demand side create problems. Particularly retail price control isolated consumer from the changing economics of the supply side of electricity markets. Retail price control discouraged energy demand reductions – energy conservation – even though energy conservation in the short run and energy efficiency in the longer run was California’s best hope for forcing decreases in wholesale prices. The California experience also brought into sharp focus the importance of managing risks associated with implementation of any policies, especially policies that radically change the system. California electricity restructuring radically changed regulations, altering an electricity system that itself has never been free of economic risk. Risks exist and they should be managed, distributing risks appropriately. One issue of managing risks involves the timing of fundamental market transformations. The tighter are the natural gas supply constraints and the electricity capacity constraints, the greater the opportunity for market gaming and exercise of market power. Fundamental transitions are less likely to lead to problems if there is enough available energy and generation capacity to limit any attempts to exercise market power. Appropriate risk management implies that such timing issues should be taken into account. The California experience highlights something well known to everyone who has managed, participated in, or observed large-scale changes in complicated organizations or economic structures. Any major restructuring of such important systems will continue to require modifications – and possibly major modifications – well after the initial changes. System operation requires monitoring and may require leadership to identify and implement changes that are needed after unintended adverse consequences of the system change become apparent. California’s electricity system includes both complicated organizations and complex economic structures. The restructuring was fundamental and sweeping. No one was able to predict with any certainty how all the changes would work. Nor could anyone rule out unintended adverse consequences of the restructuring. System flaws requiring changes could have been expected. In fact, an expectation that such a radical and complex restructuring would go smoothly – particularly given the changes in regulatory rules facing the IOUs, the creation of the CAISO and the PX, the untested incentives on the newly divested generators, and a “perfect storm” of external conditions – was entirely unrealistic. Thus, it was important that state carefully monitored the electricity system operation and it was important that leadership was available to respond to the problems once identified. The California experience shows the real-lasting harm that can befall a state when its leaders fail to take the appropriate leadership roles in the face of the unintended consequences of profound system change.

References Borenstein, S. (2000). The trouble with electricity markets: understanding California’s restructuring disaster. Journal of Economic Perspectives, 16(1): 191–211. Borenstein, S., Bushnell, J. and Wolak, F. (December, 2002). Measuring market inefficiencies in California’s restructured wholesale electricity market. American Economic Review, 92(5), 1376–1405. California Public Utilities Commission (1995). Decision D.95-05-045. California Public Utilities Commission. California’s electric services industry: perspectives on the past, strategies for the future. The Yellow Book, February 3, 1993, California Public Utilities Commission.


381

California Public Utilities Commission. Order instituting rulemaking on the commission’s proposed policies governing restructuring California’s electric services industry and reforming regulation. The Blue Book, Decision 95-12-063 (December 20, 1995) as modified by D.96-01-009 (January 10, 1996). California Public Utilities Commission. Roadmap decision. Decision D.96-03-022. 1966. Carlton, D. and Perloff, J. (1994). Modern Industrial Organization. New York: Harper Collins, pp. 139, 922. Cavanagh, R. (June 2001). Revisiting ‘the genius of the marketplace’: cures for the Western electricity and natural gas crises. Electricity Journal, 14(5). Crow, R.T. (December 2001). “Not invented here: what California can learn from elsewhere about restructuring electricity supply. Stanford Institute for Economic Policy Research Working Paper. Energy Information Administration (2001). Electricity Shortage in California: Issues for Petroleum and Natural Gas Supply. Washington, DC. website: http://www.eia.doe.gov/emeu/steo/pub/special/ california/june 01 article/caoutside.html Energy Information Administration (2001). Natural Gas Monthly. Washington, DC. website: http:// www.eia.doe.gov/emeu/steo/pub/special/california/june 01 article/caoutside.html Federal Energy Regulatory Commission. Order Directing Remedies For California Wholesale Electric Markets, Issued December 15, 2000. Federal Energy Regulatory Commission. Report On Plant Outages In The State Of California Prepared by Office of the General Counsel, Market Oversight and Enforcement, Office of Markets, Tariffs and Rates Division of Energy Markets, February 1, 2001. Federal Energy Regulatory Commission. Order on rehearing of monitoring and mitigation plan for the California wholesale electric markets, establishing west-wide mitigation, and establishing settlement conference, issued June 19, 2001. Fisher, J.V. and Duane, T.P. Trends in electricity consumption, peak demand, and generating capacity in California and the Western Grid, 1977–2000. Working Paper from the University of California Energy Institute Program on Workable Energy Regulation (POWER), September, 2001. Harvey, S. and Hogan, W. (April, 2001). On the exercise of market power through strategic withholding in California. http://ksghome.harvard.edu/~whogan/mkt_Pwr_CA_HH_042401.pdf Hogan, W.W. Electricity market restructuring: reforms of reforms, May 25, 2001, Harvard University. James L. Sweeney, (2002). The California Electricity Crisis. Hoover Institution Press; Stanford Institute for Economic Policy Research. Stanford, California. Joskow, P. and Kahn, E. (2002). A quantitative analysis of pricing behavior in California’s wholesale electricity market during summer 2000. The Energy Journal, 23(4), 1–35. South Coast Air Quality Management District. White paper on stabilization of NOx RTC prices. Web site: http://www.aqmd.gov/hb/010123a.html US General Accounting Office. Energy Markets: Results of Studies Assessing High Electricity Prices in California, June 2001. Western Governors’ Association. Conceptual Plans for Electricity Transmission in the West, August 2001.


Chapter 11 Texas: The Most Robust Competitive Market in North America1 PARVIZ ADIB, AND JAY ZARNIKAU2 1 Wholesale Market Oversight, Public Utility Commission of Texas, TX, USA; 2Frontier Associates and The University of Texas at Austin, TX, USA

The restructuring of Texas’ Electric Reliability Council of Texas (ERCOT) market benefited from a confluence of positive factors, including a phased approach where the restructuring of the wholesale market preceded the retail choice, ample generating capacity at the outset of retail competition, and an intrastate market where a single state-level regulatory authority wielded near-exclusive jurisdiction over the implementation of the state’s plan. However, many of the activities required by retail restructuring turned out to be much more difficult than anticipated. Further challenges in the coming years are anticipated as Texas seeks to implement more efficient means of coordinating the operation of generating facilities, managing transmission congestion, and ensuring the adequacy of the resources needed to meet the state’s growing demand.2 11.1. Introduction Has Texas succeeded where so many other markets have failed? That is, will the Texas electricity market provide robust competition that will provide energy consumers with the options and savings offered by a workably competitive market? Will the market provide profit opportunities to adequately reward innovative and efficient suppliers of generation and retail services? The progress achieved in the past 4 years indicates that the prospect for a robust competitive electricity market is bright, but some challenges remain. Indeed, the competitive electricity market within the ERCOT3 has avoided the market meltdown suffered in California. New retail providers have been attracted to the ERCOT 1

The opinions expressed in this chapter are those of the authors and do not represent the opinion of the PUCT or its Staff. 2 The authors would like to appreciate comments on earlier drafts provided by Drs. Shmuel Oren and Perry Sioshansi and Mr. Jess Totten, Director of Electric Industry Oversight Division, PUCT. Particular thanks go to Mrs. Lee Zarnikau for editorial advice. 3 ERCOT accounts for about 85% of Texas electric industry and has been open for retail competition since January 1, 2002. The other 15% of the State includes territories served by El Paso Electric in the West, Xcel Energy-Southwest Public Service Company in the Texas Panhandle, AEP-Southwestern Electric Power in the Northeast, and Entergy in the Southeast portions of Texas.

383

384


market, customer-switching rates are high relative to other restructured markets, and reliability has been preserved. The problems experienced in other markets appear to have been avoided through a better market design (e.g., greater reliance on bilateral contracts to hedge wholesale price risks and regulated price-to-beat (PTB) rates for small customers that can be adjusted for fuel market changes), a better market climate (e.g., higher generation reserve margins and a stronger transmission system), and many other factors (e.g., greater scrutiny following California’s problems, an absence of federal regulatory jurisdiction, and a different set of market participants). Based on this success, Texas is regarded as the most robust restructured retail market in North America and one of the top three in the world.4 Yet, the transition has been neither flawless nor painless. Many of ERCOT’s earlier difficulties resulted when it assumed the role of Central Registration Agent without adequate resources, information systems, and expertise to process customer switch requests, assign account numbers, initiate service in new buildings, and funnel billing data from the Transmission and Distribution Service Providers (TDSPs) to the appropriate retail organizations.5 Other challenges have included finding adequate resources to perform market monitoring, educating consumers about their choices, managing transmission congestion in an efficient manner to minimize subsidies and send appropriate price signals, and ensuring long-term resource adequacy. Even after the market opened to wholesale competition and retail customer choice, market design activities continued. Some major changes were implemented in ERCOT after the market opened, including changes to balanced schedule requirements, major revisions to the provider of last resort (POLR) rules, and changes to congestion management. Further sweeping changes to congestion management and a fundamental change to market design is expected in the future to further enhance system operation and achieve further market efficiency.6 In this chapter, we report on a number of successes, challenges, and lessons learned at this point in the evolution of the ERCOT electricity market.

11.2. Background on the Unique ERCOT Market A confluence of factors is responsible for ERCOT electricity market’s early success, including the presence of a single regulatory agency, ample generating capacity reserves at the outset of competition, a reasonable initial market structure, and carefully crafted transition mechanisms. 11.2.1. Relatively little federal jurisdiction In contrast to other states, most of Texas is not under the jurisdiction of the Federal Energy Regulatory Commission (FERC). This is mainly because ERCOT is not electrically synchronized with either Eastern Interconnection or with the Western Interconnection transmission networks. While ERCOT has two Direct Current (DC) ties with the Eastern Interconnection,

4

Center for Advancement of Energy Markets, Retail Energy Deregulation Index 2003, 4th Edition. ERCOT is the only known competitive electricity market that acts as Central Registration Agent for retail customer registration and switches. This responsibility goes beyond ERCOT market and covers all customers within Texas. 6 See Adib, P. (2003). 5

The Most Robust Competitive Market in North America

385

Fig. 11.1. NERC regions. Source: FERC web site.

such ties are not considered to result in interstate commerce. Hence, the Public Utility Commission of Texas (PUCT or Commission) solely regulates the electricity market within ERCOT. This has resulted in greater coherence in the retail and wholesale markets, more regulatory certainty, and closer coordination between Legislative authorities and regulators involved in restructuring the electric industry within ERCOT. Texas is covered by four Electric Reliability Councils: ERCOT, Southwest Power Pool (SPP), Southeast Electric Reliability Council (SERC), and Western System Coordinating Council (WSCC). Figure 11.1 shows various Reliability Councils within the North American Electric Reliability Council (NERC). 11.2.2. Size Texas leads the US in electricity consumption and electrical services are provided to about 8 million industrial, commercial, and residential customers (meters). Its peak demand is more than 70,000 MW served by about 90,000 MW of installed capacity. The annual value of the energy sold through the state’s wholesale electricity market is more than $22 billion, accounting for about 10% of US total sales of electricity at the wholesale level. About 85% of the electricity needs in the nation’s leading state in electricity generation and usage has traditionally been satisfied by utilities that were members of ERCOT. Figure 11.2 identifies the areas of Texas that are served through various reliability councils. Traditionally, ERCOT was dominated by a few large vertically integrated investor-owned utilities (IOUs), including Texas Utilities Electric Company (now TXU), Houston Lighting and Power Company (now split into Reliant Energy and CenterPoint Energy7), Central and Southwest Corporation8 (which later merged with American Electric Power Company), and Texas-New Mexico Power (TNMP) Company. Additionally, roughly 60 electric cooperatives operate in ERCOT. Nearly 50 municipal utilities systems also distribute (and in some cases, generate) electricity, including CPS Energy (formerly known as City Public Service 7

CenterPoint Energy’s generation was spun off to a new company, Texas Genco II, in late 2004. Two of Central and Southwest Corporation affiliates, Central Power and Light (CPL) and West Texas Utilities (WTU) operate in ERCOT. 8

386

Electricity Market Reform Texas electric area map SPP SPS LPL

WECC EPEC

ERCOT AEP-Central AEP-North BEPC COA CPS CPE TMPA TNMP TXU-Oncor

SPP SWEPCO

SERC EGS

Fig. 11.2. Areas of Texas served through various reliability councils. Source: PUCT.

Board of San Antonio and considered the second largest municipal system in the US) and Austin Energy. Prior to restructuring, generation dispatch decisions and other operational decisions were made locally in 10 control areas. However, ERCOT has operated as a single control area since July 31, 2001. The number of market participants has been growing since the introduction of wholesale competition in 1995 and retail competition in 2002. Currently, there are more than 550 market participants acting as Qualified Scheduling Entities (QSEs), Retail Electric Providers (REPs), Power Generation Companies (PGCs) and Independent Power Producers (IPPs), Power Marketers (PMs), TDSPs, and Aggregators.9 Table 11.1 lists major players in ERCOT market. 11.2.3. Generation mix Generation in Texas has traditionally been dominated by natural gas. The construction of nuclear, coal, and lignite in the late 1970s to mid-1990s reduced the share of natural gas capacity in Texas to 61% by 1995. Similarly, energy generation from natural gas declined to 9

The following link provides access to the lists of various players in ERCOT market: www.puc.state. tx.us/electric/business/index.cfm


387

Table 11.1. Major market players in ERCOT market. Power generation companies and IPPs American Electric Power Company American National Power Austin Energy Calpine Corporation Constellation Electric Coral Power CPS Energy Exelon FPL Energy Lower Colorado River Authority Sempra Power Tenaska power Texas Genco II, Limited Partnership Texas-New Mexico Power Company TXU Generation Company

Electric retail providers AEP Texas Commercial & Industrial Retail BP Energy Company Calpine Corporation Coral Constellation New Energy Direct Energy Entergy Solutions First Choice Power GEXA Energy Green Mountain Energy MPower Occidental Reliant Energy Sempra Energy Strategic Energy Suez Tenaska Power Services TXU Energy Services Utility Choice

Transmission and distribution service providers American Electric Power Company Austin Energy Brazos Electric Power Cooperative CenterPoint Energy CPS Energy Lower Colorado River Authority Texas-New Mexico Power Company TXU Electric Delivery Company

38% in the same time period.10 ERCOT experienced a similar decline in the share of natural gas capacity and energy generation during the same time period. Total gas-fired capacity and energy generation in ERCOT declined to 62% and 35%, respectively, by 1995. Since the introduction of wholesale competition in 1995, more than 33,000 MW of capacity has been added in Texas with the overwhelming majority located in ERCOT. Natural gas has been the fuel of choice while wind power has also increased significantly. Table 11.2 shows the 2004 capacity and energy generation by fuel type. There is about 8000 MW of cogeneration in Texas, a significant portion of which is solely for self-generation. The net contribution from this source to the grid, roughly 2000 MW, is included in Table 11.2.

11.3. Impetus for Restructuring and Legislative Actions A number of factors motivated the restructuring of the ERCOT electricity market, including a laissez faire political philosophy toward markets and the success of efforts to introduce competition at the wholesale level. In this section, we place the evolution and restructuring of the ERCOT market into a historical context and review some key policy considerations. Some key milestones are identified in Table 11.3.

10

These shares would be higher if about 8000 MW of cogeneration capacity, which is mainly used for self-generation, operated by independent power producers were included.

388

Electricity Market Reform Table 11.2. 2004 capacity and energy generation by fuel type within the ERCOT wholesale electricity market. Capacity (MW)

(%)

Energy (MWh)

(%)

Natural gas Coal Nuclear Wind Water Other Diesel

59,020 15,007 4737 1310 472 611 –

72.7 18.5 5.8 1.6 0.6 0.8 0.0

124,100,266 113,067,975 40,342,858 3,007,094 803,843 1,329,674 –

43.9 40.0 14.3 1.1 0.3 0.5 0.0

Total

81,157

Fuel types

100

282,651,710

100

Source: PUCT.

11.3.1. Wholesale competition Texas was the last state in the US to establish a statewide system of electric utility regulation. The Texas Public Utility Regulatory Act (PURA) was passed in 1975. At that time, the Texas electric industry included many integrated monopolies engaged in generation, transmission, and distribution of electricity. The Commission’s authority over the IOUs ranged from issuing certificates to build generation and transmission facilities to setting the rates charged to customers. Compared to other states, PURA and the regulatory rules of the Commission were considered by the investment community to foster a fair regulatory environment that encouraged adequate investment to meet the fast growing demand for electricity in Texas. Customers, particularly industrial customers, experienced reasonable electricity rates under the regulatory regime. Generally speaking, Texans enjoyed relatively low rates, but because of relatively high consumption levels, had above average electricity bills. In 1997, Texas ranked 27th highest in the US in electricity rates with an average residential rate of 7.82 cents/kWh, but with the 6th highest average annual residential bill of $1066, in part due to heavy air conditioning use. In the 1980s, the potential for competition in the generation of electricity in Texas increased. Fostered by 1978 federal Public Utility Regulatory Policy Act (PURPA) that guaranteed a market for the electricity generated by qualifying cogeneration facilities that was in excess of on-site needs, the share of electricity produced by non-utilities in Texas grew to over 10% by the early 1990s (Zarnikau and Reilly 1996). Competition in generation markets was further advanced by the Commission’s resource planning rules, which required utilities to solicit competitive bids for resources before pursuing new power plant construction projects. Senate Bill 373, enacted in 1995, required the Commission to establish rules to foster wholesale competition and create an Independent System Operator (ISO) to ensure nondiscriminatory transmission access. Additional objectives were to ensure an equitable interconnection process, facilitate generation capacity and transmission expansion, and provide customer protection.11 In the Summer of 1997, ERCOT became the first ISO to begin operation in the US. In order to eliminate barriers to entry, the Commission established generator-friendly interconnection 11

See Adib, P. and Clark, C. (1996).


389

Table 11.3. Timeline in the evolution of a competitive market in ERCOT. 1975

Passage of Texas PURA establishing the PUCT.

1978

The federal PURPA is enacted, facilitating and providing a pricing mechanism for utility purchases of power from cogeneration and small power production.

1983

Amendments to the Texas PURA to reflect the 1978 enactment of PURPA and introduction of the elements of integrated resource planning, such as a 10-year demand and resource forecast.

1992

The passage of the US Energy Policy Act.

1995

Senate Bill 373 amending the Texas PURA to introduce wholesale competition in September 1995.

February 1996

The Commission establishes the requirement for ERCOT to become an ISO and requires utilities to offer wholesale open-access transmission service.

Late 1990s

The Commission approved interconnection rule to facilitate merchant plant development.

May 1999

State Legislature passes Senate Bill 7 amending the Texas PURA to introduce retail competition on January 1, 2002.

1999

The Commission approved postage stamp rates for transmission services.

2000–2001

The Commission finalized its decision regarding functional unbundling plans for integrated utilities. In addition, several important rules were finalized by the Commission to enhance transition to competitive electricity market within ERCOT.

August 2000

The Commission established WMO to monitor market activities and detect market power abuses and other market manipulation.

June 4, 2001

The Commission finalized its decision with regard to the ERCOT Protocols that established market rules for the wholesale electricity market.

July 31, 2001

The operation of ERCOT single control area began and pilot retail program was introduced.

January 1, 2002

Customer choice began within ERCOT electricity market and PTB was established within each incumbent IOU service area became effective for residential and small commercial customers with peak load lower than 1 MW.

September 2002

RMO was established to monitor retail market and identify areas for improvements.

February 2003

Price spikes in wholesale market prompt re-examination of the use of balancing energy, wholesale price mitigations formulas, and credit requirements for REPs.

2004–2005

Regulatory proceedings to calculate stranded costs are completed.

Late 2004

Switching rates for commercial energy consumers exceeds thresholds and the PTB for commercial customers is terminated in many service areas.

September 2005

PUCT decides to transition market to a nodal structure and sets January 1, 2009 as the deadline for full implementation.

rules and allowed postage stamp transmission pricing in the ERCOT wholesale electricity market. Such policies created a level playing field among incumbent utilities and IPPs, and further encouraged merchant plant development that was essential to the success of wholesale competition. As a result, most of the major national PGCs have entered the ERCOT market since the passage of Senate Bill 373, including American Electric Power, Exelon

390


Corporation, Calpine, FPL Energy, Constellation, Mirant, American National Power, and Suez Energy. 11.3.2. Retail competition While electricity rates under regulation were below national averages in Texas, industrial energy consumers provided the initial impetus for restructuring and pursuit of retail competition as in a number of other states. The PUCT favored changes that would extend the successful experience of wholesale market restructuring to the retail level. Such initiatives gained momentum under the chairmanship of Pat Wood, who later became chairman of FERC. IOUs who were originally opposed to restructuring became advocates for retail competition by 1997. Legislative foresight was the key factor in well planned and executed legislative amendments to pave the road for electric industry restructuring and retail competition in Texas. Before retail competition was introduced a careful study of restructured markets in the US (e.g., California and Pennsylvania) and abroad was undertaken to identify potential shortcomings that should be avoided. In addition, retail competition was introduced several years after the introduction of wholesale competition, to provide adequate time for wholesale market to mature and extra capacity investment to be attracted to Texas. Any initiative to introduce competition into the retail sector would have been difficult if sufficient competition was not first achieved at the wholesale level. Fortunately, Texas has been the fastest growing state with regard to merchant plant developments. Since the opening of wholesale competition in 1995, there has been more than 33,000 MW of new capacity built in Texas. Forty-seven new power plants were added between 1995 and the introduction of retail competition on January 1, 2002, representing one-fourth of all power plants built in the nation during that period.12 Nearly all of these additions to generating capacity were constructed by entities other than incumbent IOUs, mostly using natural gas. In addition, more than 1300 MW of new wind capacity has been completed in Texas in the past 4 years and the new wind capacity is expected to pass 2000 MW by January 2009 and reach to 5000 MW by 2015.13,14 ERCOT has accounted for the vast majority of the wind power capacity additions in the US since 1995. The flow of new capacity and delay in retirement of incumbent utilities’ old generation resulted in reserve margins in excess of 30% in ERCOT in the first 4 years of this decade. However, low electricity prices due to competition and increasing natural gas prices in recent years have forced some of the old- and less-efficient units to be mothballed or retired permanently. The most recent projections indicate a reserve margin of about 17% in the Summer of 2005, declining to under 12.5% by 2010.15 Figure 11.3 presents power plant development, including a few projects by municipal utilities and electric cooperatives, in Texas between 1995 and July 18, 2005. In addition, the amount of capacity under construction, announced capacity, delayed projects, mothballed capacity, and retired capacity by location are identified. The numbers on each location are 12

See PUCT. The following link provides access to various reports regarding renewable resources in ERCOT: http://www.texasrenewables.com/reports.htm 14 Texas Senate Bill 20, which was signed by the Governor in July 2005, requires certain capacity thresholds be achieved in the future. 15 See ERCOT (2005b). 13

391


86

SPP ERCOT SERC SPP WECC

6 47

89

Amarillo Tarrant and Dallas Counties

155

44 66 74 75 80 81 82 148

Travis County

152

79

85

99 28

El Paso

WECC

117

9

8 45 1

33

121

159

92 24 49 52

34

18

116 118

124

51

29

102

95 127

154

62 67

Austin

153

San Antonio

SERC

23 114

31 54

Bexar county

Ellis County

120 39 63

57

59 3

134

SPP

32

143 Dallas

157

46

35

149 107

70 111

65 131 78 Abilene 7 10 146 90 83 165 151 ERCOT 88

38

91

168

147

106 11

58 Kiowa, OK

22

126

21 30 71 101 156

96 Choctaw Cnty, OK

109 138 139 144 145 158 166 167

150

Nolan and Taylor Counties

113

Houston

53

169

16 69 94 136 137

64

14 19

105

78 generation projects completed totaling 32,334 MW 6 generation projects under construction totaling 1331 MW 13 generation projects announced totaling 5891 MW

55 132 133 Corpus 140 160 Christi 98 87

42 13

141 68

Brazoria County 4

12

104 112

Harris county 5

20 36

40 41 56 61 69 73

27 84

76 93 100

103

115 119 123 125 128 161 162 163 164

122 17

37 77

129

135 130 25

Jefferson county

48 142

60

26 50

110 43 15

2 72

07-18-05

32 announced projects delayed or cancelled totaling 17801 MW 18 facilities with mothballed units totaling 8913 MW 22 facilities with retired units totaling 2381 MW

Fig. 11.3. New electric generating plants in Texas since 1995.

referring to a master list that chronologically shows plant additions in Texas. The PUCT web site provides access to the full list of these plants.16

11.4. Restructuring Implementation Under Senate Bill 7 (enacted in 1999), customers of most of the state’s IOUs were permitted to choose among various REPs beginning on January 1, 2002. Pilot customer choice programs were begun on July 31, 2001. Rural electric cooperatives and municipal utility systems were permitted to either participate in the plan (opt in) or decline to participate, although changes in the wholesale market would affect them regardless of their decision. Under the Texas Law, the Commission was given additional authority to conduct market oversight, investigate market manipulation, and enforce compliance with the law. The Commission’s authority also includes regulatory oversight of ERCOT. ERCOT was assigned responsibility to serve as an ISO for most areas of Texas. Stakeholders were given the opportunity to 16 Access to full plant information is provided at http://www.puc.state.tx.us/electric/maps/ gentable.pdf

392


participate in the design and operation of the electricity market through a stakeholder process established by the Commission and the ERCOT bylaws. Senate Bill 7 seeks to balance various objectives. To dilute market power, no generator is permitted to control over 20% of the installed generating capacity in ERCOT. Each PGC associated with a utility with at least 400 MW of Texas-jurisdictional installed capacity must sell at auction entitlements to at least 15% of that capacity. Furthermore, recovery of stranded costs by incumbent utilities was permitted, and a provision was added to allow additional investments designed to reduce air pollution from old generation facilities. While it was initially hoped that all of the state’s IOU service areas would open to competition in a similar time frame, competition was later delayed in the non-ERCOT areas.17 In these areas, it was thought likely that transmission constraints would hamper competition among PGCs. The non-ERCOT IOUs served in multiple states and coordination with other regulatory commissions in other states, who were less enthusiastic about introducing competition, complicated the initiative. The absence of an ISO or regional transmission organization (RTO), that could guarantee non-discriminatory open access, proved to be the primary obstacle to the introduction of retail competition in non-ERCOT areas.

11.4.1. Unbundling The Texas Law permits traditional utility providers to remain involved in both regulated and competitive activities. However, vertically integrated utilities were required to separate or “unbundle” their functions into separate entities, such as PGC, REP, and TDSP, before the start of customer choice on January 1, 2002. Generation and retail activities were largely deregulated, although various safeguards, such as the code of conduct between the regulated company and its competitive affiliates, were imposed. Regulatory oversight over the transmission and distribution of power was retained, since these operations continued to be regarded as natural monopoly activities. Limitations were placed on the exchange of information and personnel between divisions or affiliates of the same entity that were involved in regulated and competitive activities. Since the introduction of retail competition, two out of the three largest incumbent utilities have taken voluntary steps to fully divest their competitive operations from their regulated transmission and distribution services. Both CenterPoint and American Electric Power have fully divested their generation and retail services, and are operating in Texas as TDSPs. In contrast, TXU, the largest incumbent IOU, has sold some of its generation assets, but still continues operating as a vertically integrated utility with functional unbundling.18 11.4.2. Single control area and competitive wholesale markets ERCOT was comprised of 10 control areas operations until July 31, 2001when the retail pilot project went into effect. Previously, each control area was operated by the dominant 17

El Paso Electric Company was spared competition due to its emergence from bankruptcy. However, it was originally hoped that competition would be introduced in the Entergy/Gulf States, Xcel/ Southwestern Public Service Company, and AEP/Southwestern Electric Power Company areas around the same time that retail competition was introduced into the service areas of IOUs in the ERCOT market. 18 For a discussion of similar challenges associated with unbundling experienced in continental Europe see Chapter 7, authored by Haas, Glachant, Keseric, and Perez.


393

vertically integrated utility. It took a massive effort to transform ERCOT into a single control area operation. In addition to extensive investment for ERCOT operation, market participants invested significantly in their infrastructures to prepare for operation within the ERCOT wholesale electricity market. The policy makers and lobbyists responsible for crafting SB 7 focused on avoiding the market structural flaws that they anticipated would emerge in California and other markets. The resulting market structure was designed to permit an extensive reliance upon bilateral contracts, which (unintentionally) makes it similar in structure to the electricity market in Norway. The ERCOT wholesale market was designed to foster bilateral contracts between PGCs and REPs to reduce consumer exposure to hour-to-hour fluctuations in electricity prices.19 Unlike the failed California electricity market, the ERCOT ISO does not operate a centralized spot market for power, but does operate a market for Balancing Energy Service (BES) with some similar attributes. ERCOT is sometimes categorized as a “min-ISO.” It does not presently perform a central dispatch of resources. In addition to managing congestion and operating the BES market, ERCOT administers day-ahead ancillary services (A/S) and acts as the POLR for A/S for those who decide not to self-arrange for such services. Following the 2000–2001 energy crisis in California, the Texas Commission took steps to deter price spikes by imposing a $1000 offer cap on the BES market. Later, a similar cap was imposed on offers submitted for ancillary services.20 Due to the potential for local market power abuses and the fact that ERCOT market was still in transition to a more sustainable competitive market, the state’s regulators were somewhat uneasy about turning over wholesale electricity prices entirely to the forces of competition in a new and untested market structure.21 11.4.3. Maintaining a strong transmission system The Commission has appreciated the importance of transmission expansion to facilitate market operation and has taken a proactive role to enhance transmission network. In cooperation with ERCOT, transmission bottlenecks are identified and TDSPs have constantly improved the transmission network. More than $2.0 billion has been invested in transmission system since 1999 to complete over 4400 miles of transmission lines and other upgrades. In addition, ERCOT has identified about $2.8 billion in transmission projects and system upgrades to be implemented by 2011.22 11.4.4. Markets for generation, ancillary services, and transmission rights A variety of “formal” and “informal” wholesale markets have developed in ERCOT, including: ●

19

Energy markets – Bilateral (informal and not operated by ERCOT) – Day-ahead (informal and not operated by ERCOT) – BES

Market rules are presented in detail in the ERCOT Protocols. For a current version, see www.ercot.com 20 Typically, wholesale electricity prices range from the high 40 s to mid-80 s ($/MWh). 21 See Zarnikau, J. and Adib, P. (2003). 22 See ERCOT (2005a).

394

● ● ●


Ancillary services and operating reserves (day-ahead) Transmission Congestion Rights (TRCs) Renewable Energy Credit (REC)

The ERCOT ISO is engaged in some of the products offered in the wholesale market. As mentioned earlier, ERCOT is responsible for administering real-time BES and day-ahead A/S. In addition, ERCOT auctions TCRs, a hedging instrument to address zonal congestion risks, and is the program administrator for RECs. BES: Energy is bought and sold on a day-ahead basis through bilateral trades and organized markets operated by private entities, while ERCOT operates a market for BES with features similar to a spot market. ERCOT forecasts likely imbalances between supply and demand on an hour-ahead basis, and procures any additional generation (balancing up) or generation reductions (balancing down) that may be necessary in order to better match supply and demand on a near-real-time basis (i.e., with a notice period of between 10 and 20 minutes). QSEs with generation in excess of the quantities necessary to meet the needs identified in their schedules or load curtailment capability are encouraged to submit offers for balancing up energy to ERCOT. BES is settled for each 15-minute interval and is paid a zonal Market Clearing Price for Energy (MCPE) in $/MWh. Today, about 5% of ERCOT’s total generation requirements are satisfied through the BES market, with the balance met through bilateral transactions or through the “informal” market for energy. The QSEs associated with Load-Serving Entities (LSEs) that are deficient in BES are billed for the costs that ERCOT incurs in procuring these services to meet load imbalances within each operating interval. At least to some degree, the BES market improves price signals and transparency, encourages demand-side response, and fosters more efficient use of available system resources. A/S: ERCOT operates a variety of markets for ancillary services.23 While most ancillary service markets are ERCOT-wide, others are zonal or local and designed to address regional or local transmission congestion problems. Some ancillary services are only procured by ERCOT when adverse weather, transmission congestion, or other reliability problems are anticipated, while others are procured on a routine daily basis. QSEs, representing PGCs and/or Loads acting as Resources (LaaRs), submit offers to provide ancillary services on a day-ahead basis. ERCOT creates a bid stack or supply curve of all offers to provide each ancillary service obtained from QSEs, ordering all offers from lowest to highest. Offers are accepted until the market requirement is met. All winning offers receive the market-clearing price. The QSEs associated with LSEs that are deficient in ancillary services are billed for the costs that ERCOT incurs in procuring these services on behalf of LSEs. TCR: The management of transmission congestion has proven to be a key challenge in all the restructuring efforts. Congestion arises when the schedules submitted by QSEs imply power flows that would exceed transmission line limits, resulting in the need to re-dispatch generation units. Congestion costs represent the increased supply costs attributable to the transmission constraints resulting from the re-dispatch. Initially, any congestion costs were simply uplifted to the market and assigned to QSEs in proportion to the load of the LSEs they represent. As congestion costs rose and gaming

23

Ancillary services are the operating reserves and other services necessary for the efficient and reliable operation of the power grid. In ERCOT, these include Regulation Services (Up and Down), Responsive Reserve Service (RRS), Non-Spinning Reserve Service (NSRS), and Replacement Reserve Service (RPRS). A Market Clearing Price for Capacity (MCPC) in $/MW is established for these services.


395

opportunities became apparent, this was replaced, effective on February 15, 2002, with a zonal congestion management system that better assigns congestion costs to the entities responsible for creating the congestion (at least at a zonal or regional level). Initially three, and later five, congestion zones were established to ration the use of the transmission system during periods of binding transmission capacity constraints.24 Congestion between zones is managed through the deployment of balancing energy and other means. QSEs pay for congestion costs based on the QSE’s impact on Commercially Significant Constraints (CSCs) and the constraint’s shadow price or marginal cost. TCRs are auctioned to the highest bidder for monthly or annual periods and provide suppliers with a means of hedging their inter-zonal congestion cost risk.25 In contrast, congestion within each zone is managed through ERCOT’s use of out-ofmerit-order (OOM) instructions. Under an OOM instruction, a resource (i.e., a power plant or an interruptible load) that is not already scheduled for deployment may be deployed by ERCOT to address a transmission congestion problem. OOM costs are uplifted to all market participants (i.e., these costs are not yet directly assigned to the entities responsible for imposing the costs). REC: ERCOT is also the program administrator for the RECs trading program, which provides a market-oriented means of meeting the renewable energy mandate in PURA that an additional 2000 MW of generating capacity from renewable resources be installed in Texas by 2009. This will increase to 5000 MW by 2015 as a result of Senate Bill 20 from a special Legislative Session in 2005. Renewable resource owners earn RECs by their generation and REPs are required to procure a portion of their energy needs through renewable resources. The REC trading program provides a market-based mechanism to address this Legislative mandate. 11.4.5. Introduction of retail competition The introduction of retail competition (Fig. 11.4) involved creating attractive opportunities for new retailers to enter the retail market and compete with the retail affiliates of the incumbent utilities, the establishment of default pricing for smaller customers who decided not to exercise retail choice, the designation of “providers of last resort” (POLRs) who would provide service to consumers in certain situations, the establishment of profiles to estimate the hourly usage of smaller consumers between monthly meter reads, the assignment of consistent account numbers to all consumers, and the development of procedures to track customer switches and “who is serving whom.” Default retail prices were established for REPs affiliated with the incumbent utility (AREPs). AREPs were required to reduce the electricity prices charged to residential and small commercial customers26 by 6%, adjusted for fuel rate revisions and certain stipulated base-rate reductions not yet in effect by January 1, 1999. The resulting price provides a benchmark PTB for potential competitors and a default price for consumers who fail or decide not to exercise their rights to retail choice. This PTB remains in effect for 5 years (i.e., through January 1, 2007). However, the AREP can begin charging rates other than the PTB after 36 months or when the AREP loses at least 40% of its residential or small commercial

24

The current zones are delineated on the left side of Figure 11.5, which appears later in this chapter. See ERCOT (2001). 26 In Texas, the PTB applies to customers with a billing demand below 1 MW. 25

396


TEXAS Service areas offering retail competition

TXUED TXUED TNMP TNMP

AEP-North

AEP-North CNP

TNMP

AEP – Texas Central Company (AEP-Central) AEP – Texas North Company (AEP-North)

AEP-Central

CenterPoint Energy (CNP) Texas Electric Delivery (TXUED) Texas-New Mexico Power Company (TNMP)

Fig. 11.4. TDSP service areas offering competition.

customer load to competitors. After either of these events occurs, the PTB establishes only a ceiling, and the AREP may also offer lower prices. Larger energy consumers in Texas received no price cap protection. Adjustments to the fuel portion of the PTB are limited to two changes per year, unless an AREP is unable to maintain its financial integrity under the prevailing PTB rates or if transmission and distribution costs significantly change. The PUCT designated POLRs to ensure that all Texans received electric service in the event that their REPs went out of business or pulled out of the market. Initially, customers that failed to pay their electric bills were also dropped to the POLR, but later the POLR rules were redesigned so that such customers were transferred to the AREP. Under the current POLR rules, a customer whose electricity contract with a Competitive Retail Electric Provider (CR) expires is sent to the POLR if the customer fails to renew the contract or exercise any other electricity choice. The energy consumption and billing demand of energy consumers with a peak demand of over 1 MW is measured with interval data recorders. Statistical load profiles were developed for smaller customers. The assumed profile for each of the profiled customers is estimated on a day-ahead basis, based on forecasted weather, day type, and other factors. The switching of customers from one REP (e.g., the AREP) to another REP (e.g., a CR) requires the use of ERCOT’s customer databases and systems. As mentioned earlier, ERCOT serves as the central registration agent and maintains customer database for the whole state. The role of the central registration agent provides a way to roll residential consumption into the calculations for settling the balancing energy and ancillary services markets that ERCOT operates, and it minimizes disputes over which REP has the right to serve a particular customer.


397

Meter reads taken by the TDSPs are sent to ERCOT, which then forwards billing data to the appropriate REP. REPs are engaged in marketing, pricing, billing, and bill collection. Further, REPs are responsible for collecting delivery charges to compensate the owners of the transmission and distribution system for its use, as well as contributions to the System Benefit Fund charges, nuclear decommissioning costs, any transition cost or stranded cost charges, and other charges approved by the PUCT. The Systems Benefit Fund, which is set at $0.65 per each MWh sold in ERCOT, has at various times been used to fund programs to assist low-income families, provide customer education regarding customer choice, fund market monitoring and oversight activities, and compensate local school districts for any tax losses resulting from restructuring.

11.4.6. Social programs and policy goals Since the introduction of competition and the lessening of regulatory oversight were likely to compromise other policy goals, Texas’ restructuring plan included a number of special programs to advance certain social and environmental goals: ●

● ●

●

●

The TDSP’s were required to meet 10% of their growth in peak demand through energy efficiency programs. Low-income residential families received electric service at a discounted price. A renewable energy portfolio standard was imposed to increase the state’s renewable energy generating capacity by 2000 MW by 2009. A special energy conservation assistance program was established for low-income families, although funding for this program has at times been diverted to other activities. Utility generators were required to meet more stringent air-emission standards, and were allowed to recover the costs of doing so as a stranded cost.

11.5. An Assessment of the Wholesale Market ERCOT now has over 10 years of experience with the competitive wholesale market, dating back to the passage of Senate Bill 373 in the 1995 Texas Legislative Session, and over 4 years of experience with operations through a single control area (since July 31, 2001). In the following subsection, wholesale market performance will be evaluated.

11.5.1. An overview of performance The restructuring of the ERCOT wholesale market is generally regarded as a success. Reliability has been maintained, and the problems that plagued the California market during 2000 and 2001 have been largely avoided. We have seen increases in the number of market participants and merchant plant developers have continued their interest in this market. Market participants have increased their reliance on the ERCOT-operated ancillary service market from levels of around 10% in the early stages of wholesale market operation to levels of around 30–50% today. Prices for A/S services have been under $30/MW about 97% of the time. Annual energy consumption in ERCOT is about 290 million MWh. Approximately, 2–5% of total energy is transacted through the BES market where prices have been under $100/MWh in about 95% of the time.

398


By all indications, ERCOT has managed congestion cost effectively. Since February 15, 2002, inter-zonal congestion rent (zonal congestion) has been directly assigned to those causing congestion. Congestion rent has dramatically decreased from about $165 million in the first 6 months of market operation to somewhere close to an annual $30 million thereafter. In contrast, intra-zonal congestion cost (local congestion) reached over $1.0 billion by the end of April 2004 and continued to be uplifted on load-ratio basis. On annual basis, local congestion peaked above $400 million in 2003 and it currently stands around $250 million per year. Transmission improvements have significantly contributed to recent reductions in local congestion. Market participants, ERCOT, and the Commission Staff have been working to improve market operation. More than 400 Protocols Revision Requests have been initiated resulting in noticeable improvements in system operation and market efficiencies. The current wholesale electricity market has maintained effective bilateral markets, produced a healthy expansion of capacity, and provided opportunities for certain industrial customers to respond to prices. Furthermore, it has maintained full independence for market monitoring functions through the Commission’s Wholesale Market Oversight (WMO) group since August 2000.27 In addition, significant transmission expansion has taken place within ERCOT. While the restructuring of the wholesale market has been generally successful, it is expected that further improvements can be achieved by transitioning the structure of the wholesale market to a nodal design. The next subsection contrasts ERCOT’s current zonal market structure with the nodal market design that has been adopted in numerous other markets. 11.5.2. Choice of market design: zonal versus nodal A review of various electricity markets reveals that there are two dominant categories of market design used throughout the world:28 1. Nodal market design where each point (node) has its own energy price. 2. Zonal market design where one energy price is set across a large area (zone).29 Figure 11.5 includes a pictorial presentation of these two models. Nodal market design requires significant involvement by the ISO (Max ISO) where market decisions have large impacts on reliability and centralized commitment and dispatch process are the key features 27

Texas Senate Bill 408, which was signed by the Governor in June 2005, requires the creation of an Independent Market Monitor (IMM) to be fully funded by ERCOT and be hired by and report directly to the PUCT. 28 Significant amount of literature is written regarding market design since the electric industry restructuring debate begun in California in mid-1990s. see Harvey, S. and Hogan, W. (2000a, b), Hogan, W. (1998), Ruff, Larry (2004), Wolak, F. et al. (2004). The following Websites provide access to papers on electricity restructuring and market design: Professor William Hogan’s Web site at Harvard University (www.whogan.com), John F. Kennedy School of Government at Harvard University (www.ksg.harvard.edu/hepg/Competitive_Models.htm), and University of California Energy Institute at Berkeley, Center for the study of Energy Markets (www.ucei.berkeley.edu/). 29 Zonal market design is a subset of Nodal market design where a few nodes (zones) are considered. Therefore, in contrast to nodal market design where prices are determined for several thousand nodes, we have a few prices, one for each zone.

The Most Robust Competitive Market in North America Zonal

399

Nodal

NE N W S

H

Commercially significant constraint Fig. 11.5. A zonal ERCOT market versus a nodal market structure.

of this design. Most of the restructured electricity markets in the US, such as Pennsylvania– New Jersey–Maryland (PJM), New York Independent System Operator (NYISO), New England Independent System Operator (ISO-NE), and Midwest Independent System Operator (MISO), operate under a nodal design.30 Zonal market design requires minimal involvement by the ISO (Min ISO) where market decisions have a small impact on reliability, and decentralized commitment and dispatch process are the key features of this design. The California Independent System Operator (CAISO), which began with a zonal design in March 1998, is expected to switch to a nodal design by February 2007.31 ERCOT is the second US wholesale electricity market that still operates under zonal design. The Commission and stakeholders are engaged in activities that will result in ERCOT implementing a nodal design in the near future. In contrast to various Northeastern US markets that were centrally operated for many years as power pools before they were restructured, ERCOT never operated as a power pool. Therefore, it was easier and less costly to transition to a single control area, maintain historical decentralized commitment and dispatch decisions with generation entities, and give QSEs the ability to manage their plants close to real time. In addition, it was more transparent to have clearing prices seen in the bid stack while maintaining decision-making with the market participants rather than with ISO. 11.5.3. Problems with the existing zonal system Due to the nature of competitive transactions scheduled by market participants, the ERCOT market is faced with significant transmission constraints, particularly within major zones, and costs are incurred in resolving these constraints. When ERCOT’s zonal market was initially designed, it was generally expected that:32 1. local congestion would be random and infrequent; 2. zonal prices would be sufficient for siting resources; 30

See Chapters 13 and 14. See Chapter 10. 32 See Schubert, E. (2004). 31

400


3. adjusting the number of CSCs and load zones would be systematic and timely; 4. market decisions would have a small impact on reliability; 5. decentralized unit commitment and dispatch process would not result in operational inefficiency. The real experience in the past 3 years proved that these assumptions could not be sustained in ERCOT market. These expectations could not be met without compromising operation and reducing market efficiency. In reality, it became apparent that: 1. local congestion is systematic and frequent and the resulting congestion cost uplifts create opportunities for gaming and market manipulation; 2. zonal prices result in inefficient siting and dispatch of resources;33 3. the creation of new CSCs and load zones is problematic and creates commercial uncertainties; 4. market decisions have significant impact on reliability; 5. the portfolio dispatch nature of current market design creates real-time operations difficulties and results in additional costs to maintain system reliability; 6. decentralized unit commitment may not be inefficient. These issues, particularly regarding local congestion and uplift, were raised first by the WMO and its consultant.34,35 Market participants have also raised concerns with regard to the operation of the current zonal market. For example, REPs have raised concerns about:36 1. 2. 3. 4.

33

uncertainty regarding the annual configuration of zones; price volatility in the spot market; potential market manipulation; high chances of bankruptcy by smaller REPs due to lack of adequate transparency in market operation and pricing.37

In the McCamey area of West Texas, 758 MW of wind-to-energy capacity was constructed in an area with a transmission export capability of only 400 MW. It is understandable that for an intermittent resource like wind, the optimal arrangement may be to overbuild the generation, because it will still be dispatched at its maximum achievable output in a large number of hours in the year and constrained in relatively few. However, in an aggregated level, the construction of renewable generating capacity outpaced the construction of the necessary transmission capability, highlighting the absence of coordination between generation siting and transmission planning and an absence of proper pricing signals to guide investment siting decisions. 34 Market Oversight Division (MOD), Comments on Issues Related to the Transmission Congestion Workshop on September 18, 2002. Project No. 26376, Transmission Congestion Issues In The ERCOT. PUCT, September 9, 2002, Austin. 35 Dr. Shmuel Oren, from the University of California at Berkeley, is the Senior Consultant to the PUCT’s WMO. 36 Some of the listed concerns may not be affected by the zonal-nodal choice. In fact, the nodal market may require additional resources by smaller market participants to understand a more complex market and additional capital to hedge transmission risks. 37 Texas Commercial Energy, a REP working in ERCOT retail market, filed for bankruptcy on March 6, 2003. In July 2005, Azor Energy failed to meet its obligations that resulted in the transfer of about 500 customers to other REPs. Energy West also failed to meet its obligations to TDSPs and transferred its customers to the POLR.


401

In contrast, power generators have raised concerns regarding (a) inadequacy of revenues to recover capital cost in a timely manner, and (b) the presence of overly restrictive mechanisms which were designed to address market power and perform price mitigation. Additional interest has been shown by some market participants in the creation of an integrated day-ahead energy market which might be cleared using a market-based security-constrained unit commitment/economic dispatch model. It is hoped that this additional feature would address the current lack of liquidity in the BES market and provide more price transparency, which would in turn facilitate load response in the wholesale electricity market. The Commission and stakeholders have been working over the last couple of years to develop a new and improved market design that incorporates sufficient system engineering elements into economic dispatch, is incentive compatibility, results in the most efficient use of resources, minimizes opportunities for gaming and uplift, and at the same time provides flexibility to adapt to changing conditions. In a September 2005 Open Meeting, the Commission finalized its decision to fully implement a nodal market design structure by January 1, 2009. In addition, the Commission is expected by the first quarter of 2006 to finalize its decisions in several other complementary rulemaking projects that:38 1. establish an energy-only resource adequacy mechanism; 2. define market power and identify conditions to be imposed on market participants who have market power; 3. determine its mitigation mechanism that allows for scarcity pricing while mitigating prices that are not the result of competitive market forces; 4. establish its requirements regarding the creation of an Independent Market Monitor (IMM) for the wholesale electricity market in ERCOT. In finalizing its decision, the Commission is facing several policy challenges: 1. Which zonal market design flaws, that have been identified, should be addressed over the next 3 years while waiting for full nodal system implementation?39 2. How to balance the costs of implementing the new and improved design and expected potential benefits? 3. How to gain market efficiencies while mitigating the unintended consequences of the new and improved design? 4. What is a reasonable transition plan to fully implement the new and improved design?40 11.5.4. Is it workably competitive? Market concentration and market power As experience in the California market demonstrated, there is still insufficient competition in many newly restructured electricity markets at this point. In fact, the ERCOT market is still in a transitional stage; there is a need for more customer education and safeguards; 38 Project No. 24255, Rulemaking Concerning Planning Reserve Requirements, Project No. 28500, Activities Related to the Implementation of a Nodal Market for the ERCOT, Project No. 29042, Rulemaking on Definition of Market Power, and Project No. 31111, Rulemaking to Address An IMM for the Wholesale Electricity Market in ERCOT. 39 Project No. 31575, Improvements to the ERCOT Zonal Market Design. 40 Project No. 31600, Transition to an ERCOT Nodal Market Design.

402


markets lack adequate opportunities for price responsiveness by customers; too much local market power exists due to strategically located units; markets need significant improvements in their transmission infrastructures; and newly deregulated electricity markets are vulnerable to gaming and manipulation.41–44 Texas Legislative and regulatory authorities have recognized the importance of preventing market power abuses and have taken steps to address this issue through legislative amendments to the Texas PURA and the establishment of regulatory rules by the Commission. Earlier studies using simple measures of market concentration, such as Hirfindahl– Hirschman Index (HHI), suggested that the ERCOT market was either highly concentrated or moderately concentrated.45 The Commission staff also examined market concentration in sub-markets within ERCOT.46 Furthermore, the Commission observed that ERCOT was probably the most concentrated of the deregulated wholesale markets in the US.47 The foresight of the Texas legislators resulted in the inclusion in Senate Bill 7 (1999) of a 20% cap on installed capacity within ERCOT by any generation company. In addition, the restructuring legislation required most of the IOUs to auction entitlement to energy from 15% of their installed capacity. Entitlements are offered in various product types with terms that can last up to 4 years. Energy from entitlements can be offered into bilateral and ERCOT administered markets. In spite of positive impacts by these two provisions, the market power of incumbent affiliated generators is still a major concern, particularly within zones and load pockets in ERCOT. The Texas Law has given the Commission authority to address market power abuse and anticompetitive behavior. Like other electricity markets with market monitoring units, the Commission created WMO48 in August 2000 to monitor wholesale activities within ERCOT market and take steps to minimize the abuse of market power in the electricity market. Several antitrust lawsuits have been initiated before the federal courts raising allegations against TXU and some other market participants in the past 2 years.49,50 There has not been conclusive evidence to demonstrate abuse of market power51 by TXU or any other market participants. However, the reports by the Commission WMO and investigation by its consultant have demonstrated that TXU is a pivotal supplier in many instances in the BES market.52 41

PUCT, Docket No. 25755. PUC investigation into overscheduling in ERCOT in August 2001. See Adib, P. (2002b). 43 See Adib, P. (2002a). 44 See Adib, P. (2004). 45 See Zarnikau, J. and Lam, A. (1998). 46 See Synchronous Interconnection Committee (1999). 47 MOD, PUCT, Comparison of Market Designs, Project No. 26376: Rulemaking Proceeding on Wholesale Market Design Issues in the ERCOT, January 7, 2003. 48 It was originally called MOD. The Commission later established Retail Market Oversight (RMO) in September 2002. 49 Texas Commercial Energy vs. TXU Energy, Inc., et al., C.A. No. C-03-249, USDC, Southern Dist. of Texas – Corpus Christi Division (2003). 50 Utility Choice Electric and Cirro Group, Inc. d/b/a Cirro Energy Corporation vs. TXU, et al., C.A. No. 4:05-cv-00573, USDC, Southern Dist. of Texas – Houston Division (2005). 51 While the Texas Law does not prohibit having market power, the PURA prohibits the abuse of market power. 52 MOD, Staff Inquiry into Allegations Made by Texas Commercial Energy regarding ERCOT Market Manipulation, Project No. 25937, PUCT, January 28, 2004, Austin. Potomac Economics, Investigation into the Causes for the Shortages of Energy in the ERCOT Balancing Energy Market and into the Wholesale Market Activities of TXU from October 27 to December 8, 2004, April 2005. 42


403

Table 11.4 summarizes generation market share and HHI calculation by major companies in ERCOT as well as within its main five zones. All PGCs with installed capacity above 2000 MW are listed in this table. As it is evident from the figures, TXU is the dominant player in the North and Northeast as well as in the West zone. Texas Genco is the dominant player in the Houston Zone, followed by Calpine. CPS Energy, while not comparable in shares with TXU and Texas Genco, is the largest player in the South Zone. The South Zone is the most competitive area, resulting in lowest energy prices. 11.5.5. Price volatility As mentioned before, 2–5% of total energy is transacted through the BES market where prices have been under $100/MWh about 95% of the time. However, acting as the spot market in ERCOT, BES has played an important role in providing price transparency and impacts prices in ancillary services markets as well as in bilateral transactions. Such direct impact was demonstrated during an extreme weather event experienced in ERCOT during February 24–26, 2003. The colder than anticipated weather struck North and Central Texas and gas curtailments affected some generating units while a number of large generating units were down for maintenance. “Hockey stick bidding”53 by generation suppliers contributed to the price spike.54 The high $990 per MWh price observed for several hours in the BES market resulted in similar prices for the next couple of days in the ancillary services and BES market.55,56 Occasional price spikes in ERCOT’s wholesale markets have occurred since the opening of the market. Figure 11.6 provides BES prices since the beginning of 2003. While some price spikes (arguably) appear to have been the result of market power abuse or certain bidding behaviors, whether an occasional price spike is necessarily a problem is hotly debated. PGCs argue that an occasional price spike may reflect actual physical scarcity (i.e., a constraint associated with inadequate generating or transmission capacity), provides a proper price signal to encourage load response, and the resulting revenues to generators are necessary to their financial integrity. 11.5.6. Demand response As noted by FERC (2002): “Demand response is essential in competitive markets, to assure the efficient interaction of supply and demand, as a check on supplier and locational market power, and as an opportunity for choice by wholesale and end-use customers.” Traditionally, ERCOT enjoyed roughly 3500 MW of demand-side resources, including interruptible service tariffs, group load curtailment programs, direct load control, and energy storage devices. Loads served under interruptible tariffs constituted the largest resource, about 3200 MW. To retain some of this resource potential, steps were taken to facilitate demand response and permit demand-side resources an opportunity to provide ancillary services. The 53

This refers to the practice of submitting a small portion of a bidder’s resource fleet to the market at a very high price, in hopes that the small-high price increment will set the clearing price for the whole market. 54 MOD, PUCT, Analysis of Balancing Energy Price Spikes during the Extreme Weather Event of February 24–26, March 3, 2003. 55 Ibid. 56 See Saathoff, K. (2003).

404

Table 11.4. Generation market share and HHI calculation in ERCOT as of July 14, 2005. North Owners

MW

ERCOT Total HHI

26,050 –

South

%

MW

%

44.6 6.1 3.7 – 1.2 6.5 –

3146 – – – – 982 –

55.2 – – – – 17.2 –

6.0 –

1495 –

8.9 5.0 –

– – –

MW

West

Houston

%

MW

%

97 1113 1491 5468 3317 – 3379

0.4 4.8 6.5 23.7 14.4 – 14.6

3114 – – – – – –

47.6 – – – – – –

26.2 –

– 2471

10.7

– –

– – –

– 898 1020

– 3.9 4.4

– – 1034

MW – 8407 3832 – – 739 –

Total % – 55.2 25.2 – – 4.9 –

MW

%

17,964 11,122 6299 5468 3622 3421 3379

23.5 14.5 8.2 7.1 4.7 4.5 4.4

– –

– –

– –

3066 2471

4.0 3.2

– – 15.8

152 – –

1.0 – –

2463 2206 2054

3.2 2.9 2.7

17.9

80

1.4

3838

16.6

2392

36.6

2092

13.7

13072

17.1

100.0 2271

5703 –

100.0 4029

23,092 –

100.0 1424

6540 –

100.0 2833

15,222 –

100.0 3823

76,607 1019

100.0 1019

Note: Combining North with Northeast Zone results in an HHI of 2465. An application for merger between Exelon and Public Service Gas and Electric (owns TIE) is pending before regulatory authorities.


TXU 11,607 TX Genco II 1602 Calpine 976 CPS Energy – Sempra 305 FPL 1700 Austin – Energy Tenaska 1571 Lower Colorado – River Authority Exelon 2311 ANP 1308 Texas Independent – Energy All Others 4670

Northeast


405

300 UBES 2/25/03 $523.45

Energy price ($/MWh)

250

200

150

100

0

January 2003 February 2003 March 2003 April 2003 May 2003 June 2003 July 2003 August 2003 September 2003 October 2003 November 2003 December 2003 January 2004 February 2004 March 2004 April 2004 May 2004 June 2004 July 2004 August 2004 September 2004 October 2004 November 2004 December 2004 January 2005 February 2005 March 2005 April 2005 May 2005 June 2005 July 2005

50

Fig. 11.6. Energy price – on peak up balancing market, January 2003–July 2005. Source: PUCT.

demand-reduction potential and curtailments from interruptible or price-responsive customers can compete head-to-head with an offer from a power plant to provide generation or an operating reserve. In mid-May 2002, interruptible and curtailable loads were permitted to participate in the ERCOT wholesale market for responsive reserves as a resource on a limited basis, and interruptible loads were introduced into the markets for non-spinning reserves and replacement capacity in October of 2002. By July 2005, over 1800 MW of interruptible loads had been registered to participate in ancillary services markets. Most of the qualified LaaRs are large industrial chemical and refinery loads with relatively predictable load patterns. While over 1800 MW of load is qualified to provide an operating reserve, the amount of responsive reserves that can be awarded for deployment through LaaRs is limited to 1150 MW at any given time.57 As balanced schedule requirements were relaxed in November 2002, some larger energy consumers who were formerly insulated from wholesale price signals through regulated tariffs could acquire power at market-based wholesale market prices by satisfying all or a portion of their purchased power requirements with balancing energy or through other creative contractual arrangements with their REPs. Further, the market structure provides 57

Historically, LaaRs could only contribute up to 25% of required 2300 MW responsive reserves. Later, this share was increased to 50% or 1150 MW

406


some incentives for consumers to reduce power purchases during peak or high-price periods. For example, a large industrial energy consumer’s transmission charge is based upon the consumer’s contribution to ERCOT’s coincident peak demand in four Summer months, so some industrial consumers actively try to reduce energy consumption during expected peaks. Some industrial energy consumers who rely on BES to meet some or all of their electricity needs actively monitor the 15-minute balancing energy prices, and reduce electricity purchases when prices exceed threshold levels. Other industrial consumers of energy participate in curtailment or scheduling programs that are established by their REP. These types of voluntary demand responses are based on confidential contractual relationships between a REP and its large energy consumers and not announced to the market. An analysis of the 20 largest Houston area industrial energy consumers found that while in the aggregate these loads exhibited very little responsiveness to wholesale electricity prices in 2003, a couple of the individual energy consumers in this group did indeed exhibit a pronounced response to price changes.58 It is suspected that some of the industrial energy consumers with the most operational flexibility opted to provide their interruptibility to the ERCOT market as an operating reserve. Once an energy consumer commits to providing an operating reserve, it must follow a predictable load pattern and can no longer chase prices. Some demand-side resources failed to successfully transition to the new market structure. Direct load control efforts involving small energy consumers have been terminated due to metering limitations, the need for expensive system changes to recognize their impacts, and other considerations. Thermal energy storage devices have not proven economically viable in the new market due to limited differentials in the cost of electricity between on- and off-peak periods. Steel mills, due to the unpredictable nature of their load levels, are considered to be unfit to provide responsive reserves ancillary services and have also seen their interruptibility under-valued under the new market structure. Over 100 MW of group load curtailment programs are no longer in operation. 11.5.7. Wholesale assessment: a summary In summary, wholesale market operations have proceeded fairly well. Reliability has been maintained at acceptable standards and competition has increased. The wholesale market design is certainly not perfect – some gaming opportunities were quickly discovered, the local congestion management system has failed to appropriately assign costs in an economically efficient manner, and some generation-siting decisions have created market problems and operational difficulties. Yet, ERCOT’s wholesale market prior to restructuring had its own set of problems. The move to the nodal market design is expected to address most of the current zonal market design flaws and market participants remain generally optimistic that the remaining transitional problems can be resolved. However, further movement toward a nodal system of congestion management could pose a major challenge.

11.6. Retail Competition in ERCOT: An Assessment After an extremely difficult transition period, ERCOT’s retail market operations are now working well and a high degree of competition at the retail level has been achieved. 58

See Zarnikau, J., Landreth, G., Hallet, I. and Kumbhakar, S. (2005).


407

11.6.1. Early problems with switches, billing data, computer systems Initially, ERCOT had a great deal of difficulty in performing many of the customer service and billing activities that the vertically integrated utilities had performed so well.59 Some of the tasks that proved formidable to ERCOT included assigning account numbers (electricity supply industry (ESI) IDs) to new premises, customer switching, collection and distribution of billing data, and disconnection and re-connection of service (“Move-In, Move-Out”). Customer switching problems during the early months of retail competition have been traced to many causes, including software failures, vague language in ERCOT’s Protocols regarding when a TDSP can reject a switch request, higher-than-anticipated traffic on the ERCOT portal (which provides necessary information to REPs about specific customers), errors in ERCOT’s ZIP code database (which was used as a check to verify the validity of information pertaining to a customer), delays in creating ESI IDs for new buildings and premises, and typographic errors in switch requests. Further challenges emerged in ensuring that all energy consumers were assigned a REP, particularly in situations where a new home or building had been constructed and when consumers had recently moved in or moved out of a building. As a result, REPs were not certain which customers they had, and for which customers they were responsible for procuring and scheduling generation resources. This also resulted in problems in providing bills to consumers in the first year of retail competition. After a difficult 12 to 18 months, these problems were eventually resolved through numerous software changes, hardware upgrades, Protocol changes, and “manual work-around.” The early decision to require ERCOT to act as clearing house for retail customer registration and switches was made to establish trust in the system and prevent any potential barriers to entry for REPs who were interested in competing in the ERCOT market. The progress in the past 4 years shows that, although implementation was extremely difficult, it was a reasonable decision. The lessons learned by ERCOT can be effectively utilized in other markets if they decide to follow similar blueprint. 11.6.2. New market entrants and numbers of active REPs Ease of entry, ease of exit, and the occasional bankruptcy may be signs of a healthy competitive market. The ERCOT market has attracted many new entrants at the retail level, including independent REPs (e.g., Green Mountain Energy), affiliates of utilities that previously had no operations in ERCOT (e.g., Sempra, Xcel Energy Retail Services, Energy Solutions, Strategic Energy, and Constellation), IPPs or oil and gas company affiliates that entered the retail market (e.g., BP Energy, Dynegy, Calpine, Tenaska, Suez, and Coral), and new firms (e.g., Utility Choice, GEXA, and Texas Commercial Energy). In July 2005, 85 REPs were certified to provide electricity at the retail level. About seven to twelve REPs compete with the AREPs for residential loads in each incumbent service area. Competitive REPs serve about half of total sales in ERCOT (about 25% of residential sales, 55% of small commercial sales, and more than 70% of large commercial and industrial sales). Much of the greatest competition has been from AREPs competing outside of their traditional service areas (e.g., TXU Energy Solutions, Reliant Energy Solutions, and First Choice Power competing in each other’s former retail service areas). 59

As was mentioned earlier, ERCOT is the only known competitive electricity market that acts as a clearing house for retail customer registration and switches.

408


Some REPs abandoned the retail market even before competition was fully initiated, including Enron (due to its bankruptcy) and Shell. American Electric Power Company also decided to transition out of the retail market before retail choice was fully initiated – most of its remaining retail customers were switched to Centrica. Texas Commercial Energy’s lack of an “effective” hedging strategy eventually caught up with it and sent the upstart into bankruptcy. A number of REPs at least temporarily terminated efforts to recruit additional customers after failing to achieve profit targets or after running into credit problems, including AES (which later sold its retail operations to Constellation). Xcel Energy Retail Services left the market after selling its customer base to MPower. Overall, the number of retailers actively competing for customers has been quite impressive.

11.6.3. Switching activity The share of customers that switch to an alternative provider is a commonly cited indicator of competition.60 The share of residential consumers who have switched to CRs has grown to nearly 25% as indicated in Figure 11.7.61 With no PTB protections and often with greater opportunities for savings, about 70% of large commercial and industrial customers had switched by July 2005. Large commercial and industrial customers who have not switched to competitive REPs have been exposed to market-based prices since competition began in January 2002. Figure 11.8 compares the customers served by the AREP to the customers served by CRs in each service area where retail competition has been introduced. Since larger energy consumers have much higher switch rates, the energy sales by CRs are just below 50% in the areas of ERCOT opened to retail competition. Sales by AREPs are compared to sales by CRs in each service area in Figure 11.9. As in many competitive retail markets, there has been less competition for residential consumers, accounting for about 29% of total retail energy sales in Texas, than for commercial and industrial loads. Many REPs have found that it simply isn’t economical to serve the residential sector, due to high marketing and transactions costs and limited savings that can be promised to consumers relative to the regulated PTB that the AREP must charge to consumers decide not to exercise customer choice.62 REPs, which serve more than 300 MW of load, that fail to sell at least 5% of their energy sales to residential consumers face a penalty.63 However, many REPs have elected to simply pay the penalty rather than deal with smaller energy consumers. In a market where natural gas prices have a considerable impact on electricity retail prices, electricity prices will be volatile. The retail prices quoted by REPs may be honored for periods as short as a few hours. Locking-in prices at an appropriate time presents a challenge for some retail customers. In spite of less attraction for small energy consumers, switching rates have been quite high in ERCOT market relative to other markets in the US where retail competition has been introduced. It is worth noting that in Texas, the competition has been on a “one customer at a time” basis, like selling cars or groceries. This is in contrast to the competition in the eastern US for the right to provide default service through an auction process. 60

See Chapter 1 for the discussion of the value of this metric. See Hudson, P. (2005). 62 This topic is further addressed in the Introduction to this volume. 63 PURA, Sec. 39.352(g). 61

409

The Most Robust Competitive Market in North America 25.00 20.00

%

15.00 10.00 5.00 0.00

02 02 02 02 03 03 03 03 04 04 04 04 05 05 20 il 20 20 r 20 20 il 20 20 r 20 20 il 20 20 r 20 20 il 20 y y y y y y a Apr Jul obe uar Apr Jul obe uar Apr Jul obe uar Apr nu ct an ct an ct an a J O J O J O J ry

Fig. 11.7. Residential customers with competitive REP. Source: PUCT.

100

23.71%

24.34%

27.75%

29.49%

23.61%

75.7%

72.3%

70.5%

76.4%

80

%

60 40

76.3%

Affiliate Non-Affiliate

20 0 TXUED

Center point

CPL

WTU

TNMP

Fig. 11.8. Breakdown of all customers by REP status. Source: PUCT.

49.55%

47.05%

63.64%

64.11%

56.81%

50.45%

52.95%

36.36%

35.89%

43.19%

%

100 90 80 70 60 50 40 30 20 10 0

TXUED

Center point

CPL

WTU

TNMP

Fig. 11.9. Percentage of energy sold by affiliate status of REP. Source: PUCT.

Affiliate Non-Affiliate

410


11.6.4. Load aggregation By June 2005, over 130 load aggregators had registered with the PUCT. Load aggregation has proven effective, but not in the manner predicted. True aggregation (where the load diversity benefits from serving large groups of customers is recognized) is difficult in Texas, for a number of reasons. ERCOT settles each QSE’s ESI IDs separately. If customers under 1 MW without interval data recorders (IDRs) fall within the same load factor group (and the same weather zone), their load pattern is assumed to be the same as all other customers in that group. Thus, the load profiling system in ERCOT fails to recognize any load diversity among different customers within the same load factor category, and the true diversity benefits associated with serving a group’s power needs are difficult to recognize. The transmission and distribution costs associated with serving each commercial customer are billed to the REP based on the customer’s billing demand. Since this billing demand is usually based on the customer’s non-coincident peak, here again load diversity is not taken into consideration. Consequently, true aggregation is difficult unless all members of the aggregation group have IDRs. Yet, load aggregators have been able to deliver significant savings to their clients by improving their client’s bargaining power and negotiating strength and the consulting services provided through the programs. Aggregators have been particularly effective in the government arena, where they have provided a procurement service that permits customers to comply with state procurement laws in a cost-effective manner.

11.6.5. Price trends The available evidence suggests that most large commercial and industrial energy consumers have realized cost savings as a result of restructuring, although the amount of savings is a subject of dispute. There has been intense competition among REPs for the opportunity to serve this sector. Whether smaller consumers, as a whole, have realized net savings is under debate. It is not difficult to show that smaller consumers, who have exercised their choice by switching to alternative providers, have realized reasonable savings relative to the PTB. However, the majority of residential energy consumers have not yet exercised their right to switch to a competitive REP and have remained on a regulated PTB. Consumers that remained on the PTB initially realized savings due to a legislatively mandated 6% discount (relative to January 1999 prices) to base rates introduced in January 2002. However, the practice of indexing this PTB to natural gas prices, combined with the rapid increases in natural gas prices experienced in 2003 and 2004, has resulted in frequent price increases for residential and small commercial customers that remained on this PTB in 2003 and 2004. Average residential rates (considering both the PTBs offered by incumbent AREPs in their own service territories and competitive price offers) have risen at a faster rate in the areas of Texas where retail competition had been introduced than in those areas, such as non-opt-in areas within ERCOT and all Texas areas outside of ERCOT, where retail competition was not yet available.64 Constructing a baseline reflecting what electricity prices might have been had restructuring not occurred is fraught with difficulties. Yet, the available evidence suggests that the market has at least provided consumers who exercise their retail choice with opportunities for savings, relative to the PTB and (arguably) to the prices they would have paid absent restructuring. Following the expiration of the PTB default pricing for smaller consumers 64

See Zarnikau, J. and Whitworth, D. (2005).

411


0.14

December 2001 rates January 2002 price to beat rates October 2004 price to beat rates October 2004 lowest competitive offer

Dollars per kWh

0.12

0.10

0.08

0.06

0.04 TXU electric delivery

Centerpoint

AEP central

AEP north

TNMP

Transmission and distribution service area Fig. 11.10. Price to beat versus the lowest residential competitive offer by service territory.

on January 1, 2007, the market’s ability to deliver savings to consumers will be truly put to the test.65 Particularly in recent months, there has been a widening gap between the PTB and the lowest prices available from CRs in the market. Some recent data are presented in Figure 11.10 suggesting the potential savings relative to the PTB that are available to smaller customers who exercise their power to choose. 11.6.6. Retail assessment: a summary Despite a rocky start, ERCOT’s competitive retail market has achieved a relatively high degree of competition. Nearly one-half of retail sales in the competitive areas are through CRs. ERCOT’s central registration function is now operating fairly smoothly. The effect of competition upon prices requires further analysis following the expiration of the PTB.

11.7. Lessons Learned Following 10 years of experience with a restructured wholesale market and more than 3 years of experience with retail competition in Texas, we can offer some insights to policy makers and analysts in other areas where restructuring is being contemplated.

65

See we note that Rose, K. and Meeusun, K. conclude: It appears that, from the data so far, most retail customers (especially residential) in restructured states where the transition period has ended and the price is now based on the wholesale market, are seeing prices increase faster than in non-restructured states or states in transition with a price cap. At best, at this point in time, no discernable overall benefit to retail consumers can be seen from restructuring.

412


11.7.1. Creation of a central registration agent function can be an overwhelming task. There might be some intuitive appeal to assigning responsibility to a single independent entity for tracking which REP is serving which account and facilitating the flow of billing information from TDSPs to the appropriate REPs. The integrity of the market may be enhanced when an unbiased party performs function. Further benefits may accrue through of processes across the entire market. However, in ERCOT this task proved much more difficult than initially envisioned. The initial problems led to some loss of confidence in the new retail market during its first 12 to 18 months. Other markets that are similarly contemplating the establishment of a central and independent central registration agent should expect long transition times and should perhaps consider parallel operations with utility customer care operations over some period of time. The procedures, software, and networks used to assign account numbers, handle new premises, disconnect customers, identify the correct “retailer of record,” move billing data to the correct retailer, and verify switch requests must be thought out with great care. The lessons learned by ERCOT can be of great value to those markets that prefer such central and independent central registration mechanism. 11.7.2. The need for adequate resources and infrastructure for market monitoring The decision by ERCOT’s stakeholders in 2000 to require that market monitoring functions be conducted by an entity other than ERCOT resulted in the creation of the most IMM function among all existing competitive electricity markets in the US. In contrast to other operating electricity markets where the market monitoring unit has been established as a part of ISO, the WMO was established within the PUCT. Distance from real-time operators and a lack of adequate resources within the PUCT were the main difficulties that were created by this decision. The Commission took immediate steps to reallocate some of its limited resources toward market monitoring functions. Fortunately, legislative authorities were responsive to requests submitted by the Commission in the 2003 Legislative Session and allocated additional resources for the biennial fiscal years 2004–2005. Finally, the passage of Senate Bill 408 in recent 2005 Legislative Session required ERCOT to allocate additional resources through its Commission approved administrative fees to fund the creation of an IMM. The Commission is in the process of developing a rule to address resource needs and ethical standards for the market monitoring functions. The creation of IMM will allow the Commission WMO to further focus on investigation and enforcement activities in the near future, which is similar to responsibilities assigned to the FERC Office of Market Oversight and Investigation with regard to other existing electricity markets in the US. 11.7.3. Customer education Customer education is essential for the success of retail competition. Texas legislative authorities recognized the importance of customer education and directed the Commission to educate Texans about electric competition by providing neutral and non-promotional information to customers. The Commission was authorized a total budget of $36 million to be spent on customer education for a period of 3 years ending in 2003.66 A public relations 66

The Commission is further authorized to spend an annual budget of $750,000 to continue its customer education efforts beyond 2003.


413

firm was hired to work with the Commission to promote an understanding of the “power to choose” and retail competition. As a result, customer awareness was increased from 15.3% in 2000 to 75.6% in 2003, the last year in which funding was available for customer education.67 In spite of effective efforts to educate customers, more than 70% of retail customers have remained with the incumbent AREPs, and have not shopped around for the savings offered by competitive REPs. Additional customer education would be beneficial in encouraging these customers to take advantage of the best retail competitive market in the US. 11.7.4. Required transition time Adequate transition periods are essential to any restructuring effort. Texas’ strategy of reforming its wholesale market and creating an ISO long before introducing retail competition proved effective. Texas opened wholesale market in 1995 and waited until July 31, 2001 before initiating its retail pilot program. In contrast, California, the pioneer in electric restructuring, published its blue book in 1994 and opened its wholesale and retail competition at the same time in March 1998. While some may claim that proper conditions for a robust competitive market, particularly for residential customers, were not in place, it is fair to say that the policy makers in Texas recognized the risk and included adequate safeguards in the Texas Law to protect customers while allowing interested customers to shop around and get the benefits of competition. The progress in the past 4 years shows that it was a reasonable decision. Nonetheless, one area in which a longer transition time would have been prudent was in the creation of a central registration agent. The movement and consolidation of utility customer account information and billing information to the ISO and the establishment of the procedures necessary to assign account numbers, complete switch request, and transfer billing information to the appropriate organizations required much more time than allotted.

11.8. Challenges Ahead Despite our early success, we acknowledge that we have many challenges ahead in our efforts to improve upon our success with introducing competition into the ERCOT market. 11.8.1. Transition to a nodal market structure Texas policy makers recognized the importance of a strong transmission system to foster competition, and a construction boom in transmission lines is underway. Nonetheless, transmission constraints within the ERCOT market have proven to be a greater problem than originally appreciated, particularly in the Dallas-Fort Worth area, the Rio Grande Valley, Laredo, and West Texas. There is growing dissatisfaction with the present approach to congestion management. While congestion between regions or zones is now being managed in a reasonable and economical manner, the costs associated with managing local congestion are still being uplifted. Some market participants suggest adopting the method used in PJM; that is, locational marginal pricing (LMP). The PUCT has clearly expressed interest in moving toward some type of nodal pricing which would better assign all congestion costs to the market participants responsible for creating congestion. 67

For more information please visit http://www.powertochoose.org

414


The Commission is facing several policy challenges, such as (a) how to gain market efficiencies while mitigating the unintended consequences of the new and improved design and (b) what should be a reasonable transition period to fully implement the new and improved design. The Commission is expected to finalize its decision on details of the nodal pricing system by early 2006. 11.8.2. Assuring an adequate reserve margin Historically, the PUCT could ensure a minimum level of planning reserves by requiring each of the larger vertically integrated utilities to maintain 15% reserve margins. It has proven difficult to enforce generation adequacy standards under the new market structure. When some forecasts in early 2005 indicated the possibility that ERCOT’s planning reserve margin could drop below 12.5% in the Summer of 2005 following an announcement by Texas Genco that it planned to retire or mothball 3500 MW of old natural gas plants, various assumptions utilized in reserve margin calculation were re-examined.68 To ensure resource adequacy, the Commission needs to take into consideration the design of the market as well as the appropriate mitigation measures that prevents market power abuses while allowing true scarcity pricing that results from normal forces of competition. Staff recently proposed to develop a strawman regarding an energy-only resource adequacy mechanism.69 The Commission agreed with Staff and a copy of strawman was later distributed to stakeholders on August 19, 2005 for comments.70 The Commission is aware of risk factors that may be associated with such a mechanism, including the possibility of prices to spike well above currently imposed $1000 offer cap. In addition, a backstop measure is included in the proposal to prevent unintended consequences that may result in possible capacity shortages. 11.8.3. Sufficient demand response Ensuring resource adequacy and mitigating market power in a nodal market will prove much easier if sufficient demand response is present. The importance of this issue is recognized and the Commission is committed to taking necessary steps to ensure demandresponse effectiveness.71 While some progress has been made in this area, an enormous untapped potential for demand-side resources remains unutilized.

11.9. The Final Words After 10 years under the restructured wholesale market and more than 3 years of experience with retail competition in Texas, it is far too early to determine whether the new market structure will remain viable in the long term. Indeed the market is still operating under a variety of transition mechanisms, including the PTB retail price caps for consumers with a billing demand below 1 MW and offer caps at the wholesale level. Further, some of the key 68

For more information please visit PUCT Power to choose at http.//www.powertochoose.org Project No. 24255, Staff Memo to the Commission for the July 15, 2005 Open Meeting. PUCT, July 8, 2005, Austin. 70 See Schubert, E. (2005a). 71 See Schubert, E. (2005b). 69


415

features of the market structure are presently being re-examined, including educating consumers about their choices, managing transmission congestion in an efficient manner which minimizes subsidies and sends appropriate price signals, and ensuring long-term resource adequacy. Thus, the dust has not yet settled on some key market features. While there are many challenges to be addressed effectively within the next few years, the progress in the past 4 years indicates that the prospect for a robust competitive electricity market in Texas is bright. References Adib, Parviz (2002a). A California quake in Texas? the role of market oversight in the restructured electric industry. A Presentation to the Public Utility Law Section of the State Bar of Texas, Austin, October 11. Adib, P. (2002b). Mitigation measures for gaming opportunities in ERCOT wholesale electricity market, A Presentation to the Texas Legislative Oversight Committee, Market Oversight Division, Public Utility Commission of Texas, Austin, June 18. Adib, P. (2003). The evolving ERCOT market design: what should be expected from the PUCT? A Presentation to the 2003 Texas Industrial Energy Consumers’ Annual Meeting Workshop, Houston, June 11. Adib, P. (2004). Monitoring monopolistic practices: market oversight. A Presentation to the Members of the Superintendency of Electricity of the Dominican Republic, Santo Domingo, July 8. Adib, P. and Clark, C. (1996). Regulatory reform, Texas style: the electric and telecommunications industries. Texas Business Review, Bureau of Business Research, The University of Texas at Austin, August. Baldick, R. and Niu, H. (2005). Lessons learned: the Texas experience. In James Griffin and Steven Puller (eds.), Electricity Deregulation: Where to from here? The University of Chicago Press. Available: http://www.ece.utexas.edu/~baldick/papers/papers.html Center for Advancement of Energy Markets, Retail Energy Deregulation Index 2003, 4th Edition. ERCOT (2001), A guide to how the Electric Reliability Council of Texas (ERCOT) facilitates the competitive power market, Version 1.2. ERCOT (2003), Protocols, www.ercot.com ERCOT (2005a). Report on Existing and Potential Electric System Constraints and Needs, Austin, October 1, at http://www.ercot.com/NewsRoom/MediaBank/ERCOT2005ReportOnConstraintsAndNeeds 10102005.pdf ERCOT (2005b). Operations update: revised 2005–2010 ERCOT regional load forecast and reserve margin update utilizing the ERCOT econometric load forecast. A Presentation to the ERCOT Board of Directors, Austin, June 21. Federal Energy Regulatory Commission (2002). Docket No. RM01-12-000: Remedying Undue Discrimination through Open Access Transmission Service and Standard Electricity Market Design. Harvey, Scott and Hogan, W. (2000a). Nodal and Zonal Congestion Management and the Exercise of Market Power, January 10 at www.whogan.com Harvey, Scott and Hogan, W. (2000b). Nodal and Zonal Congestion Management and the Exercise of Market Power: Further Comment, February 11 at www.whogan.com Hogan, W. (1998). Competitive Electricity Market Design: A Wholesale Primer. Center for Business and Government, John F. Kennedy School of Government, Harvard University, Cambridge, Massachusetts, December 17 at www.whogan.com Hogan, W. (2005). Electricity Market Restructuring: Successful Market Design, A Presentation to the National Association of Regulatory Utility Commissioners’ Electricity Committee, Austin, July 26 at www.whogan.com Hudson, P. (2005). A perspective on Texas’ efforts. Presentation at the 65th Annual Texas Electric Cooperatives Meeting, August 2, Austin, Texas, p.23. Potomac Economics (2005). Investigation into the Causes for the Shortages of Energy in the ERCOT Balancing Energy Market and into the Wholesale Market Activities of TXU from October 27 to December 8, 2004, April. PUCT, Electricity Choice: Texas is Different from California, Undated. PUCT (2001a). Docket No. 25755: PUC Investigation into Overscheduling in ERCOT in August 2001.

416


PUCT (2001b). Power to Choose: www.powertochoose.org PUCT (2002). Market Oversight Division, Comments on Issues Related to the Transmission Congestion Workshop on September 18, 2002. Project No. 26376, Transmission Congestion Issues In The Electric Reliability Council Of Texas, September 9, Austin. PUCT (2003). Market Oversight Division, Analysis of Balancing Energy Price Spikes during the Extreme Weather Event of February 24–26. PUCT (2003). Market Oversight Division, Comparison of Market Designs, Project No. 26376: Rulemaking Proceeding on Wholesale Market Design Issues in the Electric Reliability Council of Texas. PUCT (2003). Scope of competition in electric markets in Texas. Report to the 78th Texas Legislature. PUCT (2004). Market Oversight Division. Staff Inquiry into Allegations Made by Texas Commercial Energy regarding ERCOT Market Manipulation. Project No. 25937, January 28, Austin. Public Utility Regulatory Act (Texas) Various Sections. Senate Bill 7, Chapter 39, Subchapter A. Rose, K. and Meeusun, K. (2005). 2005 Performance Review of Electric Markets, August 23. Ruff, Larry (2004). A Transitional Non-LMP Market for California: Issues and Recommendations. A Report Prepared by Charles River Associates for California ISO, September 16. (www.caiso.com/docs/ 2004/09/17/2004091712202513134.pdf) Saathoff, K. (2003). ERCOT Staff, Operations Update, Presentation to ERCOT’s Technical Advisory Committee, March 6. Schrader, T. F. (2005). ERCOT CEO Report. A Presentation to the ERCOT Board of Directors, Austin, January 18. Schubert, E. (2004). ERCOT Zonal Market in Theory and Practice. A Presentation under Project No. 28500, Public Utility Commission of Texas, Austin, December 10. Schubert, E. (2005a). Commission Staff’s Resource Adequacy Strawman. Project No. 24255, Rulemaking Concerning Planning Reserve Margin Requirements, Public Utility Commission of Texas, Austin, August 19. Schubert, E. (2005b). Staff White Paper on an Energy-Only Resource Adequacy Mechanism. Project No. 24255, Rulemaking Concerning Planning Reserve Margin Requirements, Public Utility Commission of Texas, Austin, April 14. Synchronous Interconnection Committee (1999). Report to the 76th Texas Legislature Feasibility Investigation for AC Interconnection between ERCOT and SPP/SERC, Austin. Wolak, F. et al. (2004). Alternatives to Implementing a Locational Marginal Pricing Market. Market Surveillance Committee of the California ISO, November 10 at www.caiso.com/docs/2000/09/14/ 200009141610025714.html Zarnikau, J. and Reilly, R. (1996). “The evolution of the cogeneration market in Texas.” Energy Policy, Spring. Zarnikau, J. and Lam, A. (1998). “Market Power and Market Concentration in Texas (ERCOT)”, Electricity Journal. Fall. Zarnikau, J. and Adib, P. (2003). Incentive Compatibility in the Newly Developed Competitive Markets and Potential for Unexpected Outcomes: Electric Industry and Market Oversight. A Presentation to the Students of LBJ School of Public Affairs, the University of Texas at Austin, November 11. Zarnikau, J. and Whitworth, D. (2005). “Has Electric Utility Restructuring Led to Lower Electricity Prices for Residential Consumers in Texas?” Energy Policy. Zarnikau, J., Landreth, G., Hallet, I. and Kumbhakar, S. (2005). “Industrial Energy Consumers Response to Wholesale Prices in the Restructured Texas Electricity Market.” (forthcoming).


417

Appendix Acronyms used in this chapter A/S AREP BES CPL CR CSC ERCOT FERC IOU IPP ISO LaaR LMP LSE MCPC MCPE MOD NSRS OOM PGC PM POLR PUCT PURA PURPA PTB QSE REC REP RPRS RRS RTO TCR TDSP TNMP WTU

Ancillary Services Affiliated Retail Electric Provider (a REP which is affiliated with the incumbent verticallyintegrated utility) Balancing Energy Service Central Power and Light Competitive Retail Electric Provider Commercially-Significant transmission Constraint Electric Reliability Council of Texas Federal Energy Regulatory Commission Investor-Owned Utility Independent Power Producer Independent System Operator Load Acting as a Resource (an interruptible or curtailable load which provides an ancillary service) Locational Marginal Pricing Load-Serving Entity Market Clearing Price of Capacity (used for many A/S prices) Market Clearing Price of Energy (used for BES prices) PUCT Market Oversight Division which was later renamed as Wholesale Market Oversight (WMO) Non-Spinning Reserves (an A/S) Out-of-Merit (a resource dispatch instruction used in the management of transmission congestion) Power Generation Company Power Marketer Provider of Last Resort Public Utility Commission of Texas Public Utility Regulatory Act (Texas) Public Utility Regulatory Policy Act (US) Price-to-Beat (a default retail electricity price applicable to customers with a billing demand less than 1 MW who fail to exercise customer choice) Qualified Scheduling Entity Renewable Energy Credit Retail Electric Provider Replacement Capacity (an A/S) Responsive Reserve Service (an A/S) Regional Transmission Organization Transmission Congestion Right Transmission and Distribution Services Provider Texas-New Mexico Power Company West Texas Utilities


Chapter 12 Electricity Restructuring in Canada* MICHAEL J. TREBILCOCK1 AND ROY HRAB2 1 Faculty of Law, University of Toronto, Toronto, Ontario, Canada; 2Ontario Energy Board, Toronto, Ontario, Canada

Limited electricity restructuring has occurred in Canada. This chapter focuses on the electricity sector restructuring initiatives of Canada’s most populous province, Ontario, examining the province’s restructuring process and experiences, the aftermath of re-regulation and the lessons learned. It also reviews briefly the restructuring experience in the province of Alberta. Both provinces altered their restructuring plans after experiencing unexpected price increases. In particular, the Ontario experience illustrates the importance of political commitment and how restructuring policies can be reversed quickly when a government fears a political backlash. 12.1. Introduction Most of Canada’s electricity utilities are government-owned corporations, and in many cases, vertically integrated. However, some restructuring has occurred in recent years. By April 2005, two provinces, Alberta and Ontario, had open wholesale markets and retail choice available to all consumers while five provinces (British Columbia, Saskatchewan, Manitoba, Québec and New Brunswick) had open wholesale markets with some permitting retail access for large industrial users and introducing limited competition in electricity generation. 1 However, there is little political will in the country to undertake further significant market and industry restructuring. The lack of will is in no small part caused by the problems encountered in the restructuring initiatives in Alberta, Ontario and California coupled with a long history of government ownership and provision of low-priced electricity (in many cases low-priced hydro-electric generation) (Fig. 12.1). This chapter reviews the electricity sector restructuring initiatives of Canada’s most populous province, Ontario, examining the province’s restructuring process and experiences,

* This is an updated and amplified version of an article published in the Energy Journal (Trebilcock and Hrab, 2005). The views expressed here are those of the authors and do not necessarily reflect those of the Ontario Energy Board. The authors would like to thank Joseph Doucet for providing insightful comments about Alberta’s restructuring initiative. 1 See National Energy Board (2005).

419

420


Fig. 12.1. Canada electricity generation by fuel, 2003. Source: Statistics Canada.

the aftermath of re-regulation and the lessons learned. It also reviews briefly the experience in the province of Alberta. The case of Ontario illustrates the importance of political commitment and how restructuring policies can be reversed quickly when a government fears a political backlash. 12.2. The Setting in Ontario2 Beginning in the mid-1990s, the Ontario government began to implement a number of restructuring initiatives aimed at dismantling a vertically integrated government-owned monopoly, developing a competitive generation sector and liberalizing prices. However, unexpected and extreme price increases following market opening in 2002 led the government to alter its restructuring plans significantly because of consumer (voter) resistance to price increases. The inability of the provincial government to overcome political obstacles and follow-through effectively with its initial restructuring plan created disarray and uncertainty in the province’s electricity sector such that the province, under a new government, is still struggling to bring under control (Table 12.1). With a population of 12.4 million in 2004, Ontario is Canada’s most populous province, accounting for 39% of the country’s total population. The province has approximately 4.4 million electricity consumers served by about 90 local distribution companies (LDCs), using more than 152 million MWh of electricity a year. At the end of 2004, this demand was served by the province’s installed generation capacity of 31,164 MW comprised of nuclear power (35%), hydroelectric (25%), coal (24%) and oil and gas (16%). There was minimal other

2

For up-to-date information and data on the Ontario market consult the Ontario Ministry of Energy (www.energy.gov.on.ca), the OEB (www.oeb.gov.on.ca), the IESO (www.ieso.ca), the OPA (www. powerauthority.on.ca), and the OEFC (www.oefc.on.ca).

Electricity restructuring in Canada

421

Table 12.1. Electricity sector restructuring milestones in Ontario. Event

Date

Comment

Provincial election MacDonald Report Government White Paper

1995 1996 1997

Electricity Act

1998

Ontario Hydro split Market opening delayed

1999 June 2000

Market opening date set Hydro One privatization blocked

April 2001 April 2002

Market opens Electricity Pricing, Conservation and Supply Act

May 2002 December 2002

Provincial election

October 2003

Pickering “A” Review Panel Report

December 2003

Electricity Conservation And Supply Task Force Report OPG Review Committee Report Interim price plan

January 2004

Progressive Conservative Party elected. Recommends a more market-based electricity sector. Proposes full wholesale and retail competition by 2000 and unbundling of Ontario Hydro. Legislation sets out framework for restructured market. Creation of OPG and Hydro One. Market opening (scheduled for November 2000), delayed for at least 6 months to ensure all market participants are prepared. Market opening scheduled for May 2002. Union challenge results in privatization of Hydro One, transmission grid owner, being blocked by court ruling. Retail and wholesale markets open. Following high prices during unexpectedly hot summer government freezes retail rates. Legislation freezes retail prices at 4.3 cents/kWh for low volume and other designated customers. All three main political parties pledge to maintain government ownership of electricity assets. Liberal Party wins election. In addition to maintaining public ownership, election promises include: maintaining 4.3 cent/kWh price freeze until 2006, retiring all coal-fired generation by 2007 and installing “smart” meters for all consumers. Report finds significant defects in project management and accountability for Pickering A Unit 4 restart resulted in significant delays and cost-overruns. Recommendations include creation of “conservation culture” and decreasing reliance on the spot market.

March 2004 April 2004

Electricity Restructuring Act OPG rates regulated

December 2004 February 2005

Regulated Price Plan Coal retirement delayed

April 2005 June 2005

Recommendations include maintaining government ownership of OPG and restarting Pickering A Unit 1. 4.3 cent/kWh price cap replaced with 4.7 cents/kWh for the first 750 kWh consumed per month, and 5.5 cents/kWh for consumption above that level. Regulatory framework of electricity sector restructured. OPA created. Provincial government announces that OPG baseload generation prices will be fixed and that temporary revenue limits will be placed on most of OPG non-baseload generation. OEB’s regulated retail prices in effect. Retirement deadline for all coal-fired generation delayed until 2009.

422


renewable installed capacity (e.g., wind). Between May 2004 and April 2005, coal-fired generation set the wholesale price 49% of the time, at an average price of 4.48 cents/kWh (all units in Canadian cents/dollars); hydroelectric set the price 35% of the time at 4.84 cents/kWh; and, oil and gas set the price 16% of the time at 7.42 cents/kWh.3 More than 20 companies own and operate electricity generators connected to the province’s transmission grid. The largest firm is Ontario Power Generation (OPG), which is wholly owned by the provincial government. It controls about 70% of the province’s installed capacity.

Fig. 12.2. Transmission map of Ontario. Source: Hydro One, Ontario Power Generation. 3

IESO (2005a).


423

The province relies on imports to meet its electricity needs at peak times and has the capacity to import approximately 4000 MW from two neighboring Canadian provinces (Manitoba and Quebec) and three US states (Michigan, Minnesota and New York). The province’s electricity grid is made up of about 29,000 kilometers of transmission lines, which is almost completely owned by Hydro One, a firm also wholly owned by the provincial government (Fig. 12.2). 12.3. Pre-2004 Restructuring Prior to reform, Ontario Hydro, a vertically integrated, government-owned monopoly, was responsible for meeting the province of Ontario’s electricity generation and transmission needs.4 The power produced by Ontario Hydro was purchased and distributed by about 300 local, municipally owned utility companies to consumers, who were charged a fixed price per kWh. The price bundled together generation, transmission and distribution costs. In 1999, Ontario Hydro had a provincially guaranteed debt of approximately $38.1 billion5, or about a third of total provincial indebtedness. Through the 1990s, roughly 35% of the utility’s electricity revenue went towards debt interest.6 Much of this debt was the result of overexpansion and major cost-overruns in the construction of nuclear generation facilities.7 These problems contributed to a rise in the retail price of electricity in Ontario of about 30% in the early 1990s.8 However, in 1993, Ontario Hydro’s retail prices were frozen by the provincial government. The freeze remained in place until retail and wholesale markets opened in May 2002. The poor performance of Ontario Hydro and a philosophical preference for reducing the size of government led the newly elected (June 1995) Progressive Conservative provincial government to appoint an Advisory Committee in 1995 to explore the possibilities for reforming the province’s electricity sector. In 1996, the Committee’s report9, known as the MacDonald Report, recommended a more market-based electricity sector. Subsequently, the government released a White Paper in 199710, proposing full wholesale and retail competition by 2000 and the division of Ontario Hydro into its generation and transmission components. The White Paper led to the creation of the Market Design Committee (MDC) in 1998. The MDC was responsible for designing and recommending rules for wholesale and retail competition in the province’s electricity market. In 1998, the provincial government formally set out the framework for the reformed electricity market in the Electricity Act.11 In April 1999, Ontario Hydro was split into its transmission and generation components, but both corporations remained, and still remain, wholly owned by the provincial government. Each corporation is governed by a board of directors appointed by the provincial government. Hydro One was created to own and operate the transmission grid.12 OPG was created to own and operate the generation assets, 4

This chapter builds on earlier papers, Trebilcock and Hrab (2003, 2005). For other views on the Ontario experience see Dewees (2005) and Schott (2005). 5 OEFC (2000). 6 Trebilcock and Daniels (2000), p. 163. 7 Adams (2000). 8 Ontario (1997). 9 Ontario (1996). 10 Ontario (1997). 11 Electricity Act, 1998, S.O. 1998, c. 15, Sch. A. 12 The transmission firm was originally called the Ontario Hydro Services Company. The company was renamed Hydro One Inc. in 2000.

424


accounting for approximately 90% of total provincial capacity at the time. Also, between 1998 and 2002, there was a consolidation of LDCs from approximately 300 pre-restructuring to about 90. Hydro One was the primary agent of consolidation, acquiring about 90 LDCs.13 Most LDCs are owned by municipalities.14 To discourage OPG from using its dominant position in generation to exercise market power, OPG entered into a Market Power Mitigation Agreement (MPMA) with the provincial government.15 The MPMA mandated that OPG pay a rebate to consumers on 90% of its domestic sales when the wholesale price exceeded 3.8 cents/kWh. The MPMA also required OPG to divest 65% of its price-setting generating units within the first 31⁄2 years after market opening, and 65% of its core, or base-load, facilities within 10 years of market opening. In complying with the MPMA, OPG leased its Bruce nuclear power plants to private operators in May 2001 and sold four price-setting hydro-electric plants in March 2002, reducing its control of the province’s total generation capacity to between 70% and 75% at the time of market opening. Also, the government directed Hydro One to make good faith efforts to increase inter-tie capacity with neighboring jurisdictions in Canada and the US by 50%, roughly 2000 MW, within 3 years of market opening. Two agencies were mandated to oversee the electricity market: the pre-existing Ontario Energy Board (OEB) and the newly created Independent Electricity Market Operator (IMO, renamed the Independent Electricity System Operator (IESO) in 2004). At the time of restructuring, the primary purpose of the OEB was to regulate the monopoly segments of the electricity market by setting transmission and distribution rates.16 The IMO’s responsibilities included operating the wholesale spot market and dispatch function; its independent Market Surveillance Panel (transferred to the OEB in 2004) monitors market power abuses. At market opening, the province unbundled electricity prices into separate components: transmission, energy and distribution charges. The province also assessed consumers a 0.7 cent/kWh debt retirement charge reflecting the determination that at the time of deintegration in 1999 Ontario Hydro had $19.4 billion of “stranded debt,” that is, debt that could not be serviced and retired in a competitive electricity market.17 At the end of March 2002, the value of the stranded debt was $20.1 billion.18

12.3.1. The market opens In April 2002, the month before market opening, a 10-year outlook (2003–2012) prepared by the IMO stated that “based on existing and proposed facilities, Ontario is expected to have a reliable supply of electricity for the 10-year period under a wide variety of conditions.”19

13

Ontario (2004a). Distributors’ Electricity Efficiency Policy Group (2003). 15 Trebilcock and Daniels (2000); Goulding et al., (2001). 16 The OEB also licenses all electricity market participants including generators, transmitters, distributors, wholesalers, retailers and the IMO, and is required to approve amalgamations, mergers, acquisitions and divestitures of distributors, as well as transmission-line construction. The OEB also regulates Ontario’s natural gas sector. 17 OEFC (2000). 18 OEFC (2002). 19 IMO (2002a), p. 43. 14


425

During the same month, the province’s attempt to privatize Hydro One through an IPO was blocked by a court challenge by two unions opposed to privatization. The unions argued successfully that the Electricity Act did not authorize the provincial government to sell the company’s assets.20 Despite the setback, both wholesale and retail price competition went forward on May 1, 2002. In the wholesale market, operated by the IMO, the marginal supplier set electricity spot prices every 5 minutes in response to changing levels of demand and supply. Participation in the wholesale market was voluntary; wholesale consumers could enter directly into bilateral physical or financial contracts with wholesale sellers and generators. Retail market consumers were free to enter into fixed-price contracts with retail intermediaries. Retail intermediaries purchased electricity from the wholesale market and resold it to consumers. Almost 1 million of the province’s estimated 4.4 million electricity customers entered into fixed-price contracts with retail intermediaries. Consumers not establishing a relationship with a retailer purchased electricity through their local distribution utility, which passed through the wholesale price into retail rates each billing period (i.e., there was no regulated fixed-price default option). Prices in the newly opened market increased rapidly over an abnormally hot summer (Fig. 12.3). The highest hourly price recorded was $1.03/kWh on September 3. The dramatic increase in prices reflected an environment of high demand and tight supply during the summer (further described below). From July–September, the IMO issued both power

100.00 90.00

Price ($C/MWh)

80.00 70.00 60.00 50.00 40.00 30.00 20.00 10.00 M ay 20 Au 02 gu st N ov 20 em 02 be r2 Fe 00 br ua 2 ry 20 03 M ay 20 Au 03 gu st N ov 20 em 03 be r2 Fe 00 br ua 3 ry 20 04 M ay 20 Au 04 gu st N ov 20 em 04 be r Fe 20 br 04 ua ry 20 05 M ay 20 05

0.00

Fig. 12.3. Weighted average wholesale electricity prices in Ontario, May 2002–June 2005. Source: IESO

20

Interestingly, another union, the Power Workers’ Union supported the privatization effort, arguing that immediate investment was required to modernize the transmission grid. Following the decision of the court, the government altered to the Electricity Act to permit privatization, but postponed and later abandoned plans to privatize the transmission grid.

426


warnings and power advisories, requesting consumers to reduce usage because power supplies were under strain. In October 2002, the IMO reported a supply situation significantly less optimistic than its April 2002 forecast, stating that “[t]here is a serious shortage of generation capacity to meet Ontario’s growing demand for electricity. If steps are not taken to address this situation, Ontario could face even more serious reliability problems next summer, leading to the possibility of supply interruptions and continued upward pressure on prices during periods of peak demand.”21

12.3.2. The retail price freeze In response to mounting criticism of the high summer electricity prices from consumers, on November 11, 2002, the provincial government announced that it would freeze retail prices. On December 9, 2002, the Electricity Pricing, Conservation and Supply Act, 2002 became law.22 The legislation severed the link between wholesale and retail prices by freezing the retail price of electricity for low-volume consumers using less than 150,000 kWh/year (e.g., households, small businesses and farmers), and other designated consumers23 at 4.3 cents/kWh, and included consumers who had signed fixed-price contracts with retailers. The freeze was estimated to affect about half of the electricity consumed in the province. The frozen retail rate was made retroactive to market opening, refunding any amount over 4.3 cents that consumers had already paid. In March 2003, the province extended the frozen retail price to consumers using less than 250,000 kWh/year, covering approximately an additional 7000 consumers. The weighted average wholesale price for the first year of the open market was 6.2 cents/kWh,24 44% higher than the frozen retail price. The provincial government announced that the retail price freeze was to last until at least 2006 and “continue until there is a sufficient electricity supply, at reasonable prices, to meet Ontario’s long-term needs.”25 The provincial government would pay suppliers the difference between wholesale prices and the frozen retail price. The government took a number of other actions, including placing limits on all energy rates, including transmission, distribution, wholesale market, uplift and customer charges. Further, the energy minister was placed in charge of overseeing the creation of market rules and approving changes to transmission and distribution rates. At this point, it is important to make two observations. First, the IMO examined the Ontario market to determine whether generators had abused market power during the summer of 2002. After examining almost all high-priced hours – hours where the price exceeded $200/MWh – it found no evidence of abuse of market power.26 Similar analysis of the September 2002–January 2003 period also found no evidence of abuse.27 Second, while the price of electricity in Ontario during the summer of 2002 increased significantly, Ontario

21

IMO (2002b), p. 2. Electricity Pricing, Conservation and Supply Act, 2002, S.O. 2002, c. 23. 23 Other designated consumers are municipalities, universities and colleges, public and private schools, hospitals and registered charities. 24 IMO (2003a). 25 Ontario (2002). 26 IMO (2003b). 27 In report released between May 2002 and May 2005, the MSP found no evidence of abuse of market power in the Ontario market.

22

427


prices did not vary significantly from those of neighboring jurisdictions. In fact, Ontario electricity prices almost never exceeded those of neighboring jurisdictions (Figs. 12.4 and 12.5).

12.4. Implementation Problems Why was the Ontario experiment so short lived? Some argue that Ontario’s problems were the inevitable result of privatization and de-regulation.28 However, a number of problems were evident with Ontario’s restructuring attempt. A reduction in domestic generation capacity, an increasing reliance on imports, limited import capacity and extreme temperatures all helped drive prices higher. With the exception of the weather, most of these developments did not suddenly emerge in the summer of 2002; many were apparent during the years leading up to the market opening. During the summer of 2002, the demand for electricity exceeded available capacity at peak times (Fig. 12.6). In October 2002, the IMO reported that, “the percentage by which total available capacity exceeds the summer peak demand for energy has fallen from 19.2% in 1996 to ⫺ 1.5% in 2002.”29 The increased summer demand – the peak demand was 25,414 MW on August 13, 2002 – and diminished hydroelectric capacity were primarily caused by above-average summer temperatures. There was also a longer than expected outrage of a reactor at the Bruce Nuclear Power Station-B during the summer of 2002. The high temperatures

70.00 60.00

Price ($C/MWh)

50.00 40.00 30.00 20.00 10.00

New England ISO

2 r2

D

ec

em

be em ov N

be

r2

00

2 00

02 er

ct ob O

te

m be

r2

20

00

2

02

NY ISO Zone OH

Se p

st Au gu

ly Ju

ne Ju

IMO

20

20

02 20

02 20 M ay

02

0.00

PJM Western Hub

Fig. 12.4. Average hourly Ontario prices relative to neighboring control areas, off peak, 2002. Source: IESO. 28 29

For example, see Hampton (2003). IMO (2002c), p. 132.

428


120.00

100.00

Price ($C/MWh)

80.00

60.00

40.00

20.00

IMO

NY ISO Zone OH

2

2

00

00

r2 be em D

N

New England ISO

ec

ov

O

ct

em

ob

be

er

r2

20

00 r2

Se

pt

em

be

st gu Au

02

2

02 20

02 20 Ju

ly

ne Ju

M

ay

20

20

02

02

0.00

PJM Western Hub

Fig. 12.5. Average hourly Ontario prices relative to neighboring control areas, on peak, 2002. Source: IESO.

30,000

Megawatts

25,000 20,000 15,000 10,000 5000 0

1996

1997

1998

1999

Peak demand

2000

2001

2002

2003

Available capacity

Fig. 12.6. Peak summer demand versus available capacity in Ontario, 1996–2003. Source: IESO.

429

Electricity restructuring in Canada 4000 3500

Megawatts

3000 2500 2000 1500 1000 500 0

1996

1997

1998

1999

2000

2001

2002

2003

Fig. 12.7. Net imports for summer peak hours in Ontario, 1996–2003. Source: IESO.

increased demand for electricity for air conditioners and reduced the amount of water available for hydro-electric generation. From 1984 to 2001, the average annual growth of primary energy demand in Ontario was 1.6%30; the first year of the restructured Ontario market experienced a demand increase of 5.5%.31 Other contributing factors were actions by the provincial government and Ontario Hydro over the decade leading up to market opening. First, in 1993, in an effort to stabilize Ontario Hydro’s financial outlook, the provincial government directed the cancellation of a number of planned and in-progress generation projects.32 Second, a number of nuclear plants were taken offline for reliability and safety reasons. Between 1995 and 1998, the four units of the 3300 MW Bruce Nuclear Power Station-A (Bruce A) were taken offline. In 1997, the four units of the 2060 MW Pickering Nuclear Power Station-A (Pickering A) were removed from service. In late 1998, Ontario Hydro announced plans to initiate the restart of the four Pickering A units. The restart experienced numerous delays and substantial cost overruns. It was expected that the first unit would be restarted by June 2000 for use during the winter of 2000/2001; however, the first unit did not return to commercial service until the end of September 2003. Had the Pickering restart been completed on schedule, the price increases and volatility experienced during the summer of 2002 would have been partially mitigated, although it should be recognized that the Pickering delays were known prior to market opening. The lack of available domestic generation capacity has required Ontario to import electricity to balance supply and demand during summer demand peaks since 1997 (Fig. 12.7). Thus, increasing inter-tie capacity prior to market opening would have partially mitigated

30

IMO (2002d). IMO (2003a). 32 Hampton (2003), pp. 139–52. 31

430


price increases. The IMO made 38 emergency import purchases during the summer of 2002 to maintain system reliability.33 The large amount of imports strained transmission inter-tie capacity with other jurisdictions. The province’s inter-ties with Manitoba, Quebec, New York, Minnesota and Michigan all experienced varying degrees of congestion during the summer of 2002.34 The province was importing the maximum amount of electricity – roughly 4000 MW – that the transmission system could physically accommodate. The peak amount scheduled for import was 4273 MW on September 20, 2002.35 Additionally, by 2002, Hydro One expected phase-shifting transformers at the Michigan inter-tie, which would increase import capacity by 500 MW, to become operational. However, the transformers experienced significant technical difficulties and were not available for the summer of 2002. 12.4.1. Market structure: supply With the reduction of domestic capacity and increasing reliance on imports it would seem that profitable opportunities for private-sector investment in generation existed in Ontario leading up to market opening. And yet, only two new private-generation projects amounting to 620 MW became operational during the first year of the restructured market. The reasons for this limited private-sector interest revolve around market design problems and raise questions about the provincial government’s commitment to the restructuring initiative. First, industry participants cited the delay and uncertainty regarding market opening as a significant factor contributing to the failure of the province to attract private investment. The Ontario market was originally scheduled to open in November 2000, but market opening was delayed to May 2001 and later May 2002. Market opening was delayed to ensure system reliability and to allow thorough testing of the hardware and software acquired by the IMO, wholesale market participants, service providers and retailers to implement the wholesale and retail market design. The delay was costly because investors lost confidence in the electricity sector following the California crisis during 2000–2001 and the collapse of Enron during 2001–2002. As a result, investors who may have invested in generation capacity in 2000 came to view the North American electricity sector as too risky, and were no longer interested or able to raise sufficient capital for new generation capacity when the Ontario market opened. Additionally, the multiple delays of the market opening date, regardless of the reasons, created uncertainties as to the government’s commitment to restructuring. Conditions within Ontario prior to the California crisis also contributed to the lack of private investment. During 1998 and 1999, the private sector expressed a reluctance to invest in Ontario’s electricity industry because of continued OPG ownership and control of generation assets and the prolonged decontrol timetable set out in the MPMA. The provincial government did little to allay investor concern regarding OPG’s dominance of the sector. This reluctance may have been influenced by the provincial government’s desire to avoid controversy because of the prospect of an election that ultimately occurred in June 1999 (won by the incumbent Progressive Conservative Party). And yet, even though the Progressive Conservatives were re-elected, it should be recalled that no decontrol of OPG assets occurred until May 2001.

33

IMO (2002c), p. 48. Ibid. 35 IESO Website (www.ieso.ca) 34


431

Further, the provincial government created uncertainty regarding its commitment to reducing OPG’s dominance. For example, the announcement of the Pickering nuclear units restarts suggested continuing OPG dominance in the generation sector. Further, in 2000, the province placed a temporary freeze on the sale of OPG’s coal-fired generation plants, citing the need for environmental safeguards prior to privatization. And during the summer of 2002, the province blocked the sale of two OPG coal-fired generating plants.36 Such actions did not encourage private investment and raised concerns regarding the government’s commitment to restructure the market. Another problem with the restructuring process relates to the distribution sector. About 50% of Ontario’s LDCs, on average, serve less than 5000 customers each.37 Yet estimates from Ontario indicate that the efficient scale of operation for a distribution firm is reached at 20,000–30,000 customers.38 The Ontario government could have promoted the consolidation of many of its LDCs to increase efficiency and foster a better environment for the use of forward contracts to attract more private-sector investment in generation. Clearly, if it so desired, the provincial government could have been more active in restructuring the sector prior to market opening. It could have privatized more assets and introduced competition in the generation sector to improve the environment for private investors as done in other jurisdictions, such as in Britain (Chapter 4) and the state of Victoria in Australia (Chapter 6). For example, the government could have divided OPG into a number of separate companies and privatized them. Instead, it precluded the MDC from making such a recommendation. The privatization of Hydro One could have also occurred. Instead, the provincial government did not announce its intentions to privatize the transmission grid until December 2001, less than a year before market opening, and then later abandoned the initiative. Additionally, the province could have consolidated LDCs and privatized them. Instead, most of the consolidation that did occur was driven by the acquisition of LDCs by Hydro One, the government-owned transmission monopoly. 12.4.2. Market structure: demand Competitive markets are usually defined as possessing demand and supply responsiveness. Thus, one avenue for overcoming short-term capacity constraints, as well as ensuring long-run efficiency in electricity markets, is to increase demand responsiveness. The nature of electricity supply and demand suggests that a slight decrease in demand may mitigate price spikes and preserve system reliability in a period of tight supply. For example, during its electricity crisis, California had to resort to rolling blackouts with a supply shortage of only 300 MW.39 In Ontario, and most jurisdictions, real-time metering and billing was, and, for the time being, continues to be, limited almost exclusively to large industrial and commercial consumers. During the summer of 2002, Ontario had 90 industrial consumers, comprising approximately 15% (about 3000 MW in a typical peak hour) of demand, directly connected 36 Ontario blocked sale of coal-fired plants: Industry questions commitment to creating market, National Post, (November 4, 2002), p. FP2: It was reported that a spokesman for the Minister of Energy stated that the sale price “did not meet our standard of ensuring maximum value for Ontario taxpayers and electricity consumers,” and that the buyer would not commit to converting the plants to natural gas fired facilities. 37 Ontario (2004a). 38 Yatchew (2001). 39 Hunt (2002), p. 76.

432


to the transmission grid with interval meters that measure and report hourly consumption, allowing the user to be billed at the actual hourly spot price.40 The IMO estimated that up to a further 20% of total demand was comprised of industrial consumers not directly connected to the grid but possessing interval meters.41 These consumers reduced consumption in response to the rising prices of the summer of 2002 by shifting consumption from peak to off-peak hours, maintaining system reliability. 12.4.3. Market rules The structure of the Ontario market also interacted with some market rules to further undermine the restructuring initiative. OPG’s continued ownership and control of generation assets resulted in a large proportion of electricity sold in the province being subject to the MPMA rebate, reducing incentives for consumers to enter into forward contracts with private generators. It is widely argued that through long-term bilateral contracting, companies seeking to enter the generation market can secure a future revenue stream, mitigating sunk-cost problems and enabling firms to obtain long-term project financing.42 The lack of interest in contracts was evident from the fact that about 60–70% of electricity was purchased in the Ontario spot market during the summer of 2002. Some industry participants argued that market rules discourage investment in the province because the price of imported electricity does not set the market clearing price, although this was likely not a major issue as other jurisdictions (e.g., Alberta) have similar rules and significant generation entry. For technical reasons, accepted imports are scheduled 1 hour in advance of delivery, and cannot be dispatched on a 5-minute basis as domestic generators can. Import prices are not used to calculate Ontario’s wholesale price. Instead, if an import is accepted, the importer is guaranteed the offer price in cases where the Ontario market clearing price is below it. The guaranteed payment system was implemented to improve reliability. However, when Ontario demand is very high the guaranteed payment can create situations where it is more profitable to sell electricity to the Ontario market from outside than inside the province. For example, on one occasion in July 2002, out-of-province generators were receiving $2/kWh for electricity while Ontario generators were receiving 47 cents/kWh. 12.4.4. Consumer expectations and information Despite the reduction in domestic capacity, Pickering restart delays, an increase in imports leading up to market opening and a decade of frozen prices, some politicians and industry participants told consumers the province would have a surplus of generating capacity and there would be no price spikes following market opening. Indeed, some proponents of market restructuring told consumers to expect price reductions. With such expectations, it is not surprising that consumers were surprised and outraged at the run-up in electricity prices during the summer of 2002. This outrage was vented on the Progressive Conservative provincial government, which had just elected a new leader who was expected to call an election in the near future. These factors put intense pressure on the government to intervene in the market, but there were other contributing factors. 40

IMO (2002c), p. 17. Ibid., pp. 17–18. 42 See White (1996). 41


433

Many residential consumers were not well informed about the OPG rebate mechanism contained in the MPMA. If they had been more fully aware that they would be receiving a rebate from OPG at the end of the year, it is likely there would have been less public anger over rising prices. The IMO estimated that the OPG rebate would have reduced the average electricity charge from 6.2 cents/kWh to approximately 5 cents/kWh for the first year of the open market.43 Further, consumers lacked information about how to compare their prior bundled electricity charges with the unbundled rates following de-regulation. Some consumers signed retail contracts before the market opened to purchase unbundled electricity at nearly 6 cents/kWh, mistaking the energy price for the total bundled rate. The electricity charge accounts for one-half to two-thirds of the total price of electricity delivered to a residential customer.44 This resulted in many consumers receiving electricity bills that were much more expensive than their expectations. This outcome resulted in allegations by some consumers and politicians that some electricity retailers used deceptive marketing practices. In retrospect, the government could have made a stronger and more compelling case for de-regulation and restructuring, as well as better preparing consumers for restructuring. First, the province could have clearly stated that electricity prices in Ontario had been capped at a level below cost for years, emphasizing that the initial gains from restructuring would come from reduced debt charges through the sale of electricity assets, and that over the long term, competition would control costs better than the Ontario Hydro monopoly, resulting in either lower prices or lower price increases. Second, the province could have accelerated the rebate payments to occur on a quarterly or monthly basis, mitigating consumer concerns about higher prices, reducing political pressure to freeze rates. On March 21, 2003, the province belatedly announced that consumers using more than 250,000 kWh/year would receive OPG rebates quarterly, rather than annually, with the rebate fixed at 50% of the value by which the average wholesale price exceeded $38/MWh. Alternatively, the province could have taken an incremental approach to price de-regulation, as done in other jurisdictions, such as Victoria and England & Wales, by first opening the wholesale market and phasing in deregulated retail prices by size of the user, beginning with the largest. Last, the province could have unbundled electricity charges prior to market opening, enabling consumers to make more informed comparisons between offers from competitive retailers and the standard supply offer of the LDCs.

12.5. Results of the Price Freeze The three primary results of the province’s 2002 retail price freeze and retreat from restructuring were: an elimination of incentives to reduce electricity consumption; the creation of substantial financial burdens on the province; and a lack of incentives for investment in new generation and transmission by the private sector. 12.5.1. Unrestrained consumption The retail price freeze exacerbated the supply demand imbalance in Ontario’s electricity market. Previously, increasing retail prices indicated that supply was tight relative to demand, 43 44

IMO (2003a), p. 4. National Energy Board (2001), p. 35.

434


signaling consumers to reduce their use or face an increased electricity bill. The frozen retail prices eliminated consumers’ incentives to limit demand in times of tight supply or shift consumption to off-peak periods, with the result that demand became unresponsive to changes in the wholesale price. Thus, one result of the retail price freeze was higher wholesale prices than if there had been no retail freeze, resulting in no reduction in demand and exacerbating the supply demand imbalance that the province was experiencing. 12.5.2. The government’s financial exposure The retail price freeze and its impact on consumption, coupled with market-based wholesale prices exposed the provincial government to significant financial commitments. The Ontario decision to make taxpayers responsible for paying the difference between wholesale and retail prices can be contrasted to California’s decision (Chapter 10) to make the shareholders of electricity utilities pay for the difference. The freeze was financed by OPG’s fund for rebates of revenues in excess of its 3.8-cent wholesale price cap and by the Ontario Electricity Financial Corporation (OEFC). The OEFC is a government agency created by the provincial government to manage the provincially guaranteed debt and other legacy liabilities of the former Ontario Hydro. At the time, the Ministry of Energy stated that the price freeze would be “revenue neutral,” claiming the program would “pay for itself” when wholesale prices fell below the frozen retail price. This did not occur. During the first year of the price freeze, the OEFC was required to finance approximately $730 million of the costs of the freeze. 12.5.3. Discouraging private investment The retail price freeze chilled plans by private entrants into the Ontario market. New privategeneration plans were delayed or cancelled following the initial freezing of retail prices in 2002. For example, in November 2002, Sithe Energies Inc. suspended plans to construct two power plants with a combined capacity of nearly 1700 MW. The plants had obtained the necessary regulatory approvals and could have been online by 2005. The company’s reasons for delaying construction included OPG’s market dominance and the potential for government policy changes. In March 2003, the IMO reported that only about 2200 MW of approximately 8800 MW of planned generation was under construction.45 Further, a planned 975 MW merchant transmission line with Pennsylvania through Lake Erie was halted in November 2002 because of the financial uncertainty surrounding such companies following the Enron scandal and policy reversals in Ontario.46 The retail price freeze created incentives for the provincial government to prevent further OPG decontrol. Under the MPMA, decontrol by OPG would decrease the amount of rebates that OPG would have been required to pay, increasing the government’s direct financial exposure to the retail price freeze. 45

IMO (2003b). Merchant transmission is defined by the Harvard Electricity Policy Group as: “Commercial transmission investments made in response to market-based incentives. The return on investment depends on a combination of sales of transmission rights or profits from locational arbitrage of energy prices. The investment does not add to a regulated rate base or qualify for a regulatory recovery mechanism. The full market risk and reward accrue to the transmission investors.” See http://www.ksg.harvard.edu/ hepg/Merchant_transmission.htm 46


435

12.6. A New Vision The policy reversals and mixed signals have put tremendous pressures on the province to increase its installed generation capacity, its ability to import power and encourage conservation, resulting in a multitude of new policies directed at the electricity market being developed. For the summer of 2003, the province contracted with private generators to provide extra power at an estimated cost of $70 million. The estimated costs of potential larger-scale generation projects are substantial. In June 2003, the provincial governments of Ontario and Manitoba announced that they would investigate the possibility of jointly financing the construction of a $5 billion, 1250 MW hydro-electric station located in Manitoba. The Progressive Conservative government also considered a second transmission–expansion project. The Manitoba generation project would require a new transmission line, estimated to cost $1.4 billion. Additionally, in July 2003, Ontario and Quebec renewed discussions regarding the construction of a $300 million, 1500 MW transmission line between the two provinces; if approved, the line was expected to take 2 years to construct. Additionally, the Progressive Conservative government announced a number of tax incentives to promote the purchase of energy efficient appliances by consumers and development of “green” (i.e., clean and renewable) generation.47 On October 2, 2003, Ontario held a provincial election. All three main parties, including the Progressive Conservatives, pledged to keep Ontario’s electricity assets under government ownership. The election was won by the Ontario Liberal Party. During the election, the Liberals made a number of other commitments with respect to the electricity sector, including: maintaining the retail price freeze until 2006; installing “smart” meters in residential households by 2006; and retiring all of the province’s coal-fired generation plants, which accounted for 7560 MW (approximately 25%) of the province’s total generation capacity at the time of the election, by 2007 for environmental and health reasons. The coal plants were to be replaced with natural-gas fired and renewable generation. 12.6.1. Pricing, regulation and consumer response As noted, during the election the Liberals pledged to maintain the retail price freeze until 2006. In 2006, the freeze was to be replaced with an administered price system, involving a set of prices for low- and high-demand periods, as well as a set price for a basic amount of consumption and a higher price for consumption over the basic amount. However, soon after taking office the new government found that it had inherited a substantial budget deficit and asserted that the frozen retail price was financially unsustainable and could not be maintained until 2006. On April 1, 2004, a new interim pricing plan was implemented. Under the new pricing regime consumers covered by the 4.3 cent/kWh freeze began paying 4.7 cents/kWh for the first 750 kWh consumed per month, and 5.5 cents/kWh for consumption above that level. The OEB was mandated to design a new rate structure to replace the interim structure. The new government proceeded to restructure the regulatory framework underpinning the electricity sector. These actions were based on the government’s assessment that Ontario required approximately $25 billion to $40 billion in investment “to refurbish, rebuild, replace or conserve 25,000 MW of generating capacity by the year 2020,” about 80% of Ontario’s installed capacity at the time, to satisfy demand growth and shutdown all of 47

Ontario (2002).

436


Fig. 12.8. Institutional structure of the Ontario electricity sector, 2005. Source: IESO.

Ontario’s coal-fired generating plants.48 On December 9, 2004, the new government passed the Electricity Restructuring Act, 2004, setting out a new direction for the province’s electricity sector. The Act altered the regulatory environment of the electricity significantly49 (Fig. 12.8). The Act established a new body called the Ontario Power Authority (OPA). The Power Authority would assume some of the responsibilities currently assigned to the IMO (e.g., long-term demand forecasting). Most importantly, however, the OPA would have the responsibility and ability to ensure, plan and procure new generation and transmission capacity for the province through designing an Integrated Power System Plan to be reviewed by the OEB. A Conservation Bureau was created within the OPA to develop conservation programs. Further, the IMO was renamed IESO, continuing to operate the wholesale market and ensuring operation and reliability of Ontario’s power system. Additionally, the Market Surveillance Panel was transferred from the IESO to the OEB. It is too early to tell whether the new regulatory structure is an improvement over the previous regime. On the other hand, it has imposed costs as the regulatory bodies have had to sort out and clarify their new roles and responsibilities. Further, the OPA has required time to build the institutional infrastructure required to fulfill its mandate. 48 49

Ontario (2004b). Ontario (2004c).


437

As part of its plan to stabilize prices, the provincial government removed OPG’s baseload hydroelectric and nuclear generation from competitive market pricing (about 40% of the province’s installed generation capacity) in February 2005.50 This was accomplished by setting the price of OPG’s baseload hydro-electric generation at 3.3 cents/kWh, and the price for OPG’s nuclear generation at 4.95 cents/kWh. According to the Ministry, “The prices on OPG’s regulated assets are based on projected costs of operation, plus a 5% return on equity.” The regulated baseload prices will remain in effect until the OEB develops its own pricing regime, no later than March 31, 2008. When announcing the regulated baseload prices, the provincial government also announced that the MPMA would be replaced with a temporary (lasting from April 1, 2005 to April 30, 2006) upper revenue limit of 4.7 cents/kWh on 85% of the output from OPG’s unregulated assets (about 33% of the province’s generation capacity). In February 2006, the government extended the revenue cap on OPG’s unregulated assets for three years. The OEB released a new retail rate structure in March 2005.51 The new structure went into effect on April 1, 2005. Under the new Regulated Price Plan, from April 1 to October 31, 2005, consumers covered by the previous retail freeze would face a price of 5 cents/kWh for the first 750 kWh consumed per month and 5.8 cents/kWh for consumption above that level. From November 1, 2005, to April 30, 2006, designated consumers would face a price of 5 cents/kWh for the first 1000 kWh consumed per month and 5.8 cents/kWh for consumption above that level. Consumers not wishing to participate in the regulated price regime are free to enter into fixed-price contracts with electricity retailers or acquire interval meters and pay the spot market price. The OEB also released details for its “smart” meter, time-of-use price schedule.52 This is the result of the new government’s plan to have “smart” meters installed into every home by December 31, 2010, with an interim target of 800,000 meters installed by December 31, 2007. The capital cost of installing the meters is estimated at $1 billion. The OEB’s price schedule will be seasonal (summer and winter) and have three price categories during the day: off-peak, mid-peak and on-peak.53 The goal of the “smart” meter program is to increase demand responsiveness. Charging users prices more reflective of demand and supply conditions should encourage conservation and demand shifting, as well as increasing the demand for products that conserve electricity, creating incentives for manufacturers to develop appliances and equipment that consume less electricity, eliminating the need for more interventionist methods of promoting conservation and appeals to ration.54 Beginning April 1, 2008, all consumers, except residential and general service consumers under 50 kW, are expected to face “smart” meter prices, unless they choose a different retail option; consumers under 50 kW will remain eligible for the prices in the two-tiered Regulated Price Plan. It is worth commenting further on the possible implications of the government’s decision to regulate OPG’s prices, enter into power purchase agreements with private generators 50

Ontario (2005a). OEB (2005a); see also Faruqui and George (2005). 52 OEB (2005b). 53 For example, the summer weekday schedule (May 1–October 31) designates 10 PM–7 AM as off-peak at a price of 2.9 cents/kWh, 7 AM–11 AM and 5 PM–10 PM as mid-peak at a price of 6.4 cents/kWh, and 11 AM–5 PM as on-peak at a price of 9.3 cents/kWh. All hours of weekends and holidays are designated as off-peak. 54 Joskow (2000), Heath (2003), Cicchetti and Long (2000). 51

438


and introduce a time-of-use pricing regime. The regulation of OPG’s baseload generation prices and the creation of the OPA to contract for new supply will reduce the degree to which the province’s electricity prices are determined in the spot market and competition between generators. While the use of long-term purchase agreements will attract much needed private-sector investment and stabilize prices, the net fiscal benefits resulting from the province, or the OPA, directly entering into long-term power contracts are unclear. The strategy being pursued will entail, to a greater or lesser degree, shifting market risks from investors to electricity ratepayers and potentially taxpayers. Experience shows that government contracting for power is an extremely risky and expensive proposition. Ontario Hydro entered into about 90 long-term power purchase agreements (some of the contracts do not expire until 2048) involving billions of dollars with non-utility generators in the late 1980s and early 1990s at what many consider to be above-market prices, although some credit these arrangements (representing about 1700 MW of generation capacity) for keeping the lights on during the summer of 2002. Following its electricity crisis, estimates indicated that California entered into approximately US$40 billion worth of power contracts that are likely to have a value of only US$20 billion.55 The costs of these contracts will be paid for through higher rates to electricity ratepayers or through higher rates to taxpayers. Nonetheless, the government claims to have designed contracts for 2225 MW of natural gas-fuelled generation with private firms that ensures that “all construction, operational, performance and efficiency risks will rest with the contract winners,”56 stating that “the contract structure is based on a contingent support payment, with all new generation priced through Ontario’s competitive electricity market. The contract winners are assured that they will have sufficient ongoing revenue to meet their fixed-project costs, such as capital and financing, if they operate efficiently according to the pre-agreed standards. When market revenues exceed these fixed-cost requirements, the contracts stipulate that 95% of the surplus will flow back to ratepayers.”57 It remains to be seen whether these agreements will allocate risks between consumers and investors efficiently in practice.58 55

Sweeney (2002), p. 224. Ontario (2005d). 57 Ibid. 58 Also, while recognizing that government intervention in the Ontario market is necessary at present to secure adequate supply, it should be noted that arguments by generators that long-term contracting is necessary to attract private investment are not entirely compelling. In some cases the claim may be an attempt to capitalize on the government’s pressing need for new generation, coupled with information asymmetries between electricity producers and the government; Ontario may be particularly vulnerable to such arguments at present. For example, Rothwell and Gomez (2003, p. 111) argue that forward contracts in electricity markets between private producers and consumers typically do not last more than a few years and that financiers consider other factors when making lending decisions: “The price volatility of power markets is the rationale for both sellers and buyers to make long-term contracts to hedge against uncertainty. These contracts usually do not last more than a few years and they are rarely the basis on which financiers award construction loans for generation plants. More frequently, financiers back plants because the borrower (typically an established utility that can assume risk) can provide collateral. However, the expectation of future cash flows is currently the main market force driving generation expansion in competitive markets.” Moreover, in many industries, such as automobile manufacturing, advanced technology and natural resources, plants and facilities of substantial cost are built without long-term contracts pre-selling the final product. However, in order to enhance incentives to maintain sufficient available capacity to meet all contingencies, it may be necessary to consider capacity payments (as in England and Wales until recently) or obligations on all major load-serving entities to meet capacity reserve requirements through contract coverage (as in the Pennsylvania–New Jersey–Maryland (PJM) market). 56


439

Therefore, the regulated retail prices blend together the price of OPG’s regulated electricity generation, the price of electricity procured through contracts with private generators and the price of electricity determined in the spot market. This will be a complex exercise and it is unclear whether regulated prices will be able to reflect market prices accurately in the long term in such an environment. 12.6.2. Supply The new provincial government has stated it will not sell any publicly owned electricity assets, such as transmission and generation facilities. Further, a review committee appointed by the new government, the OPG Review Committee, concluded that OPG should continue to be owned by the government of Ontario and that it should maintain ownership of its nuclear, major hydro-electric and fossil-fuel generating assets.59 At the same time, the government faced the immediate challenge of contracting for the construction of 25% of its current generation capacity to meet its commitment to phase out all coal-fuelled plants by 2007 (subsequently delayed to 2009, see below). To meet the province’s generation needs, the OPG Review Committee recommended that OPG proceed with restarting Pickering “A” Unit 1. The provincial government announced in July 2004, that it was directing OPG to restart Unit 1 at an expected cost of $900 million. The decision to restart Pickering’s Unit 1 is interesting because in December 2003, the Pickering “A” Review Panel, appointed by previous Progressive Conservative government to investigate the Pickering restart delays, issued its final report.60 In 1999, OPG estimated that the total cost of the restart would be $1.1 billion with the last unit being restarted by December 2002. In contrast, the Panel found that returning all four units would cost an estimated $3–$4 billion with the last unit being restarted between October 2006 and August 2008. The Panel also found that the project suffered from significant management and accountability deficiencies.61 In addition to the Pickering Unit 1 restart, the provincial government has announced or approved a number of other generation projects. By June 2005, the government announced that 9145 MW of additional generation capacity and demand-side projects had either been finalized or were under negotiation.62 At the same time, the government pushed back its deadline for shutting-down all of the province’s coal-fired generation capacity from 2007 to 2009, determining that network reliability would be compromised if the original deadline was pursued, given the current state of replacement and new project development. The first coal plant, the 1140 MW Lakeview plant, was closed in April 2005. The remaining coal plants will be closed beginning in 2007. Maintaining a reliable supply of electricity also requires investment in the province’s generation capacity and transmission grid in many regions of the province.63 Much of the province’s transmission grid was installed in the 1960s, with some of it dating back to the 1940s. Also, investments are required to maintain reliability as coal plants are shutdown while other investment is necessary because investment in generation and transmission have not kept pace with demand growth since the restructuring process began in 1995.

59

OPG Review Committee (2004). Pickering “A” Review Panel (2003). 61 Following the report, the government accepted the resignation of OPG’s CEO and other senior executives and the Board of Directors, and appointed replacements. 62 Ontario (2005b). 63 IESO (2005b). 60

440


12.6.3. Conservation As stated above, the Liberal government created a Conservation Bureau within the OPA to promote conservation. Additionally, the government introduced a number of policies designed to reduce electricity demand in the province. These policies focus on promoting conservation to the public and stricter energy efficiency standards aimed at reducing peak electricity demand by 5 percent by 2007 by creating a “culture of conservation.” For example, in January 2004, the province appointed a Conservation Action Team to “to seek out and promote best conservation practices and ideas across Ontario, and to identify existing barriers to conservation;” the team released its report in May 2005.64 12.7. Lessons from the Ontario Experience Ontario is a vivid warning that jurisdictions should not approach reform lightly. The costs of failure, whether caused by flawed market design, political interference or both, can be very high. Rather than reducing the province’s electricity-related debt, lowering prices and encouraging private-sector investment in generation, Ontario’s restructuring experiment resulted in an increase in electricity-related debt, higher prices, appeals to ration consumption and private investors demanding consumer/taxpayer-backed contracts to build new capacity. Before restructuring was implemented, centralized long-term integrated system planning and investment was carried out. Following restructuring, for the most part (with the exception of the system operator), it was expected that this model would be replaced by decentralized decision-making by private investors. Instead, the uncertainty surrounding restructuring resulted in integrated system planning being in a state of limbo for years and little investment by the private sector. Ontario’s electricity problems are far from over. The result of the planning and investment vacuum is that 3 years after the retreat from privatization and retail price de-regulation the province’s power system remains vulnerable to extreme weather in the short term. Significant investments are required to ensure system reliability in the long term. An extremely hot summer in 2005 saw electricity demand hit record levels, resulting in the IESO issuing public appeals to reduce consumption and, during 2 days in August, implementing voltage reductions at peak times. Furthermore, continued uncertainty about the sustainability of its electricity system could have significant economic implications for Ontario. First, continued uncertainty could result in Ontario becoming a less attractive location for international and domestic investment relative to other jurisdictions. Second, the costs of maintaining the system could compromise the provincial government’s ability to fund other public spending priorities such as health care, education and other infrastructure. Even if no restructuring had been attempted, the province would still face the current problems of a lack of demand responsiveness and the substantial financial investments needed to sustain a stable electricity system. At the same time, private-generation projects that were planned, but cancelled due to the previous government’s policy reversals in 2002, could have been completed or would be near completion at present, providing a more solid supply situation than Ontario has currently. The current government’s pledge to retire all coal-fired plants has put further pressure on the province to add new, more expensive capacity. Whether the electricity sector is privately or publicly owned, system planning must be carried out and consumers and producers of electricity need to face prices that reflect actual conditions of supply and demand. Exposing consumers and producers to 64

Ontario (2004d, e; 2005c).


441

prices that reflect market conditions creates demand-side incentives to reduce and shift consumption, use products that conserve electricity and supply-side incentives to investment. However, it is politically difficult to achieve efficient prices when “efficient prices” means higher prices, especially in the immediate short term. In Ontario, it remains to be seen if residential consumers will react differently to a higher regulated tiered-rate pricing structure and higher peak rates under time-of-use pricing than they did to the higher prices experienced during the short period of retail price de-regulation. Since the decade-long price freeze that preceded attempts at sector restructuring Ontario electricity consumers are accustomed to very low and stable prices and easily recognize and resist any change to this situation. Restructuring initiatives must address political realities directly to better ensure success and prevent costly policy reversals and backtracking. As stated earlier, an obvious way to minimize political resistance to restructuring is to ensure that there is thorough publicity of the benefits of the plan, for example, in the case of Ontario, reduced public debt. While education is far from a panacea for political opposition, there are some obvious pitfalls that should be avoided (e.g., claiming that restructuring will lead immediately to lower prices). However, education done well will not necessarily eliminate political resistance to restructuring and de-regulation. There may be losers from such a change who will resist regardless of the social good, and there also will be those who rationally do not invest in understanding the benefits. Moreover, some of the benefits only arise if there is a firm commitment by the government to adhere to its agenda (time consistency). Thus, another strategy that the government can pursue is to adopt policies that are irreversible, or at least very costly to reverse. For example, if attracting private investment and fostering competition is a goal of the restructuring policy, privatizing whatever government corporations are involved in the industry creates a political constituency in favor of restructuring: the firms (and their workers) that have invested in competing in the restructured market. Once this constituency is active, the government will face countervailing pressure not to renege on its plans, although other problems could still emerge (e.g., private generators demanding government-backed power purchase agreements could occur with privatization). Other options exist. Incremental retail price de-regulation (beginning with the largest consumers) to avoid policy reversals caused by sudden and unexpected shocks that result in higher prices for a large proportion of consumers in the early period of sector restructuring. Also, in the short term, the adoption of less-than-ideal interventionist strategies to prevent immediate upwards price shocks, such as capacity reserve requirements or a limited subsidy for a basic level of residential consumption, could be prudent strategies. Prior work regarding government intervention in response to economic shocks has noted that “policies that do not address the root problem but rather deal only with the consequences of a shock on an ad hoc basis create future costs and can exacerbate the impact of a particular shock.”65 As this review has attempted to point out, political considerations resulted in Ontario reacting to unforeseen electricity price increases with policies (e.g., retail price freezes) that imposed high costs on taxpayers and exacerbated supply demand imbalances. This reaction resulted in further policies to restore order to the province’s electricity sector. It remains to be seen whether the current government can remain faithful to its new restructuring plans and whether this new vision will be workable technically, financially and politically. Problems have emerged. For example, in December 2005, the OPA submitted its Supply Mix Advice Report, a 20-year, $70 billion investment plan, to the government. The report included a recommendation that the province build new nuclear generation capacity. Environmental groups criticized the report sharply and demanded that the government hold extensive public consultations before approving a long-term supply plan. 65

Iacobucci, Trebilcock and Haider (2001), p. 167.

442


12.8. The Alberta Restructuring Experience in Brief Alberta is an energy-rich province in both coal (possessing 60% of Canada’s total reserves) and natural gas (accounting for over 80% of Canadian production). The province’s electricity system (Fig. 12.9) is owned and operated by investor owned and municipally owned

Fig. 12.9. Transmission map of Alberta. Source: AESO.


443

companies. As of May 2005, the province of Alberta had 12,099 MW of installed capacity. The capacity is comprised of coal (48.3%), gas co-generation (26.8%), gas (13.8%), hydro (7.4%), and wind and biomass (3.7%).66 The province has the ability to import 950 MW67 of power from the neighboring provinces of British Columbia and Saskatchewan. Alberta has no direct interconnections with the US. Alberta began to reform its electricity market in 1995 through the Electric Utilities Act. The motive for restructuring (Table 12.2) was to attract private-sector investment in new generation to meet growing demand and a perception that the existing regulatory environment was not conducive to competition.68 Prior to reform the Alberta’s market was primarily composed of three vertically integrated utilities with assigned service areas. These three firms accounted for about 90% of the province’s total generation capacity.69 However, unlike Ontario, the provincial government of Alberta did not own any electricity assets. Two of these utilities were investor owned and the other was municipally owned by the city of Edmonton (EPCOR). The private utilities were regulated by Alberta’s Energy Utility Board (AEUB), as were EPCOR’s transmission and generation assets. Electricity was purchased by the

Table 12.2. Electricity sector restructuring milestones in Alberta. Event

Date

Comment

Electric Utilities Act

1995

Wholesale market opens Electric Utilities Act Amended PPAs auctions

1996 1998

Electricity Rebate Program

November– December 2000 November– December 2000

Legislation established the open-access, competitive power pool. Wholesale market operations begin. Established structure for PPAs and timeline for customer choice. Rights to sell generation of incumbent utilities auctioned. Government announces that consumers will receive rebates funded from the proceeds of PPA sales. A 5-year RRO established for home and farm customers. A 3-year RRO established for eligible small commercial and industrial customers. RRO to take effect from January 2001. Retail choice available for all consumers. Incumbent Progressive Conservative Party re-elected. AESO created, replacing Power Pool of Alberta and Transmission Administrator. RRO extended to July 2006 for all RRO eligible customers. Recommendations for improving retail and wholesale markets. RRO extended to 2010.

RRO

2000

Full retail access Provincial Election

January 2001 March 2001

New Electric Utilities Act

June 2003

RRO extended

November 2003

Alberta’s Electricity Policy Framework

June 2005

66

For up-to-date information and data on the Alberta market consult the Alberta Department of Energy (www.energy.gov.ab.ca), the AESO (www.aeso.ca), the Alberta Energy and Utilities Board (www.eub. gov.ab.ca) and the Balancing Pool (www.balancingpool.ca). 67 In practice imports and exports are more limited than this because of technical considerations on transmission. 68 Alberta (2004). 69 Daniel et al., (2003).

444


government at a regulated, cost-of-service rate (different for each generator) and then resold to utilities’ distribution divisions at an averaged, uniform price.70 The Electric Utilities Act mandated that the utilities transfer control of their transmission assets to an independent Transmission Administrator. A mandatory uniform price spot market pool, the Power Pool of Alberta, was established in January 1996. The Act mandated that from 1996 to 2000, the generation price would continue to be set at a cost-of-service rate for regulated generators (i.e., incumbent generation built before 1995).71 Transmission and distribution rates continued to be set by the AEUB. The province did not require incumbent utilities to divest generation capacity. Instead, to facilitate competition in the wholesale market, the government mandated that the incumbent utilities divest the production rights of their generation assets, rather than require divestiture outright.72 The generation rights, called Power Purchase Arrangements (PPAs), would be auctioned to private investors. The PPA required the bidder to purchase a fixed output of an individual generating unit for up to 20 years (the length of the PPA was based on the remaining life of the regulated unit). The owners of the generation rights are required to bid all of their power into the Power Pool, but were permitted to enter into long-term hedging contracts with their customers.73 The government created the Balancing Pool to manage the PPAs (whether sold or unsold). A PPA contract required the buyer to pay the incumbent owner the marginal cost of generation of each unit of electricity produced and bid into the Power Pool, plus a fixed monthly payment determined by regulators.74 In exchange, the owners of PPAs could bid the power produced by the plants covered by the PPA into the Power Pool, retaining the revenue.75 Two auctions took place in 2000.76 In the first auction, in August 2000, the regulated generation was split into 12 PPAs. Seven bidders participated with five bidders winning PPAs; eight PPAs (amounting to 4249 MW) were sold for about $1.1 billion, leaving four PPAs, representing thousands of megawatts, unsold. It has been argued that the low number of bidding firms resulted in a bids being lower than they otherwise would have been. With a low number of bidders and firms faced with a 20% total capacity limit, firms may have had a tacit understanding not to bid up prices, allowing one firm to win individual auctions.77 An alternative explanation for the lack of bidders may be that the private sector was not convinced that Alberta’s provincial government committed to restructuring during this early stage of the restructuring period and were also concerned about the risks of not owning or operating the generation units. Contributing to these concerns may have been that the PPAs had terms of up to 20 years. These risks may have contributed to heavily discounted PPA bids. It should also be noted that some of the unsold units were simply not economic in a competitive market and not attractive to private investors. The second auction to sell the power not sold in the August auction was held in December 2000. A different approach was taken in the second auction, referred to as the Market Achievement Plan (MAP) auction. The PPAs were sold for terms of 1–3 years and in much smaller quantities of capacity than the first auction and not associated with specific generation units. 70

Jaccard (2002), p. 24. Daniel et al., (2003). 72 Jaccard (2002), p. 19. 73 Ibid., p. 20. 74 Daniel et al., (2003). 75 Ibid. 76 See Balancing Pool (2001). 77 Silk (2001), p. 15. 71

445

Electricity restructuring in Canada 140.00

Price ($C/MWh)

120.00 100.00 80.00 60.00 40.00 20.00 0.00

1996

1997

1998

1999

2000

2001

2002

2003

2004

Fig. 12.10. Yearly average real-time wholesale electricity price in Alberta, 1996–2004. Source: AESO.

The result was a more competitive auction, consisting of 62 bidders with 45 successful bidders. About 2800 MW of electricity contracts were sold. About $1 billion in revenue was raised from the second auction. It is likely that market participants had gained more confidence in the government’s commitment in the time elapsing between the first and second auction. Additionally, the California crisis was continuing to unfold and the price of energy was higher. Wholesale prices consistently increased between 1996 and 1999 (Fig. 12.10). Between 1999 and 2000 wholesale prices rose dramatically from an average of $42.74/MWh in 1999 to $133.22/MWh in 2000. It has been argued that prices increased because exports and imports were permitted to set the market-clearing price (British Columbia was importing electricity from Alberta for export to California); in November 2000, Alberta changed its wholesale pricing system to exclude imports and exports from setting pool prices.78 Rising natural gas prices also contributed to rising wholesale prices because many marginal generation plants in Alberta were gas fired.79 This was the result of most of Alberta’s new generation (i.e., post-1995) being gas fired, resulting in a greater proportion of marginal generation being gas fired. Further, the province had little import capacity as its transmission grid was interconnected with only British Columbia and Saskatchewan.80 It was also argued that the initial linkage of price to cost in 1995 suppressed the price of electricity in the province and discouraged investment in new generation.81 Wholesale prices fell significantly after 2000, following the resolution of the supply crisis across western North America (e.g., California), a reduction in natural gas prices and new generation entry. Approximately 2500 MW of new capacity (primarily gas fired) came into service in the Alberta system between 1998 and 2002. Also, in response to the rising prices during the summer of 2000, the Alberta Market Surveillance Administrator (AMSA) conducted an investigation into power pool pricing

78

Daniel et al., (forthcoming). Ibid. 80 Alberta Advisory Council on Electricity (2002). 81 Jaccard (2002), p. 9. 79

446


behavior.82 The AMSA report argued that there may have been evidence that economic and physical capacity withholding had occurred. Specifically, the investigation claimed that electricity was being sold at prices that exceeded the marginal cost of production,83 although such an observation does not necessarily imply an abuse of market power by generators. Second, the report stated that evidence indicated that generators may have kept capacity off-line without a valid physical or operational justification.84 Two significant policies were implemented following consumer complaints of high prices in 2000. In the initial restructuring plan, it was scheduled that all consumers would have to choose a competitive retailer for retail service. However, in November and December 2000 the government announced a default capped retail rate for consumers that did not switch to a competitive retailer.85 Residential, farm, small commercial and industrial customers were given the option of choosing to remain on a default regulated rate option (RRO) with their incumbent utility. A 5-year RRO was established for home and farm customers. A 3-year RRO established for eligible small commercial and small industrial customers. A province wide RRO was set at 11 cents/kWh for 2001. At the same time the government initiated a $2 billion electricity rebate program for 2001. The program funded through the Balancing Pool with the proceeds of the PPA sales.86 Residential and farm customers received a $40/month rebate while non-residential and commercial farm customers received a 3.6 cent/KWh rebate. These policy announcements also came shortly before a provincial election held in March 2001. The incumbent Progressive Conservative party, which had initiated the restructuring program, was re-elected. In 2002, the government ceased to set a province wide RRO. Energy suppliers request approval for rate changes within the RRO from their designated regulator (i.e., AUEB, municipal regulator or a Rural Electrification Association) to recover or refund the amount by which the RRO differed from their procurement costs through customer bills. The Alberta decision to have consumers ultimately pay for the difference between wholesale and retail prices can be contrasted to that of Ontario (difference paid by taxpayers) and California (difference paid by shareholders). The RRO has been extended twice from its initial 3- and 5-year lifespan. Currently, the RRO is scheduled to remain in place until 2010. These delays would appear to indicate that the outcomes in California and Ontario, as well as Alberta’s own electricity price spikes in 2000, have led Alberta policy makers to approach further retail price de-regulation with caution. The RRO has resulted in few small consumers leaving their incumbent utility. By April 2005, approximately 7% of the small consumer market had signed a contract with a competitive retailer.87 Switching rates have been more significant for commercial and industrial consumers. Approximately 37% of small commercial consumers and over 70% of industrial and large commercial consumers have switched to a competitive marketer.88 In June 2003, Alberta further restructured its electricity market with a new Electric Utilities Act. The new Act created the Alberta Electric System Operator (AESO), which merged the responsibilities of the Power Pool of Alberta and the transmission administrator. The AESO

82

Alberta Market Surveillance Administrator (2002). Ibid., p. 17. 84 Ibid. 85 Jaccard (2002), p. 21. 86 See Balancing Pool (2001). 87 Alberta (2005), pp. 9–10. 88 Ibid. 83


447

is responsible for planning and operating Alberta’s electricity system as well as identifying the need to upgrade and enhance the province’s transmission grid. 12.8.1. Observations from the Alberta experience First, as in other jurisdictions the restructuring initiative in Alberta was set back by reluctance to pass higher prices on to consumers. Second, while significant new generation entry has occurred, the PPAs appear to have been an unsatisfactory substitute for facilities-based competition; divestiture would have likely been a more suitable, although politically difficult, course of action to enhance competitive conditions in the early period of restructuring. Third, Alberta illustrates how a lack of import capacity can limit a jurisdiction’s ability to avoid price shocks, although greater export capacity could have exacerbated the price shock in 2000. Fourth, the willingness of the province to allow the RRO to vary within the province appears to be preferable to an inflexible, uniform, rate freeze (e.g., Ontario). Fifth, it must be recalled that Alberta possesses significant energy resources in coal and natural gas. The Alberta resource advantage is an important factor that must be taken into account when comparing the Alberta and Ontario experience in electricity sector restructuring and their differing paths with respect to government policy commitment in the face of unexpected and significant price increases. Sixth, while Alberta has backed away from and delayed full retail price de-regulation, it has continued to support and refine its restructured market89 and the consistent policy commitment by the Alberta government to private ownership and a competitive generation sector has had the obvious benefit of inducing substantial new private-sector investment in generation, and reduction and stabilization of electricity prices (in contrast to Ontario). Between 1998 and May 2005, 3853 MW (32% of May 2005 installed capacity) of new generation capacity was added to the Alberta system and a further 4300 MW have been announced for further development. The fact that Alberta’s electricity system, unlike Ontario’s, was mostly investor owned prior to restructuring and lacked a long-term strong government presence in electricity provision, is likely a main factor behind the different policy decisions made by the two provinces following price shocks. References Adams, T. (2000). From Promise to Crisis: Lessons for Atlantic Canada from Ontario’s Electricity Liberalisation. Atlantic Institute for Market Studies. Alberta (2004). Department of Energy. Facts on Electricity Deregulation. Alberta (2005). Department of Energy. Alberta’s Electricity Policy Framework. Alberta Advisory Council on Electricity (2002). Report to the Alberta Minister of Energy. Alberta Market Surveillance Administrator (2002). Report on Power Pool of Alberta Prices – Summer 2000. Balancing Pool (2001). Annual Report: 2000. Cicchetti, C.J. and Long, C.M. (2000). Politics as usual: a roadmap to backlash, backtracking, and re-regulation. Public Utilities Fortnightly, 138(18), 34–44. Daniel, T., Doucet, J. and Plourde, A. (forthcoming). Electricity industry restructuring: the Alberta experience. In Andrew N. Kleit (ed.), The Challenge of Electricity Restructuring, Rowman and Littlefield, London. Dewees, D.N. (2005). Electricity restructuring in Canada. In G. Bruce Doern (ed.), Canadian Energy Policy and the Struggle for Sustainable Development. University of Toronto Press, Toronto, pp. 128–150.

89

For example, see Alberta (2005).

448


Distributors’ Electricity Efficiency Policy Group (2003). Delivering Value: The Next Evolution of Electricity Distribution in Ontario, October (Draft). Electricity Supply and Conservation Task Force (2004). Tough Choices: Addressing Ontario’s Power Needs. Faruqui, A. and George, S.S. (2005). Preventing Electric Shocks: What Ontario – and Other Provinces – Should Learn About Smart Metering. C.D. Howe Institute Commentary 210. Goulding, A.J., Frayer, J. and Uludere, N.N. (2001). Dancing with Goliath: prospects after the breakup of Ontario hydro. Public Utilities Fortnightly, 139(5), 22–33. Hampton, H. (2003). Public Power: The Fight for Publicly Owned Electricity. Insomniac Press, Toronto. Heath, J. (2003). Collective Power Struggle. University of Toronto Bulletin. September 8: 20. Hunt, S. (2002). Making Competition Work in Electricity. John Wiley & Sons, New York. Iacobucci, E.M, Trebilcock, M.J. and Haider, H. (2001). Economic Shocks: Defining a Role for Government. C.D. Howe Institute, Toronto. Iacobucci, Edward, Trebilcock, Michael and Winter, Ralph A. (2006). The Canadian experience with deregulation. University of Toronto Law Journal, 56(1), 1–74. IESO (2005a). Your Road Map To Ontario Wholesale Electricity Prices. IESO (2005b). 10-Year Outlook: An Assessment of the Adequacy of Generation and Transmission Facilities to Meet Future Electricity Needs in Ontario, July. IMO (2002a). 10-Year Outlook: An Assessment of the Adequacy of Generation and Transmission Facilities to Meet Future Electricity Needs in Ontario, April. IMO (2002b). Market Surveillance Panel Monitoring Report on the IMO-Administered Electricity Markets for the First Four Months, May to August 2002, Executive Summary, October. IMO (2002c). Market Surveillance Panel Monitoring Report on the IMO-Administered Electricity Markets for the First Four Months, May to August 2002, October. IMO (2002d). Ontario Demand Forecast from January 2003 to December 2012, April. Jaccard, Mark (2002). California Shorts a Circuit: Should Canadians Trust the Wiring Diagram? C.D. Howe Institute Commentary 159. IMO (2003a). The Ontario Wholesale Electricity Market: Year in Review, June. IMO (2003b). Market Surveillance Panel Monitoring Report on the IMO-Administered Electricity Markets for the Period from September 2002–January 2003, March. IMO (2003c). 18-Month Outlook: An Assessment of the Reliability of the Ontario Electricity System from October 2003 to March 2005 September. Joskow, P.L. (2000). Deregulation and regulatory reform in the U.S. electric power sector In Sam Peltzman and Clifford Winston (eds.), Deregulation of Network Industries: What’s Next? AEI-Brookings Joint Center for Regulatory Studies, Washington, D. C., pp. 113–188. National Energy Board (2001). Canadian Electricity Trends and Issues. National Energy Board (2005). Outlook for Electricity Markets 2005–2006. OEB (2005a). Changes to electricity prices for Ontario effective April 1, 2005, March. OEB (2005b). Smart Meter Implementation Plan, January. OEFC (2000). Annual Report. OEFC (2002). 2002 Annual Report. Ontario (1996). Advisory Committee on Competition in Ontario’s Electricity System. A Framework for Competition. Ontario (1997). Direction for Change: Charting a Course for Competitive Electricity and Jobs in Ontario. Ontario (2002). Ministry of Energy. Action Plan to Lower Your Hydro Bill. Ontario (2004a). Ministry of Energy. Electricity Transmission and Distribution in Ontario – A Look Ahead, December. Ontario (2004b). Ministry of Energy. McGuinty Government Unveils Bold Plan to Restructure Electricity System, June. Ontario (2004c). Ministry of Energy. McGuinty Government Brings Stability and Balance to Ontario’s Electricity Sector, December. Ontario (2004d). Ministry of Energy. Building a Conservation Culture in Ontario: Release of the Government’s Conservation Action Team. Backgrounder, May. Ontario (2004e). Ministry of Energy. New Energy Efficiency Standards Help Protect the Environment and Save Consumers Money, March.


449

Ontario (2005a). Ministry of Energy. Ontario Government Introduces Fair And Stable Prices For Electricity From Ontario Power Generation, February. Ontario (2005b). Ministry of Energy. McGuinty Government Unveils Bold Plan to Clean Up Ontario’s Air, June. Ontario (2005c). Ministry of Energy. Conservation Action Team. Building a Conservation Culture in Ontario. Ontario (2005d). Ministry of Energy. Contract Structure and Pricing. May. OPA (2005). Supply Mix Advice Report. OPG Review Committee (2004). Transforming Ontario’s Power Generation Company. Pickering “A” Review Panel (2003). Report of the Pickering “A” Review Panel. Rothwell, G. and Gomez, T. (eds.), (2003). Electricity Economics: Regulation and Deregulation. IEEE, Hoboken, NJ. Schott, S. (2005). Sustainable and socially efficient electricity production: how will Ontario satisfy the criteria? In G. Bruce Doern (ed.), Canadian Energy Policy and the Struggle for Sustainable Development. University of Toronto Press, Toronto, pp. 175–199. Silk, W. (2001). Alberta’s road to competitive electricity markets. Canadian Bar Association Annual Fall Conference on Competition. Sweeney, J.L. (2002). The California Electricity Crisis. Hoover Institution, Stanford, CA. Trebilcock, M.J. and Daniels, R. (2000). Electricity restructuring: the Ontario experience. Canadian Business Law Journal, 33(2), 161–192. Trebilcock, M.J. and Hrab, R. (2003). What Will Keep the Lights on in Ontario: Responses to a Policy ShortCircuit. C.D. Howe Institute Commentary 191. December. Trebilcock, M.J. and Hrab, R. (2005). Electricity restructuring in Ontario. Energy Journal. 26(1), 123–46. White, M.W. (1996). Power struggles: explaining deregulatory reforms in electricity markets. Brookings Papers on Economic Activity, Microeconomics, 201–250. Yatchew, A. (2001). Incentive regulation of distribution utilities using yardstick competition. Electricity Journal, 14(1), 56–60.


Chapter 13 The PJM Market JOSEPH BOWRING PJM Interconnection L.L.C., Pennsylvania, USA

Chapter Summary PJM Interconnection, L.L.C. (PJM) manages the largest centrally dispatched control area in North America and operates the largest competitive wholesale electricity market in the world. PJM operates a bid-based, security constrained, economically dispatched, locationally priced, competitive wholesale electricity market with open-access transmission and financial transmission rights (FTRs). PJM has achieved its success to date based on a number of factors including PJM’s and PJM members’ historical experience as a power pool; the complete set of market rules developed by PJM and its members prior to PJM markets implementation; the governance structure of PJM including the independence of PJM and its Board; the membership process and the behavior of the members in proposing, evaluating and approving multiple changes to the market rules; an incremental approach to meeting significant market challenges; an independent market monitoring function; and the incremental approach to implementing retail competition by the states and therefore the member utilities.

13.1. Introduction The PJM market comprises, as of mid-2005, generating capacity of approximately 164,000 megawatts (MW) and about 330 market buyers, sellers and traders of electricity serving more than 51 million people in all or parts of Delaware, Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, West Virginia and the District of Columbia. PJM grew substantially in 2004 and 2005 as the result of the integrations of new members in Illinois, Indiana, Kentucky, Michigan, North Carolina, Ohio, Tennessee, Virginia and West Virginia. PJM covers about 164,000 square miles and includes about 56,000 miles of transmission lines and about 1,200 generating units. Peak demand is about 135,000 MW. PJM operates a bid-based, security constrained, economically dispatched, locationally priced, competitive wholesale electricity market with open-access transmission and FTRs. Each of these components of the market description is essential to the market design of PJM and comparable RTOs. The market is based on the daily, competitive bids and offers of loads and generators. PJM dispatches the generating units to meet the instantaneous loads of electric customers in economic merit order, consistent with the physical characteristics of 451

452


the transmission system (security constrained). The result is expressed as locational marginal prices (LMP) or prices at every electric node in the system that are based on the marginal costs to meet the loads using the existing generating resources, based on the physical constraints imposed by the transmission system. Access to the transmission grid is open to all market participants and is priced at flat, license plate rates. PJM also provides financial hedges against the congestion that arises in locational markets to loads that purchase firm transmission rights. PJM has achieved its success to date based on a number of factors including PJM’s and PJM members’ historical experience as a power pool; the complete set of market rules developed by PJM and its members prior to PJM markets implementation; the governance structure of PJM including the independence of PJM and its Board; the membership process and the behavior of the members in proposing, evaluating and approving multiple changes to the market rules; an incremental approach to meeting significant market challenges; an independent market monitoring function; and the incremental approach to implementing retail competition by the states and therefore the member utilities. The PJM market rules have been central to the success of PJM and in particular the fact that the rules establish a bid-based, security-constrained market with centrally operated economic dispatch and nodal pricing with full participant flexibility including the ability to enter into bilateral contracts, to selfschedule generation and to self-supply. In addition, a detailed generator interconnection process for new generation contributes to ease of competitive entry; the requirement that new capacity resources be deliverable contributes to a robust transmission network and a comprehensive transmission planning process ensures a reliable transmission system. Other specific features of the market rules that have contributed significantly to the success of PJM as a working, competitive market include the availability of effectively unlimited non-firm transmission service willing to pay congestion; a limit of one offer per day by generators that incents competitive offers; a strong and unambiguous local market power mitigation rule; and an overall energy market offer cap of $1,000/MWh.1 The balance of this chapter provides details on the functioning of the various components of the PJM market. PJM has demonstrated how these essential building blocks can be assembled into a competitive, wholesale power market that works. The chapter reviews the characteristics of generating units in PJM; the structure, behavior and performance of the energy market including congestion and FTRs; the structure, behavior and performance of the capacity market; the structure, behavior and performance of the ancillary services markets; and the cost effectiveness of PJM operations.

13.2. PJM History PJM was created as a power pool in 1927 by Public Service Electric & Gas, Philadelphia Electric Company and Pennsylvania Power & Light, three utilities in Pennsylvania and New Jersey. PJM Agreements evolved to cover economic dispatch, capacity planning and transmission planning and to incorporate additional members. When PJM filed with the Federal Energy Regulatory Commission (FERC) to become an Independent System Operator (ISO) in 1997, PJM included eight interconnected utility systems: Atlantic City Electric Company, Baltimore Gas & Electric Company, Delmarva Power & Light Company, General Public Utilities (Jersey

1

See the PJM Operating Agreement and the PJM Open Access Transmission Tariff at http://www. pjm.com/documents/agreements.html.

The PJM Market

453

Central Power & Light Company, Metropolitan Edison Company, Pennsylvania Electric Company), Potomac Electric Power Company, Pennsylvania Power & Light Company, Public Service Electric & Gas Company and PECO Energy Company. PJM became an ISO in 1998 and a Regional Transmission Organization (RTO) in 2002. PJM uses a two-tiered governance structure, consisting of the PJM Board of Managers and the PJM Members Committee. The PJM Board is independent of all PJM market participants. The Members Committee includes a representative of every member, organized in five voting sectors representing generators, transmission owners, electric distributors, power marketers and consumers. The Members Committee provides advice to the Board on a variety of market-related and governance issues.

PJM Transmission Zones

454


PJM is regulated by the FERC and is authorized by FERC to manage the reliability of the electric transmission system and the operation of the wholesale power market in a defined control area. FERC has been supportive of PJM’s market model.2 PJM also has an active relationship with the states in the PJM footprint. The state Public Utility Commissions recently formed an organization to manage their relationship with PJM, the Organization of PJM States, Inc., or OPSI.

13.3. PJM Market Overview The PJM Market consists of a number of components, all of which are necessary parts of the whole. Each component is designed to provide a market mechanism for, and therefore a price for, each service that is required in order to operate a competitive wholesale power market. PJM market design has evolved since 1997 and continues to evolve. In 2005, for example, PJM filed to modify the capacity market design and PJM continues to examine the structure of operating reserves markets. PJM introduced nodal energy pricing with market-clearing prices based on offers at cost on April 1, 1998, and nodal market-clearing prices based on competitive offers on April 1, 1999. Daily Capacity Markets were introduced on January 1, 1999, and Monthly and Multimonthly Capacity Markets were introduced in mid-1999. PJM implemented an auction-based FTR Market on May 1, 1999. PJM implemented the Day-Ahead Energy Market and the Regulation Market on June 1, 2000. PJM modified regulation market design and added a market in spinning reserve on December 1, 2002. PJM introduced an Auction Revenue Rights (ARR) allocation process and an associated Annual FTR Auction effective June 1, 2003. PJM operates the Day-Ahead Energy Market, the Real-Time Energy Market, the Daily Capacity Market, the Interval, Monthly and Multimonthly Capacity Markets, the Regulation Markets, the Spinning Reserve Markets and the Annual and Monthly Auction Markets in FTRs. PJM also administers the Open Access Transmission Tariff (OATT).

13.4. Generating Resources – Basic Characteristics Information on the basic characteristics of the generation resources can be useful in evaluating market outcomes. Data is presented on installed capacity by fuel type, on MWh generated by fuel type and on the fuel type of units that are marginal and therefore set the market price. Outage data is also presented.

13.4.1. Capacity by fuel type On December 31, 2004, PJM installed capacity was about 144,000 MW. Of the total installed capacity, 59,800 MW, or 41.5%, was coal; 40,900 MW, or 28.4%, was natural gas; 27,400 MW, or 19.1%, was nuclear; 10,100 MW, or 7.0%, was oil; 5,300 MW, or 3.7%, was hydroelectric; and 400 MW, or 0.3%, was solid waste. (See Fig. 13.1)3

2

This chapter should be read in conjunction with Chapter 14, which provides additional details on the FERC perspective. 3 Unless otherwise indicated, all data in this chapter is from PJM sources.

455

The PJM Market Hydro Solid waste 5,301 413 3.7% 0.3% Nuclear 27,426 19.1% Coal 59,760 41.5%

Gas 40,866 28.4%

Oil 10,112 7.0%

Fig. 13.1. PJM capacity by fuel type: December 31, 2004. Generation by Fuel Source (GWH) Solid waste 2,781 0.6%

Hydro 10,303 Wind 2.3% 379 0.1%

Nuclear 165,128 36.9%

Coal 233,217 52.1%

Gas 31,411 7.0%

Oil 4,738 1.1%

Fig. 13.2. PJM generation by fuel type (GWh): Calendar year 2004.

13.4.2. Generation by fuel type In calendar year 2004, coal and nuclear units generated 88.9% of the total electricity. Coal was 52.1%, nuclear 36.9%, natural gas 7.0%, oil 1.1%, hydroelectric generation 2.3%, solid waste 0.6% and wind generation 0.1% of total generation. (See Fig. 13.2)

456

Electricity Market Reform Table 13.1 Type of fuel used by marginal units: Calendar years 2000–2004.4 Fuel type

2000 (%)

2001 (%)

2002 (%)

2003 (%)

2004 (%)

48 0 18 2 31

49 0 18 1 32

55 0 23 0 21

52 0 29 1 18

56 0 31 0 12

Coal Miscellaneous Natural gas Nuclear Petroleum

13.4.3. Marginal units by fuel type Table 13.1 shows the fuel type of marginal units. Marginal units set the price in PJM’s nodal pricing system so these shares reflect the proportion of the time that different fuel types set the price in PJM. The share of coal has risen to 56%, the share of natural gas has increased to 31% and the share of petroleum has declined to around 12%. 13.4.4. Generator performance From 1996 to 2001, the average PJM equivalent demand forced outage rate (EFORd) trended downward, reaching 4.8% in 2001, but then increased to 5.2% in 2002 and 7.0% in 2003. In 2004, the average PJM EFORd continued its upward trend, reaching 8.0%. Approximately half the increase in EFORd from 2003 to 2004 was the result of increased forced outage rates of fossil steam units, while the balance of the increase was the result of increased forced outage rates of combustion turbine (CT), nuclear and hydroelectric units. These forced outage rates are for the PJM Mid-Atlantic Region and the AP Control Zone only. The forced outage rate in 2004 was 7.9% for all zones within the PJM Capacity Market (including the AEP, DAY and ComEd Control Zones).5

13.5. Energy Market 13.5.1. Energy market design In PJM, market participants wishing to buy and sell energy have multiple options. Market participants decide whether to meet their energy needs through self-supply, bilateral purchases from generation owners or market intermediaries, through the Day-Ahead Energy Market or the Real-Time Energy Market. Energy purchases can be made over any timeframe from instantaneous Real-Time Energy Market purchases to long-term bilateral contracts. Energy purchases may be made from generation located within or outside PJM. Market participants also decide whether and how to sell the output of their generation assets. Generation owners can sell their output within PJM or externally and can use generation to meet their own loads, to sell into PJM markets or to sell bilaterally. Generation owners can sell their output over any timeframe from instantaneous Real-Time Energy Market

4

The primary fuels contained in the miscellaneous category include methane, petroleum coke, refuse, refinery gas, waste coal, wood and wood waste. 5 In some cases the data for the AEP, DAY and ComEd Control Zones may be incomplete for the year 2004 and as such, only data that have been reported to PJM were used.

The PJM Market

457

sales to long-term bilateral arrangements. Market participants can use purely financial increment offers and decrement bids in the Day-Ahead Energy Market to hedge positions or to arbitrage expected price differences between markets. The PJM Energy Market comprises all types of energy transactions, including the sale or purchase of energy in PJM’s Day-Ahead and Real-Time Energy Markets, bilateral and forward markets and self-supply. Energy transactions analyzed in this chapter include those in the PJM Day-Ahead and Real-Time Energy Markets. These PJM markets provide key benchmarks against which market participants may measure results of other transaction types. In 2004, for example, Real-Time Energy Market activity represented 35% of average loads for all hours (Fig. 13.3) while Day-Ahead Energy Market activity represented 26% of average total Day-Ahead loads for all hours.6 Both Real-Time and Day-Ahead Energy Market transactions are referred to as spot market activity because they are transactions made in a short-term market. Spot market activity as a proportion of load in the Real-Time Energy Market was 40% in 2003, 38% in 2002 and 21% in 2001. Spot market activity as a proportion of load in the DayAhead Energy Market was 31% in 2003, 32% in 2002 and 15% in 2001. As a general matter, day-ahead spot market activity is included in the real-time spot market activity. The alternatives to such spot market transactions are self-supply and bilateral arrangements. The fact that transactions occurred in the PJM spot market does not mean that the parties to the transactions were not hedged and does not mean that they were exposed to real time prices. Participants may use the PJM spot market as a mechanism for transactions even when those transactions are hedged via bilateral contracts with separately specified

80,000

2004 PJM average load Average spot volume

70,000

ComEd integration

60,000 Load (MW)

AEP and DAY integration

50,000 40,000 30,000 20,000 10,000

Ju ly Au g u Se pt st em be r O ct ob N ov er em b D ec er em be r

ne Ju

ay M

ril Ap

ch ar M

ua br

Fe

Ja

nu

ar

y

ry

0

Fig. 13.3. PJM average hourly load and spot market volume: Calendar year 2004.

6

All hours are presented and all hourly data are analyzed using Eastern (United States) Prevailing Time (EPT).

458


financial terms, for example, contracts for differences. Such bilateral contracts can and do clear through the PJM Energy Markets. A substantial proportion (approximately half in some years) of the reported spot market activity consists of transactions between parties with underlying bilateral financial contracts including, for example, generation and load serving affiliates of the same parent company. This means that market participants use the PJM settlements system and the PJM credit mechanisms to facilitate their underlying bilateral contracts. 13.5.2. Competition in the energy market The competitiveness of the PJM Energy Market can be evaluated, like any market, by examining the structure of the market, the behavior of market participants and the performance of the market. Key measures of PJM Energy Market structure, behavior and performance include concentration ratios, residual supplier indices, price-cost mark up, net revenue, prices, and the extent to which market power mitigation is required. The PJM Energy Market results have been competitive since the introduction of markets in 1999. The analysis of energy markets begins with market structure which provides the framework for the actual behavior or conduct of market participants. The analysis also examines participant behavior in the context of market structure. In a competitive market structure, market participants are constrained to behave competitively. In a competitive market structure, competitive behavior is profit-maximizing behavior. Finally, the analysis examines market performance results. The ultimate test of the markets is the actual performance of the market, measured by price and the relationship between price and marginal cost. For example, at times market participants behave in a competitive manner even within a non-competitive market structure. This may result from the relationship between supply and demand and the degree to which one or more suppliers are singly or jointly pivotal even in a highly concentrated market. This may also result from a conscious choice by market participants to behave in a competitive manner based on perceived regulatory scrutiny or other reasons, even when the market structure itself does not constrain behavior. 13.5.3. Energy market structure Concentration ratios are a summary measure of market share, a key element of market structure. High concentration ratios indicate comparatively smaller numbers of sellers dominating a market, while low concentration ratios mean larger numbers of sellers splitting market sales more equally. High concentration ratios indicate an increased potential for participants to exercise market power, although low concentration ratios do not mean that a market is competitive or that participants cannot exercise market power. Analysis of the PJM Energy Market indicates moderate market concentration overall. Further, analyses of supply curve segments indicate moderate concentration in the baseload segment, but high concentration in the intermediate and peaking segments. Several geographic areas of PJM exhibit moderate to high levels of concentration when transmission constraints define local markets. No evidence exists, however, that market power has been exercised in these areas, both because of generator obligations to serve load (equivalent to a contract) and because of PJM’s rules limiting the exercise of local market power. A generation owner, or group of generation owners, is pivotal if the output of the owners’ generation facilities is required in order to meet market demand. When a generation owner or group of owners is pivotal, that owner or group of owners has market power. The residual

459

The PJM Market

supply index (RSI) is a measure of the extent to which generation owners are pivotal suppliers. Like concentration ratios, the RSI is an indicator of market structure. When the RSI is less than 1.00, a generation owner or group of owners is pivotal. As with concentration ratios, the RSI is not a bright line test. While an RSI less than 1.0 clearly indicates market power, an RSI greater than 1.0 does not guarantee that there is no market power. As an example, suppliers can be jointly pivotal. The average RSI was 1.66 in 2003 and 1.64 in 2004 while a single generation owner in the PJM Energy Market was pivotal for only 8 hours in 2004 and for 6 hours in 2003, less than 0.1% of all hours in each case. The RSI results are consistent with the conclusion that the PJM Energy Market results are competitive. Markets require both a supply side and a demand side to function effectively. The demand side of the wholesale energy market is underdeveloped. This underdevelopment is one of the basic reasons for maintaining an offer cap in PJM and other wholesale power markets, although it is not the only reason. There is a relatively small amount of demand-side resources in PJM as measured by participation in programs operated by PJM or by utilities within the PJM footprint. For example, the maximum level of demand-side response resources available in PJM during the 9-month period ended September 30, 2004, was 11,562 MW. Thus, the maximum total demand-side resources, including PJM programs and additional programs reported by PJM customers in response to a survey, were approximately 15% of PJM’s peak demand at the time, although actual demand response was substantially less than the identified potential. 13.5.4. Energy market performance Price-cost markups are a measure of market power. The price-cost markup reflects both participant behavior and the resultant market performance. The price-cost markup index is defined here as the difference between price and marginal cost, divided by price for the marginal units in the PJM Energy Market. Overall, data on the price-cost markup are consistent with the conclusion that PJM Energy Market results are reasonably competitive. Net revenue is an indicator of generation investment profitability, and thus is a measure of overall market performance as well as a measure of incentives to add generation to serve PJM Markets. Net revenue quantifies the contribution to capital cost received by generators from all PJM Markets. The net revenue calculations reflect environmental costs, unit class-specific, forced outage factors, annual planned outages, the hourly effects of ambient and cooling water temperature on plant performance and unforced capacity, and the reactive revenue requirements for each plant class. In general, net revenues have not been adequate to cover the first year fixed costs of a new entrant CT, a combined-cycle plant (CC) or a pulverized coal plant (CP) since the inception of PJM markets. (See Tables 13.2–13.4) Table 13.2 Net revenues – combustion turbine – economic dispatch (in dollars).

1999 2000 2001 2002 2003 2004 Average

Energy

Capacity

Reactive

Total

55,612 8,498 30,254 14,496 2,763 919 18,757

16,667 20,200 30,960 11,516 5,554 5,376 15,047

2,254 2,254 2,254 2,254 2,254 2,254 2,254

74,543 30,952 63,468 28,266 10,571 8,549 36,058

460

Electricity Market Reform Table 13.3 Net revenues – combined cycle – economic dispatch (in dollars).

1999 2000 2001 2002 2003 2004 Average

Energy

Capacity

Reactive

80,546 24,794 54,206 38,625 27,155 27,389 42,119

16,999 19,643 29,309 10,492 5,281 5,241 14,494

3,816 3,816 3,816 3,816 3,816 3,816 3,816

Total 101,361 48,253 87,331 52,933 36,252 36,446 60,429

Table 13.4 Net revenues – pulverized coal steam – economic dispatch (in dollars).

1999 2000 2001 2002 2003 2004 Average

Energy

Capacity

Regulation

Reactive

Total

92,935 108,624 95,361 96,828 159,912 124,497 113,026

17,798 20,755 30,862 11,493 5,688 5,537 15,355

5,596 3,492 1,356 2,118 2,218 1,399 2,697

2,988 2,988 2,988 2,988 2,988 2,988 2,988

119,317 135,859 130,567 113,427 170,805 134,420 134,066

For reference purposes, the 20-year levelized fixed costs of a combustion turbine are $72,207/MW, of a combined-cycle plant are $93,549/MW and of a pulverized coal steam unit are $208,247/MW. PJM’s LMPs are a direct measure of market performance. Price level is a good general indicator of market performance, although the number of factors influencing the overall level of prices means it must be analyzed carefully. For example, overall average prices subsume congestion and price differences over time. Annual price comparisons can be made in several ways and each must be interpreted carefully. Table 13.5 shows PJM average prices for the period from 1999 through 2004. Table 13.6 shows PJM average load-weighted prices that reflect hourly load levels. Table 13.7 shows PJM average fuel-cost-adjusted, load-weighted, average prices that reflect the impact of increased fuel costs. Table 13.5 shows that in 2004, PJM average prices increased by 10.8% over 2003. Table 13.6 shows that in 2004, when accounting for the actual loads paying these prices in each hour, PJM average prices increased by 7.5% over 2003. Finally, Table 13.7 shows that in 2004, when accounting both for actual hourly loads and for the fact that fuel costs increased for the units setting prices in each hour, PJM prices decreased by 4.2%. Energy Market results in PJM have reflected supply–demand fundamentals. While the existing structure of the Energy Market does not guarantee competitive outcomes, actual market performance results have been competitive. Nonetheless, given the structure of the Energy Market, tighter markets or a change in participant behavior are potential sources of competitive concern in the Energy Market.

461

The PJM Market Table 13.5 PJM average hourly LMP (Dollars/MWh): Calendar years 1998–2004. LMPs (in dollars)

1998 1999 2000 2001 2002 2003 2004

Year-to-Year changes

Average

Median

Standard deviation

Average LMP (%)

Median LMP (%)

Standard deviation (%)

21.72 28.32 28.14 32.38 28.30 38.27 42.40

16.60 17.88 19.11 22.98 21.08 30.79 38.30

31.45 72.42 25.69 45.03 22.40 24.71 21.12

N/A 30.4 ⫺0.6 15.1 ⫺12.6 35.2 10.8

N/A 7.7 6.9 20.3 ⫺8.3 46.0 24.4

N/A 130.3 ⫺64.5 75.3 ⫺50.3 10.3 ⫺14.5

Table 13.6 PJM load-weighted, average LMP (Dollars/MWh): Calendar years 1998–2004. Load-weighted average LMP (in dollars)

1998 1999 2000 2001 2002 2003 2004

Year-to-Year changes

Average

Median

Standard deviation

Average LMP (%)

Median LMP (%)

Standard deviation (%)

24.16 34.07 30.72 36.65 31.58 41.23 44.34

17.60 19.02 20.51 25.08 23.40 34.95 40.16

39.29 91.49 28.38 57.26 26.73 25.40 21.25

N/A 41.0 ⫺9.8 19.3 ⫺13.9 30.6 7.5

N/A 8.1 7.9 22.3 ⫺6.7 49.4 14.9

N/A 132.8 ⫺69.0 101.8 ⫺53.3 ⫺5.0 ⫺16.3

Table 13.7 PJM fuel-cost-adjusted, load-weighted LMP (Dollars/ MWh): Calendar years 2003–2004.

Average LMP Median LMP Standard deviation

2003 ($)

2004 ($)

Change (%)

41.23 34.95 25.40

39.49 34.47 20.81

⫺4.2 ⫺1.4 ⫺18.1

13.5.5. Market power mitigation in the energy market PJM’s market power mitigation goals have focused on market designs that promote competition (a structural basis for competitive outcomes) and on limiting market power mitigation to instances where market structure is not competitive and thus where market design alone cannot mitigate market power. In the PJM Energy Market, this occurs only in the case of local market power. PJM agreements include only very limited and targeted market power mitigation rules. PJM agreements have explicit market power mitigation rules only for the case of local market

462


power. These rules were proposed by the initial eight utility members of PJM and have been in effect since the beginning of PJM markets. Local market power exists when transmission constraints create small markets with limited generation resources that can operate to solve the constraints. In such cases, PJM applies a structural test to determine if the local market is competitive, applies a behavioral test to determine if generator offers exceed competitive levels and applies a market performance test to determine if such generator offers would affect the market price. PJM rules provide that the local market is analyzed, using a three pivotal supplier test, to determine if the local market structure is competitive. If the local market is competitive, no further action is taken. If the local market is not competitive, PJM limits the energy offers of generators that fail the three pivotal supplier test, that offer energy at a greater than competitive level and that would otherwise increase the market price. The offers are limited to the competitive level, incremental cost plus 10%, although such units always receive the higher of market prices or their offer cap. The levels of such mitigation have been low, consistently less than 1% of total MW or the total number of units. Mitigation for local market power does not have a significant, negative impact on unit net revenues.

13.6. Energy Imports and Exports PJM market participants import energy from, and export energy to, external regions on a continuous basis. Such transactions may fulfill long-term or short-term bilateral contracts or take advantage of price differentials. PJM’s external interfaces are the seams between PJM markets and markets in other regions. The integration of additional service territories into PJM in recent years has resulted in significant changes to its external interfaces. 13.6.1. Import and export activity PJM was a net importer of power during its early years, but became a net exporter of power in mid-2004 as a result of the integration of additional service territories. As of late 2004, PJM’s two largest net exporting interfaces were the interface between PJM and the New York ISO (NYISO) and the interface between PJM and Michigan Electric Coordinated System, each with about 20% of net exports.7 About 90% of the net import volume occurs at three interfaces: PJM/Illinois Power, PJM/Ohio Valley Electric Corporation and PJM/First Energy, each with about 30%. 13.6.2. Import and export issues Outside of LMP-based energy markets, energy is scheduled and paid for based on contract path despite the fact that the associated actual energy deliveries flow on the path of least resistance. Outside of LMP-based markets, potential or actual overloads on transmission facilities are managed by the direct curtailment of transactions that affect such transmission facilities. These curtailments are called transmission loading relief procedures (TLRs). At times, PJM has had to call TLRs to curtail flows coming from non-LMP markets because there were no market-based redispatch options available to PJM. 7

Interfaces are named after adjacent control areas. As is true of the control areas themselves, this naming convention does not imply anything about any company operating within the control areas.

The PJM Market

463

7,000,000 6,000,000 5,000,000

Gross imports Gross exports Net interchange

MWh

4,000,000 3,000,000 2,000,000 1,000,000 0

⫺2,000,000

January 1999 April 1999 July 1999 October 1999 January 2000 April 2000 July 2000 October 2000 January 2001 April 2001 July 2001 October 2001 January 2002 April 2002 July 2002 October 2002 January 2003 April 2003 July 2003 October 2003 January 2004 April 2004 July 2004 October 2004

⫺1,000,000

Fig. 13.4. PJM import and export transaction volume history: Calendar years 1999–2004.

The number of TLRs issued by PJM declined after the integration of additional control areas. The integrations meant that PJM could redispatch generating units to relieve constraints on facilities in the newly integrated areas where PJM had previously relied on TLRs for constraint control. The result was a drop in the number of TLRs called by PJM. After the creation of an LMP-based market in the Midwest ISO (MISO), an agreement was reached under which PJM and MISO redispatch generation in a coordinated manner to resolve constraints in both PJM and MISO that are affected by flows in both PJM and MISO. The “Joint Operating Agreement between the Midwest Independent Transmission System Operator, Inc., and PJM Interconnection, L.L.C.” (JOA) provides for relief of constraints on certain coordinated flowgates. The integration of new control areas into PJM affects the interfaces with adjacent markets. New interfaces also require new pricing points. When an LMP market is adjacent to markets that rely on contract path schedules rather than redispatch, the physical configuration creates the potential for power schedules (contracts), but not physical power flows, to bypass a control area. Outside of PJM’s LMP-based Energy Market, energy is scheduled and paid for based on contract path despite the fact that the associated actual energy deliveries flow on the path of least resistance. Loop flow can arise from transactions scheduled into, out of or around the PJM system on contract paths that do not correspond to the actual physical paths that the energy takes. Without management, a profitable game can be created and congestion within the LMP market can be affected. PJM has managed the contract path loop flow issue by creating pricing points that recognize the location of generation and the physical path of power flows. The price of imports into and exports from PJM are based on the actual physical path, to the extent possible, rather than the contract path.

464


13.7. Congestion Congestion occurs when available, low-cost energy cannot be delivered to all loads because of limited transmission capabilities. When the least cost available energy cannot be delivered to load in a transmission-constrained area, higher cost units must be dispatched in this constrained area to meet that load.8 The result is that the price of energy in the constrained area is higher than elsewhere because of the transmission limitations. LMPs reflect the price of the lowest cost resources available to meet loads, taking into account actual delivery constraints imposed by the transmission system. Thus LMP is an efficient way of pricing energy supply when transmission constraints exist. Congestion reflects this efficient pricing. As PJM integrated new transmission zones during 2004, the patterns of congestion changed, reflecting additional transmission and generation resources with new cost structures, load requirements and transmission system characteristics. In general, the increase in congestion costs associated with the integration of new areas into PJM was the result of making explicit the underlying price differentials that already existed. The inclusion of new facilities in PJM dispatch did not change the underlying physical transfer capabilities of the transmission system. Those physical limitations on the ability to transfer lower cost western power to the east, for example, existed prior to the integration of new areas to the west of the original PJM control area. The inclusion of these new areas did mean that these transmission limitations were priced explicitly and efficiently and were thus defined as congestion in an LMP system. The use of security constrained economic dispatch meant the more efficient use of these transmission facilities than was possible under the prior system that restricted all power transfers when a transmission limit was reached. 13.7.1. Overview Total congestion costs have ranged from 6 to 9% of PJM annual total billings since 2000. Total congestion costs were $808 million in calendar year 2004. Total congestion costs are a function both of the size of the PJM market footprint and of individual constraints.

Table 13.8 Total annual PJM congestion (Dollars (millions)): Calendar years 1999–2004.

1999 2000 2001 2002 2003 2004 Total

8

Congestion charges ($)

Percent increase

Total PJM billing ($)

Percent of PJM billing

53 132 271 430 499 808 2,193

N/A 149 105 59 16 62 N/A

N/A 2,300 3,400 4,700 6,900 8,700 N/A

N/A 6 8 9 7 9 N/A

This is referred to as dispatching units out of economic merit order. Economic merit order is the order of all generator offers from lowest to highest cost. Congestion occurs when loadings on transmission facilities mean that the next unit in merit order cannot be used and that a higher cost unit must be used in its place.

465

The PJM Market

In PJM, FTRs have been available to firm point-to-point and network service transmission customers as a hedge against congestion costs since the inception of locational energy pricing on April 1, 1998. FTRs have been paid at from 95 to 100% of the target allocation level. Customers holding FTRs were thus protected from paying most congestion costs. To provide an approximate indication of the geographic dispersion of congestion costs, LMP differentials are shown for the 2001 to 2004 period (Fig. 13.5), for control zones in the PJM Mid-Atlantic and Western Regions as they existed at year end 2004. If the zonal LMP differential is positive, average zonal LMP was greater than the LMP at the PJM West Hub while if the zonal LMP differential is negative, average zonal LMP was less than the LMP at the PJM West Hub. Persistent congestion in areas within PJM and the overall level of congestion costs suggest the importance of PJM’s continuing efforts to improve the sophistication of its congestion analysis. Congestion analysis is central to implementing the FERC order to develop an approach identifying areas where investments in transmission would relieve congestion where that congestion might enhance generator market power and where such investments are needed to support competition.9 PJM implements a regional transmission expansion planning protocol (RTEPP) in order to ensure that transmission projects required for system reliability are identified and that the transmission owners build the identified projects. Per direction from FERC, the PJM regional transmission planning protocol has been expanded to include economic planning. PJM will, when appropriate, initiate upgrades or expansions of the transmission system to enhance the economic and operational efficiency of wholesale electricity markets in PJM. PJM’s economic $15 2001 2002 2003 2004

$10

$5

$0

⫺$5

EP A

Y A D

Ed C om

D PL

O

PL

EC A

JC

PS EG

PE C O

-E d et M

PP L

C EL E

B G E

PE N

PE PC

⫺$15

O

⫺$10

Fig. 13.5. Annual zonal LMP differences (Reference to Western Hub): Calendar years 2001–2004. 9

96 FERC ¶61,061 (2001).

466


planning process identifies transmission upgrades needed to address unhedgeable congestion. PJM defines unhedgeable congestion as the cost of congestion attributable to the portion of load affected by a transmission constraint that cannot be supplied by economic generation or hedged with available FTRs.10 First, market forces are relied upon through the opening of a 1-year market window during which merchant solutions are solicited through the introduction of incentives and the posting of relevant market data. If market forces do not resolve unhedgeable congestion within an appropriate time period, PJM will determine, subject to cost–benefit analysis, transmission solutions that will be implemented through the RTEPP. 13.8. FTRs and ARRs In PJM, FTRs have been available to firm point-to-point and network service transmission customers as a hedge against congestion costs since the inception of locational energy pricing on April 1, 1998. These firm transmission customers have access to FTRs because they pay the costs of the transmission system that enables firm energy delivery. Firm customers receive requested FTRs to the extent that they are consistent both with the physical capability of the transmission system and with other eligible customers’ FTR requests. Effective from June 1, 2003,11 PJM replaced the allocation of FTRs with an allocation of ARRs and an associated Annual FTR Auction. The process for allocating ARRs is identical to the previous process for allocating FTRs, but the revenues received for the allocated ARRs are based on the results of the Annual FTR Auction, in which any market participant can buy FTRs. Firm transmission customers have the option either to take ARRs or to take the underlying FTRs through a process called self-scheduling. PJM also runs monthly auctions designed to permit bilateral FTR sales and to allow eligible participants to buy any residual system FTRs. For the 2003 to 2004 planning period, PJM introduced 24-hour FTRs into the monthly auctions. At the same time, PJM also added annual and monthly FTR options. Unlike standard FTRs, the options can never be a financial liability. ARRs and FTRs are both financial instruments that entitle the holder to receive revenues or to pay charges based on nodal price differences. ARRs provide holders with revenues or charges based on the locational price difference between ARR sources (origins) and sinks (destinations) determined in the Annual FTR Auction.12 In other words, ARR revenues are a function of FTR auction participants’ expectations of locational price differences in the Day-Ahead Energy Market. FTRs provide holders with revenues or charges based on the locational price differences actually experienced in the Day-Ahead Energy Market. ARR and FTR holders do not need to deliver energy to receive ARR or FTR credits, and neither instrument represents a right to the physical delivery of power. Both can, however, help protect load-serving entities (LSEs) and other market participants from congestion costs in the PJM Day-Ahead Energy Market. Market participants can also hedge against real-time congestion by matching real-time energy schedules with day-ahead energy schedules. ARRs were available throughout the PJM Mid-Atlantic Region for the 2004 to 2005 planning period, while both ARRs and direct allocation FTRs were available to eligible market participants in the AP and ComEd Control Zones. Eligible customers in the AEP and DAY Control Zones received phase-in FTRs to carry them to the start of the next planning period. 10

104 FERC ¶61,124 (2003). 87 FERC ¶61,054 (1999). 12 These nodal prices are a function of the market participants’ annual FTR bids and binding transmission constraints. An optimization algorithm selects the set of feasible FTR bids that produces the most net revenue. 11

The PJM Market

467

13.8.1. FTR/ARR market structure ARR demand is limited by the total amount of network and long-term, firm point-to-point transmission service. ARR supply is limited by the capability of the transmission system to simultaneously accommodate the set of requested ARRs and numerous combinations of ARRs are feasible. PJM market rules automatically reassign ARRs and their associated revenue when load switches among LSEs. Individual MW of load may be reassigned multiple times over a period. Under the Annual FTR Auction, there is no limit on demand. FTR supply is limited by the capability of the transmission system to accommodate simultaneously the set of requested FTRs, and numerous combinations of feasible FTRs.

13.8.2. FTR market performance For the 2004 to 2005 planning period, just over 80% of Mid-Atlantic Region annual FTRs were purchased for less than $1/MWh and 90% for less than $2/MWh. The overall average prices paid for annual FTR obligations were $1.27/MWh for 24-hour, $0.16/MWh for on-peak and $0.13/MWh for off-peak FTRs. Comparable prices for the 2003 to 2004 planning period were $1.09/MWh for 24-hour, $0.34/MWh for on-peak and $0.15/MWh for off-peak FTRs. Annual and Monthly FTR auction revenue is allocated to ARR holders based on ARR target allocations. Congestion revenues are allocated to FTR holders based on FTR target allocations. ARRs were 100% revenue adequate during the 2003 to 2004 and the 2004 to 2005 planning periods. ARR holders received credits valued at $311 million during the 2003 to 2004 planning period, with an average hourly ARR credit of $1.23/MWh. ARR holders will receive credits valued at $345 million during the 2004 to 2005 planning period, with an average hourly ARR credit of $1.17/MWh. FTRs were 98% revenue adequate during the 2003 to 2004 planning period, receiving credits valued at $680 million. The Annual ARR Allocation and Annual FTR Auction together provide long-term, firm transmission customers with a mechanism to hedge congestion and provide all eligible market participants increased access to long-term FTRs. The Annual FTR Auction allows a market valuation of FTRs that is consistent with the most efficient use of such financial instruments. The FTR auction process results have been competitive and succeeded in providing all qualified market participants with equal access to FTRs. The rules remove a potential barrier to competition by explicitly providing that beneficial ARRs follow load as load shifts among load-serving entities in response to market forces.

13.9. Capacity Market 13.9.1. Background PJM and its members have long relied on capacity obligations as one of the methods to ensure reliability. Before the introduction of wholesale competition or retail restructuring, the original PJM members had determined their loads and related capacity obligations annually and several years in advance. Combined with state regulatory requirements to build and incentives to maintain adequate capacity, this system created a reliable pool, where capacity and energy were adequate to meet customer needs and where capacity costs were borne equitably by members and their loads.

468


Capacity obligations continue to be critical to maintaining reliability and contribute to the effective, competitive operation of PJM Energy Markets. Adequate capacity resources, equal to or greater than expected load plus a reserve margin, help to ensure that energy is available on even the highest load days. For the PJM Capacity Markets, forecast peak-loads plus a system reserve margin are the basis for capacity obligations. An LSE’s unforced capacity obligation is its forecast peak load plus the reserve margin. In PJM, each LSE must own or purchase capacity resources greater than, or equal to, its capacity obligation. The unforced capacity position of every LSE is calculated daily to determine if any LSE is short of capacity resources. Deficient entities must contract for capacity resources to satisfy their deficiency. Any LSE that remains deficient must pay a substantial penalty. On January 1, 1999, in response to retail restructuring requirements, PJM introduced a transparent, PJM-run market in capacity credits. New retail market entrants needed a way to acquire capacity credits to meet the capacity obligations associated with competitively gained load. Existing utilities needed a way to sell excess capacity credits when load was lost to new competitors. The PJM Capacity Credit Market (CCM) provides mechanisms to balance supply of and demand for capacity unmet by the bilateral market or self-supply. The PJM Capacity Credit Market consists of the Daily, Interval, Monthly and Multimonthly Capacity Credit Markets. The Capacity Credit Market is intended to provide a transparent, market-based mechanism for competitive retail LSEs to acquire the capacity resources needed to meet their capacity obligations and to sell capacity resources when no longer needed to serve load. The PJM Daily Capacity Credit Market permits LSEs to match capacity resources with short-term shifts in retail load while Interval, Monthly and Multimonthly Capacity Credit Markets provide mechanisms to match longer term obligations with capacity resources.

13.9.2. Rationale Wholesale electric power markets, apparently without exception, are affected by externally imposed reliability requirements. A regulatory authority external to the market makes a determination as to the acceptable level of reliability, typically measured as an acceptable loss of load probability level. This level of reliability is enforced through a requirement to maintain a target level of installed or unforced capacity, which, based on planning models, is considered to be a level that will produce the desired loss of load probability. The requirement to maintain a target level of installed capacity can be enforced via a variety of mechanisms including government construction of generation, full requirements contracts with developers to construct and operate generation, state utility commission mandates to construct capacity, or capacity markets of various types. The target level of installed capacity is typically measured using a reserve margin, the proportion of capacity required in excess of some measure of expected peak load. Regardless of the enforcement mechanism, the exogenous requirement to construct capacity in excess of what is constructed in response to energy market signals has an impact on energy markets. The impact of having capacity in excess of the equilibrium level likely to result from the operation of an energy market alone is to reduce the level and volatility of prices and to reduce the duration of high prices. This in turn reduces net revenues to generation owners, which reduces the incentive to invest. A capacity market is a formal market-based mechanism used to allocate the costs of maintaining the level of capacity required to maintain the reliability target. Ideally, a capacity market would include a mechanism for equilibrating energy and capacity market revenues such

The PJM Market

469

that, in equilibrium, generators receive a market-based return for investing in capacity from all markets taken together. A capacity market is also an explicit mechanism for valuing capacity and is preferable to non-market and non-transparent mechanisms for that reason. For example, if the primary mechanism for maintaining the target level of reliability were that individual state utility commissions required that jurisdictional utility companies build capacity under rate base rate of return regulation, this would be a non-market and nontransparent mechanism. One immediate consequence of such a mechanism would be that potential competitors to the jurisdictional utility companies, that is, merchant generators that are not regulated by the states, could not compete in the wholesale power markets. Such generators would be missing a key revenue stream because their net revenues would depend entirely on the energy market and would have no capacity market component. PJM proposed a modified capacity market construct, the Reliability Pricing Model (RPM) in a filing with FERC on August 30, 2005. 13.9.3. Capacity resources Capacity resources are defined as MW of net generating capacity meeting PJM-specific criteria. They may be located within or outside of PJM, but they must be committed to serving load within PJM. All capacity resources must pass tests regarding the capability of generation to serve load and to deliver energy. This latter criterion requires adequate transmission service. The sale of a generating unit as a capacity resource within the PJM Control Area entails obligations for the generation owner. These requirements, listed below, are essential to the definition of a capacity resource and contribute directly to system reliability. These requirements define the capacity product that generation owners sell in the capacity markets. 13.9.3.1. Energy recall right PJM rules specify that when a generation owner sells capacity resources from a unit, the seller is contractually obligated to allow PJM to recall the energy generated by that unit if the energy is sold outside of PJM. This right enables PJM to recall energy exports from capacity resources when it invokes emergency procedures. The recall right establishes a link between capacity and actual delivery of energy when it is needed. Thus, PJM can call upon energy from all capacity resources to serve load within the Control Area. When PJM invokes the recall right, the energy supplier is paid the PJM real-time energy market price. 13.9.3.2. Day-ahead energy market offer requirement Owners of PJM capacity resources are required to offer their output into PJM’s Day-Ahead Energy Market. When LSEs purchase capacity, they ensure that resources are available to provide energy on a daily basis, not just in emergencies. Since day-ahead offers are financially binding, PJM capacity resource owners must provide the offered energy at the offered price if the offer is accepted in the Day-Ahead Energy Market. This energy can be provided by the specific unit offered, by a bilateral energy purchase, or by an energy purchase from the Real-Time Energy Market. 13.9.3.3. Deliverability To qualify as a PJM capacity resource, energy from the generating unit must be deliverable to load in the PJM Control Area. Capacity resources must be deliverable to the total system

470


load. In addition, for external capacity resources used to meet an obligation within PJM, capacity and energy must be deliverable to the metered boundaries of the RTO through firm transmission service. 13.9.3.4. Generator outage reporting requirement Owners of PJM capacity resources are required to submit historical outage data to PJM.

13.9.4. The operation of PJM capacity markets Each organization serving PJM load must own or acquire capacity resources to meet its respective capacity obligations. LSEs can acquire capacity resources by entering into bilateral agreements with capacity internal or external to PJM, by participating in the PJM-operated Capacity Credit Market or by purchasing or constructing generation. Collectively, all arrangements by which LSEs acquire capacity are known as the Capacity Market. 13.9.4.1. Overview Key measures of PJM Capacity Market structure and performance include concentration ratios, prices, outage rates and reliability. While there remain serious market power concerns based on market structure issues in the Capacity Market and there was a significant exercise of market power in 2001, modified rules and market conditions have otherwise resulted in reasonably competitive outcomes in PJM Capacity Markets. The analysis of capacity markets begins with market structure which provides the framework for the actual behavior or conduct of market participants. The analysis also examines participant behavior in the context of market structure. In a competitive market structure, market participants are constrained to behave competitively. In a competitive market structure, competitive behavior is profit maximizing behavior. Finally, the analysis examines market performance results. The ultimate test of the markets is the actual performance of the market, measured by price and the relationship between price and marginal cost. For example, at times market participants behave in a competitive manner even within a non-competitive market structure. This may result from the relationship between supply and demand and the degree to which one or more suppliers are singly or jointly pivotal even in a highly concentrated market. This may also result from a conscious choice by market participants to behave in a competitive manner based on perceived regulatory scrutiny or other reasons, even when the market structure itself does not constrain behavior. 13.9.4.2. Capacity market structure PJM Daily Capacity Credit Markets exhibit moderate concentration while PJM longer term markets exhibit high concentration levels. Electricity distribution companies (EDCs) and their affiliates accounted for a large share (generally 75 to 80%) of the PJM Capacity Markets’ load obligations. 13.9.4.3. Capacity market performance Figure 13.6 shows prices and volumes in PJM’s Daily and longer term Capacity Credit Markets from 2000 through 2004. Since the interval system was introduced in June 2001, overall volume in the Monthly and Multimonthly Capacity Credit Markets has increased and prices in both the daily and longer term markets have declined and remained relatively stable with the

471

Daily CCM Monthly CCM Monthly weighted-average price Daily weighted-average price

175,000 150,000

$250

$200 125,000 $150

100,000 75,000

$100

50,000 $50 25,000

Weighted-average capacity clearing price ($/MW-day): Lines

Volume of credits transacted (MW): Bars

The PJM Market

$0 July 2004

October 2004

April 2004

January 2004

October 2003

July 2003

April 2003

January 2003

July 2002

October 2002

April 2002

January 2002

July 2001

October 2001

April 2001

January 2001

July 2000

October 2000

April 2000

January 2000

0

Fig. 13.6. PJM daily and monthly capacity credit market (CCM) performance: Calendar years 2000–2004.

exception of the second interval of 2004. Although daily volume has risen to pre-June 2001 levels, capacity obligations have increased by more than 25%. The share of load obligation traded in the PJM Daily Capacity Market has declined since the introduction of Interval Markets, while the share of load obligation traded in Monthly and Multimonthly Capacity Markets has increased. Daily capacity market volume declined from 2.5% of average obligation in 2000 to 1.6% in the last two intervals of 2003. In comparison, average daily capacity credit market volume in 2004 increased to 1,062 MW from 907 MW in 2003, but as a percent of obligation, 2004 volume remained approximately the same at 1.4% of obligation. Monthly and Multimonthly Capacity Market volume increased from 3.0% of obligation in 2000 to 5.2% of average obligation in the last two intervals of 2003. In comparison, average monthly and Multimonthly Capacity Credit Market volume in 2004 increased to 3,966 MW from 3,435 MW in 2003, but 2004 volume as a percent of obligation declined slightly to 5.1% from 5.2% in 2003. With the integration of the AEP and DAY Control Zones, by virtue of their participation, total volume traded has increased. Nonetheless, because of the new participants’ reliance on their own resources, volume as a percent of obligation has declined once again, with values approaching 1% for the Daily Capacity Credit Market and 4% for the Monthly and Multimonthly Capacity Credit Markets since October 2004. With the exception of a price spike in the daily capacity market in the summer of 2004, daily capacity market prices have trended close to zero and the prices in the longer term capacity markets have trended down. The capacity market prices have been, with the noted exception in 2001, the result of market fundamentals; overall the PJM capacity market is long. The volume-weighted average price for all Capacity Credit Markets was $52.86/MW-day in 1999, $60.55 in 2000, $95.34 in 2001, $33.40 in 2002, $17.51 in 2003 and $17.74 in 2004. The volumeweighted average price for the Monthly and Multimonthly Capacity Credit Markets was $70.66/MW-day in 1999, $53.16 in 2000, $100.43 in 2001, $38.21 in 2002, $21.57 in 2003 and

472


$17.88 in 2004, while the price in the Daily Capacity Credit Market averaged $3.63/MW-day in 1999, $69.39 in 2000, $87.98 in 2001, $0.59 in 2002, $2.14 in 2003 and $17.21 in 2004. 13.9.4.4. Capacity market summary Given the basic features of market structure in the PJM Capacity Market, including high levels of concentration, the relatively small number of non-affiliated LSEs, the capacity-deficiency penalty structure facing LSEs, supplier knowledge of the penalty structure and supplier knowledge of aggregate market demand if not individual LSE demand, the likelihood of the exercise of market power is high. Market power is endemic to the structure of PJM Capacity Markets. Supply and demand fundamentals have, with the noted exception in 2001, offset these market structure issues in the PJM Capacity Market, producing competitive results. The proposed redesign of the PJM capacity markets in the RPM filing directly addresses the market power issues via an explicit set of market power mitigation rules.

13.10. Ancillary Service Markets The FERC defined six ancillary services in Order 888: (1) scheduling, system control and dispatch service; (2) reactive supply and voltage control from generation sources service; (3) regulation and frequency response service; (4) energy imbalance service; (5) operating reserve – spinning reserve service; and (6) operating reserve – supplemental reserve service.13 PJM currently provides regulation and spinning through market-based mechanisms and PJM also provides energy imbalance service through the Real-Time Energy Market. PJM provides the remaining ancillary services on a cost basis. Regulation matches generation with very short-term changes in load by moving the output of selected generators up and down via an automatic control signal. Regulation is provided, independent of economic signal, by generators with a short-term response capability (less than 5 minutes). Longer term deviations between system load and generation are met via primary and secondary reserves and generation responses to economic signals. Spinning reserve is an ancillary service defined as generation that is synchronized to the system and capable of producing output within 10 minutes. Spinning reserve can, at present, be provided by a number of sources, including steam units with available ramp, condensing hydroelectric units, condensing CTs and CTs running at minimum generation. All of the units that participate in the Spinning Reserve Market are categorized as either Tier 1 or Tier 2 spinning. Tier 1 resources are those units that are online following economic dispatch and able to respond to a spinning event by ramping up from their present output. All units operating on the PJM system are considered potential Tier 1 resources, except for those explicitly assigned to Tier 2 spinning. Tier 2 resources include units that are backed down to provide spinning capability and condensing units synchronized to the system and available to increase output. PJM introduced a market for spinning reserve on December 1, 2002. Before the Spinning Reserve Market, Tier 1 spinning reserve had not been compensated directly and Tier 2 spinning reserve had been compensated on a unit-specific, cost-based formula. Under the Spinning Reserve Market rules, Tier 1 resources are paid when they respond to an identified spinning event as an incentive to respond when needed. Tier 1 spinning

13

75 FERC ¶61,080 (1996).

The PJM Market

473

payments or credits are equal to the integrated increase in MW output above economic dispatch from each generator over the length of a spinning event, multiplied by the spinning energy premium less the hourly integrated LMP. The spinning energy premium is defined as the average of the 5-minute LMPs calculated during the spinning event plus $50/MWh. All units called on to supply Tier 1 or Tier 2 spinning have their actual MW monitored. Tier 1 units are not penalized if their output fails to match their expected response as they are only compensated for their actual response. Tier 2 units assigned spinning by market operations are compensated whether or not they are actually called on to supply spinning so they are penalized if their MW output fails to meet their assignment. Both the Regulation and Spinning Reserve Markets are cleared on a real-time basis. A unit can be selected for either spinning reserve or regulation or neither, but it cannot be selected for both. The Spinning Reserve and Regulation Markets are cleared simultaneously and cooptimized with the Energy Market to minimize the cost of the combined products. PJM does not provide a market for reactive power, but does ensure its adequacy through member requirements and scheduling. Generation owners are paid according to the FERCapproved reactive revenue requirements. Charges are allocated to network customers based on their percentage of load, as well as to point-to-point customers based on their monthly peak usage. PJM generally operates two Regulation Markets: one for the Mid-Atlantic Region and a second for the Western Region. However, PJM is, in the summer of 2005, experimenting with having a single Regulation Market over the entire PJM footprint. PJM operates four Spinning Reserve Markets in four spinning zones: the PJM MidAtlantic Region spinning zone, the ComEd spinning zone, the AP-AEP-DAY spinning zone and the Dominion spinning zone. Key measures of structure, conduct and performance for the identified Regulation Markets and the Spinning Reserve Markets include concentration of ownership, price and availability. These markets have functioned effectively and produced competitive results because the markets clear with cost-based offers rather than price-based offers. The structure of each of the Regulation and Spinning Reserve Markets has been evaluated and, with the exception of the Regulation Market in the PJM Mid-Atlantic Region, these markets are not structurally competitive as they are characterized by high levels of supplier concentration and inelastic demand. As a result, these Ancillary Service Markets are operated as markets with market-clearing prices and with offers based on the marginal cost of producing the service plus a margin. The conduct of market participants within these market structures has been consistent with competition, and the market performance results have been competitive. The Regulation Market in the PJM Mid-Atlantic Region is cleared based on participants’ price offers. All suppliers are paid the market-clearing price, which is a function of the supply curve and PJM-defined demand. The supply curve is offered MW and their associated offer price, which is a combination of unit-specific offers plus opportunity cost (OC)14 as calculated by PJM. The Regulation Market in the West is cleared on cost-based offers because the market is not structurally competitive. The price of regulation in the West was based on unit-specific, cost-based offers plus unit-specific OC. The cost-based regulation offer prices are defined to be the unit-specific incremental cost of providing regulation plus a margin of $7.50/MWh plus OC calculated by PJM. 14

As used here, the term “opportunity cost” (OC) refers to the estimated lost opportunity cost (LOC) that PJM uses to create a supply curve on an hour-ahead basis. The term, “lost opportunity cost,” refers to opportunity costs included in payments to generation owners.

474


The Spinning Reserve Markets in PJM are cleared based on cost-based offers because these markets are not structurally competitive. The cost-based offers for spinning reserve include incremental cost plus a margin of $7.50/MWh plus OC calculated by PJM. Prices for spinning were market-clearing prices determined by supply and PJM-defined demand.

13.11. Conclusion The experience of PJM demonstrates that competitive wholesale power markets can work if properly designed and managed. While PJM has evolved steadily and continued to improve its market design, PJM began with all the essential elements of a wholesale power market. From the beginning on April 1, 1999, PJM has been bid based, security constrained and economically dispatched with locational marginal pricing in a context of open access transmission with financial hedges against congestion. PJM has also demonstrated the efficient management of the grid. Grid management requires fixed costs, operating costs and labor costs. The costs of buildings, hardware, the essential software of markets and labor are not a linear function of the MWh size of markets. PJM’s costs per MWh increased with early market development between 1999 and 2003 and have declined since then as PJM’s growth in market size has exceeded the growth in costs. The expected cost of PJM’s services, excluding FERC fees, is about $.36/MWh for 2006 or less than 1% of the average LMP.15 Figure 13.7 illustrates this historical pattern for PJM as well as providing a comparison of annual operating costs for U.S. ISO/RTOs. PJM provides an example of the successful creation and growth of a market and market institutions. This success began with a history of cooperation among members and the

1.20

Dollars/MW hr

1.00 0.80 0.60 0.40 0.20 0.00 1999 ERCOT

2000

2001

PJM

ISO-NE

2002

2003

NYISO

2004 CAISO

2005 MISO

Proposed 2006 SPP

Fig. 13.7. ISO/RTO Tariff rates.

15

The year to date LMP through September 2005 was $54.69/MWh. Thus, PJM’s costs were 0.7% of that LMP.

The PJM Market

475

creation of a power pool, but success required PJM and its members to meet significant challenges in the transition from pool to energy market and from energy market to a complete set of markets with complex but consistent interactions. The cooperation of PJM members and PJM was institutionalized in a set of clearly delineated market rules and in a set of PJM market and operating practices, all based on economic theory and electric network realities. This set of market rules and consistent operating practices is the foundation of the market and because its characteristics make it scalable, this set of rules and practices made the significant and successful integrations of 2004 and 2005 possible. Both PJM and PJM members have developed with the markets. PJM and PJM members have maintained a commitment to market solutions throughout. Nonetheless, PJM continues to face challenges. The net revenue performance of the markets over 6 years illustrates that additional market modifications are necessary if PJM is to pass the ultimate test of a market, the successful provision of long-term incentives to invest. A market design cannot be deemed truly successful until it results in the retirement and replacement of a significant portion of the existing investment in generating assets, based on incentives endogenous to the market design. While there continues to be an active debate, PJM has filed a modified locational capacity market design intended to address a key aspect of the longterm investment incentive. In addition, PJM faces challenges in areas including how to implement scarcity pricing, how to mitigate market power, how to increase the level of demand response, how to provide appropriate incentives for transmission investment and how to manage seams with other RTOs and non-market areas. PJM Market milestones. Year

Month

Event

1996 1997

April April November April January March March April April June June July June April June

FERC Order 888 Offer-based Energy Market FERC approval of PJM ISO status Cost-based Energy LMP Market Daily Capacity Market FERC approval of market-based rates for PJM Monthly and Multimonthly Capacity Market Competitive energy LMP Market FTR Market Regulation Market Day-Ahead Energy Market Customer Load-Reduction Pilot Program First PJM Emergency and Economic Load-Response Programs Integration of the AP Control Zone into PJM Western Region Second PJM Emergency and Economic Load-Response Programs Spinning Reserve Market FERC approval of full PJM RTO status Annual FTR Auction Integration of the ComEd Control Area into PJM Integration of AEP Control Zone into PJM Western Region Integration of DAY Control Zone into PJM Western Region

1998 1999

2000

2001 2002

2003 2004

December December May May October October

476


Acknowledgment The material in this chapter draws heavily on the Market Monitoring Unit’s PJM 2004 State of the Market Report. That report also includes substantial additional detail on the material presented here. I want to acknowledge the fact that the State of the Market Report was based on substantial work by all the members of the PJM MMU, including Kevin Bazar, Jerry Bell, Tom Blair, Susan Cawley, Bridgid Cummings, Andrew Engle, Beatrice Gockley, Howard Haas, Ellen Krawiec, Don Kujawski, Yan Lin, Mark Million, Jack O’Neill, David Picarelli, Frank Racioppi, Paul Scheidecker and Tom Zadlo.

Bibliography Key FERC Orders re competitive electricity markets Order No. 888, “Transmission Open Access. Promoting Wholesale Competition Through Open Access Nondiscriminatory Transmission Services by Public Utilities; Recovery of Stranded Costs by Public Utilities and Transmitting Utilities”, April 24, 1996. Order No. 888-A, “Reaffirming & Clarifying Terms of Order 888 in Regards to Open Access Transmission Services & Recovery of Stranded Costs”, March 4, 1997. Order No. 888-B, “Order Affirming, With Certain Clarifications, The Fundamental Calls Made in Order 888-A in Regard to Promoting Wholesale Competition Through Open Access Non-discriminatory Services etc.”, November 25, 1997. Order No. 889-B, “Final Order: Order Denying Rehearing of Order 889-A in Regards to Open Access SameTime Information System & Standards of Conduct (OASIS)”, November 25, 1997. Order No. 888-C, “Order on Rehearing, Denying Otter Power Co. Rehearing Request”, January 20, 1998. Order No. 2000, “Final Rule, Regional Transmission Organization (RTO)”, December 20, 1999. Order No. 2000A, “Regional Transmission Organization (RTO), Order on Rehearing”, February 25, 2000. Docket Nos. EL01-118-000, EL01-118-001, “Market Behavior Rules”, November 17, 2003. Docket No. EL01-18-000, “Order Clarifying Market Behavior Rules”, May 19, 2004.

Key FERC orders re PJM electricity markets Docket Nos. OA97-261-000, ER97-1082-000, “Order Accepting For Filing and Suspending Proposed PoolWide and Single-System Holding Company Open Access Transmission Tariffs and Revised Tariffs, and Deferring Further Action”, February 28, 1997. Docket Nos. OA97-261-000 et al. “Order Conditionally Accepting Open Access Transmission Tariff and Power Pool Agreements, Conditionally Authorizing Establishment of an Independent System Operator and Disposition of Control over Jurisdictional Facilities and Denying Rehearings”, November 25, 1997. Docket No. ER99-196-000, “Order Accepting, as Revised, PJM Capacity Credit Markets”, January 13, 1999. Docket No. ER97-3729-000, “Order Approving PJM Supporting Companies’ Request for Market-Based Pricing Authority”, March 10, 1999. Docket No. ER98-3527-000, “Order Approving Market Monitoring Plan as Modified”, March 10, 1999. Docket No. ER99-2028-000, “Order Conditionally Accepting Compliance Filing”, April 13, 1999 (Tariff revisions for FTR Auctions). Docket No. ER00-1630-000, Letter Order accepting amendments to the Open Access Transmission Tariff and Operating Agreement providing for market-based pricing and new market rules for regulation service, April 12, 2000. Docket No. ER00-1849-000, Letter Order accepting revisions to the Open Access Transmission Tariff and Operating Agreement, containing a modified two-settlement system that incorporates increment and decrement bids into the day-ahead energy market, May 18, 2000. Docket No. ER00-3090-000, “Order Accepting and Suspending Filing”, July 26, 2000 (Load Reduction Pilot Program).

The PJM Market

477

Docket No. ER01-1671-000, “Order Accepting Tariff Sheets as Modified” May 30, 2001 (Emergency and Economic Load-Response Programs). Docket No. RT01-2-000, “Order Provisionally Granting RTO Status”, July 12, 2001. Docket Nos. ER02-2519-000, ER02-2519-001, ER02-2519-002, “Order Accepting Spinning Reserve Market”, October 31, 2002. Docket Nos. RT01-2-001, RT01-2-002, “Order Granting PJM RTO Status, Granting In Part and Denying In Part Requests For Rehearing, Accepting And Directing Compliance Filing, and Denying Motion For Stay”, December 20, 2002. Docket No. ER03-406-000, “Order Accepting For Filing Proposed Tariff Changes, As Modified”, March 12, 2003 (Annual FTR Auction).

Additional PJM resources Market Monitoring Unit, PJM Interconnection. 1999 State of the Market Report, June 2000. Market Monitoring Unit, PJM Interconnection. 2000 State of the Market Report, June 2001. Market Monitoring Unit, PJM Interconnection. 2001 State of the Market Report, June 2002. Market Monitoring Unit, PJM Interconnection. 2002 State of the Market Report, March 5, 2003. Market Monitoring Unit, PJM Interconnection. 2003 State of the Market Report, March 4, 2004. Market Monitoring Unit, PJM Interconnection. 2004 State of the Market Report, March 8, 2005. Additional analytical reports produced by the Market Monitoring Unit can be found at http:// www.pjm.com/markets/market-monitor/reports.html. Additional resources about PJM Markets can be found at www.pjm.com.

Additional electricity market resources Crampton, P. and Steven S. (2005). A capacity market that makes sense. Electricity Journal, 18 (August/ September), 43–54. Hobbs, Benjamin F., Javier I. and Steven E. Stoft. (2001). Installed capacity requirements and price caps: oil on the water, or fuel on the fire? Electricity Journal, 14, July 43–54. Hogan, William W. (1998). Competitive Electricity Market Design: A Wholesale Primer. John F. Kennedy School of Government, Harvard University, Cambridge, MA.Available at http://www.whogan.com/. Hogan, William W. (1992). Contract Networks for Electric Power Transmission., John F. Kennedy School of Government, Harvard University, Cambridge, MA. Available at http://www.whogan.com/. Hogan, William W. and Brendan J. Ring. (2003). On Minimum-Uplift Pricing for Electricity Markets. John F. Kennedy School of Government, Harvard University, Cambridge, MA. Available at http://www. whogan.com/. Kahn, Edward P., Peter Cramton, Robert Porter and Richard Tabors. (2001). Pricing in the California Power Exchange Electricity Market: Should California Switch from Uniform Pricing to Pay-as-Bid Pricing? Blue Ribbon Panel Report. Commissioned by the California Power Exchange, January 23, 2001, available at http://www.ksg.harvard.edu/hepg/Papers.htm. Lambert, Jeremiah D. (2001). Creating Competitive Power Markets: the PJM Model. PennWell Corporation. Tulsa, OK. Schweppe, Fred C., Michael C., Caramanis, Richard D., Tabors and Roger E. Bohn. (1988). Spot Pricing of Electricity. Kluwer Academic, Boston. Stoft, S. (2002). Power System Economics. IEEE Press, Piscataway, NJ.


Chapter 14 Independent System Operators in The USA: History, Lessons Learned, and Prospects* RICHARD O’NEILL,1 UDI HELMAN,1 BENJAMIN F. HOBBS2 AND ROSS BALDICK3 1 US Federal Energy Regulatory Commission, Washington, DC; 2Department of Geography and Environmental Engineering, The John Hopkins University, Baltimore, USA; 3Department of Electrical and Computer Engineering, The University of Texas at Austin, Texas, USA

Summary This chapter provides a cross-cutting perspective on the currently operating Independent System Operator (ISO) markets and Federal Energy Regulatory Commission (FERC’s) broader public policy agenda. The chapter provides perspective on regulatory reforms in the USA that began the movement towards organized, competitive markets. These reforms were accompanied by intense debate over the appropriate design of such markets, including whether and how to mitigate market power. The chapter examines what motivated these debates and how different ISO regions experimented with alternative approaches. It explains why certain design features eventually were considered best practices and what design issues remain for consideration. It examines what measures of market performance have been implemented and how they should be interpreted. Finally, the chapter provides a summary of the state of market design and market performance across each US ISO market.

14.1. Introduction The goal of regulatory reform of the electricity industry in the USA and around the world has been to achieve greater efficiency in provision of electricity through market competition. Whereas generation was widely believed before the 1970s to be part of a natural monopoly, technological developments since then have made scale economies in generation construction and operation less important, particularly in large, well-interconnected markets. At the same time, some very costly and inefficient investments in generation approved under

*Portions of this chapter were reprinted, with permission, from Ross Baldick, Udi Helman, Benjamin F. Hobbs and Richard P. O’Neill, “Design of Efficient Generation Markets,” Proceedings of the IEEE, Special Issue on Electric Power Systems: Engineering and Policy, 93(11), November 2005, © 2005 IEEE.

479

480


monopoly regulation suggested that competitive markets could make better decisions. Thus, competition among suppliers of generation came to be seen as a viable alternative for spurring efficiency in electricity production. Concurrently, increasing transmission interconnection between utilities for purposes of reliability and improved capability to co-ordinate system operations suggested the ability to better use the existing grid by reducing impediments to obtaining transmission service. Regulatory reform started in the USA in 1978 with the passage of the Public Utility Regulatory Policies Act (PURPA), and began overseas with power market restructurings in Chile (1982) and in England & Wales (1989). In the US, regulatory reform was accelerated over the latter half of the 1990s with the advent of the open access transmission regime in 1996 (FERC 1996) and the subsequent formation of several large regional spot markets under non-profit (regulated) ISOs and, later, Regional Transmission Organizations (RTOs). For our purposes, ISOs and RTOs are basically the same type of organization and we will use the term ISO generically. This chapter will largely focus on how ISOs were established in the USA and their experiences in wholesale market operations and administration. Ten years into the regulatory reforms, about two-thirds of US electricity demand is in regions under ISOs (Chapter 1 in this volume includes a map showing the geographical boundaries of US ISOs). As discussed in the Foreword and the Introduction to this volume, ISOs and the markets that they oversee are components of the ‘textbook’ model for restructuring and competition. As such, ISO market design and performance can make the difference between success and failure in the transition to competitive markets. While there remain problems to solve – particularly in some areas of market design, operating costs and organizational incentives – this chapter maintains that the US ISOs have made great strides and continue to advance on the basis of lessons learned and market participant needs. Most notably, heretofore they have been the surest way to establish open access to transmission, since they control access, pricing and assignment of tradable property rights on the transmission network, which in the USA remains under the ownership and maintenance of private- and government-owned utilities. They also operate automated, centralized auction markets for day-ahead (DA) and real-time (RT) power and certain ancillary services (AS). In these markets, reliability is maintained through dispatch instructions given to hundreds of generators simultaneously and spot prices are calculated every five or so minutes at tens of thousands of locations on the grid across multi-state regions, displaying a feat of computational capability, operational co-ordination and monitoring thought to be intractable just a few years ago. Wholesale buyers have the choice hourly of purchasing power from the ISO markets or scheduling their own supply (owned or contracted). Forward and spot markets, although not yet seamlessly integrated, are being developed on the basis of efficient market design. These are significant improvements in the utilization of US electricity assets. While some ISOs have had high start-up costs, the empirical benefit–cost analysis generally suggests that the benefits will outweigh the costs over the long term. The chapter is organized as follows. Section 14.2 presents a brief overview of US regulatory reforms in support of ISO development and a timeline of significant events (Table 14.1). This section explains how the particular laws and regulations of the USA and the division of federal and state responsibility have affected this process. Section 14.3 provides reference to general principles of efficient markets and how and why ISO markets do or do not achieve efficiency. Sections 14.4–14.6 and 14.8, examine details of the design of ISO markets for generation services, demand-response programs, resource adequacy and transmission property rights. Rather than provide case-studies of particular ISOs, the chapter provides a cross-cutting and historical perspective on design topics. Section 14.7 reviews FERC market

Independent System Operators in The USA: History, Lessons Learned, and Prospects

481

Table 14.1. Major milestones in the US regulatory reform of the wholesale electricity markets and formation of ISOs. Event

Date

Comments

PURPA

1978

Energy Policy Act

1992

California “Blue Book” proposal FERC Orders 888 and 889

1994

Requires utilities to purchase power at avoided costs from qualifying facilities – small renewable generation and co-generation. Created category of non-utility wholesale generators and required transmission access. First proposal for an organized competitive power market in the USA. Requires open access to transmission. Provides principles for ISO formation. March 31. PX operates DA and hourly markets with zonal pricing. ISO operates ancillary services and RT market and undertakes congestion management. April 1. First implementation of RT energy market with LMP and FTRs. However, for the first year, spot offers were capped at marginal cost. An ICAP market was already in operation. May 1. Bid-based auction markets for RT energy, regulation, three types of operating reserves and two types of capacity. Single zone energy pricing with uplift payments to “out of merit” generation. November 15. First implementation of DA and RT energy markets with LMP and FTRs. Also bid-based markets for regulation, three types of operating reserves, and zonal ICAP. Proposes voluntary formation of RTOs. California PX terminated. Market redesign, including DA and RT energy markets with LMP and FTRs, was first proposed in 2002 and is currently scheduled for implementation in 2007. July 31. Zonal market for energy along with regulation and operating reserves. Proposed the standardization of market designs in all FERC-jurisdictional utilities; this proposal was officially terminated in 2005. New England adopts the PJM market rules.

1996

California ISO and PX market start

1998

PJM ISO market start

1998

ISO New England market start

1999

New York ISO market start

1999

FERC Order 2000 California market crisis

1999 2000–2001

ERCOT (Texas) market start FERC SMD proposal

2001 2002

New England market rule revisions PJM expansion

2003

Midwest ISO market start

2005

SPP proposal given preliminary approval

2005

Energy Policy Act

2005

2004

PJM ISO expands west to incorporate portions of Ohio and Illinois. March 31. First implementation of an LMP energy market with multiple control areas under central dispatch. Currently scheduled for implementation in 2006. This region would operate a central dispatch energy balancing market, but would retain physical scheduling rights for transmission. Gives FERC enforcement role for reliability. Increases FERC penalty authority and transmission siting authority. Requires long-term transmission rights in ISO markets.

482


power monitoring and mitigation requirements and some specifics on the ISO rules. Section 14.9 briefly examines the important topic of ISOs and reliability, while Section 14.10 provides perspective on the evaluation of ISO performance, including costs and benefits of ISOs, advances in information technology, incentives for efficient management and innovation, and quantitative tools for design decisions. Section 14.11 is the conclusion. The chapter does not provide extensive review of prices and quantities in the US ISO markets. For such information, see the chapters on Pennsylvania–New Jersey–Maryland (PJM), Texas and California in this volume. Also, each ISO prepares and posts on its web site an annual state of the markets report that provides detailed market data; the ISOs also post many other periodic reports and data. Another source of cross-cutting market data and analysis are FERC state of the markets reports (e.g., FERC, 2005c).

14.2. Overview and Timeline of US Regulatory Reforms in Support of ISOs US regulatory reforms to support competition in the electricity sector have taken place both at the federal and the state levels. Under the statutory requirements and amendments to the Federal Power Act, the federal government regulates wholesale transmission rates and competition in the wholesale power markets (by utilities subject to federal jurisdiction).1 FERC, an independent government agency, administers the Act.2 The states regulate retail competition and other aspects of the industry important to the development of markets, such as siting of generation and transmission (although the Energy Policy Act of 2005 has enhanced FERC’s ability to influence transmission siting). This chapter will not attempt a broad survey of US federal and state regulatory reform, but will rather focus on the formation of the wholesale ISO markets (for some other views on US reforms, see the Introduction and chapters by Bowring, Sweeney, Adib and Zarnikau, and Tschamler, this volume). The Federal Power Act has standards that governed the cost-based regulation of franchised, monopoly utilities under the prior regulatory regime and which continue to govern market-based regulation. In particular, the rates set by jurisdictional utilities must be judged “just-and-reasonable” before they are charged or be subject to refund. These standards continue to be influential in the design and regulatory oversight of the competitive markets for power. What has changed is the interpretation of how to meet the standards under marketbased regulation. In general, FERC seeks to establish through its review and approval process that market rules will ensure that wholesale markets are efficient, competitive and “well-functioning.”3 Under the just-and-reasonable standard, prices must also sustain the market by providing the opportunity to earn at least normal profits for prudent

1

The Federal Power Act was set forth by the US Congress in 1935 in recognition that electricity markets were crossing state boundaries and hence that the federal government should exert jurisdiction over inter-state transmission and wholesale power rates. Most of the state of Texas has not been subject to federal jurisdiction. 2 FERC has limited authority over municipal, state, or federally owned generating and transmission facilities. Also, with the exception of hydropower projects, FERC has no jurisdiction over the construction or maintenance of power generating plants and until recently, over transmission lines. However, the Energy Policy Act of 2005 increased FERC’s ability to require siting of transmission lines, particularly in critical infrastructure corridors. 3 In Hope, the Supreme Court announced the principle that FERC is not bound to the use of any particular ratemaking method, as long as the outcome is just-and-reasonable. The courts have interpreted this rate-setting responsibility as ensuring efficient allocation of resources and protecting consumers; see


483

investments. In addition, FERC must also prevent “monopoly profits.” In terms of enforcement, economic regulation under the Federal Power Act is different from US antitrust standards and some other competition laws. Until the Energy Policy Act of 2005 (which increased penalties), the Federal Power Act provided little authority to act after the fact (e.g., no treble damages and no jail time). This has also shaped FERC’s market-based regulations in the new phase of competition.

14.2.1. Significant market and regulatory developments prior to Order 888 Regulatory reforms to promote competition in electric power began in earnest following the energy crises of the 1970s. The first federal attempt to promote competition in generation, PURPA, required utilities to buy power from “qualifying” independent power producers that could be small renewable generators or co-generation. The power contracts were to be at the utilities’ (projected) avoided cost, and were typically long-term. PURPA lead to an initial period of growth in independent generation capacity, but over the course of the 1980s and early 1990s, declining fossil fuel prices, reductions in renewable energy subsidies, qualifying tests and other factors, lead to the decline of PURPA-driven investments and this phase of market development subsided. In some states, particularly California, high priced PURPA contracts contributed significantly to the high electricity prices in the early 1990s that created pressures for regulatory reform to support competition. Motivated by the growth of independent power production under PURPA, and seeking to expand competition, the Energy Policy Act of 1992 created a new category of wholesale generators that was exempt from utility regulation and sought to promote the access of such generators to transmission. By this period, it was understood that the lack of transmission open access was a limiting factor to the growth of electricity competition. By the early 1990s, US railroads, airlines, telephony and natural gas pipelines had all been though the first phase of the open access debate and in some cases reforms (Winston, 1993). FERC had implemented open access to natural gas pipelines in the late 1980s. However, throughout most of the 1980s and early 1990s, the electric utilities remained opposed to opening access to the transmission grid, arguing that such access would threaten system reliability. The state of California was the first, in 1994, to contemplate a substantial reorganization and restructuring of its state utilities and transmission system operations around market competition, including generator competition and transmission open access. The federal initiative on mandatory electricity transmission open access would begin in 1996. For organized electricity markets in the USA, several other industry and market developments in the two to three decades prior to 1996 stand out. The first was the evolution of tight power pools in the Northeastern USA, which had developed central dispatch over large

3

(continued) Maryland Peoples Counsel v. FERC, 761 F.2d 780, 786, D.C. 1985. The courts have also approved the prices that result from well-functioning, competitive markets as just-and-reasonable. Commonly cited court opinions include the following: “(W)hen there is a competitive market the FERC may rely upon market-based prices ... to assure a ‘just and reasonable’ result,” Elizabethtown Gas Co. v. FERC, 10 F.3d 866, 870–71 (D.C. Cir. 1993); and “In a competitive market, where neither buyer nor seller has significant market power, it is rational to assume that the terms of their voluntary exchange are reasonable, and specifically to infer that the price is close to marginal cost, such that the seller makes only a normal return on its investment,” Texas Power Corp. v. FERC, 908 F.2d 998, 1004 (D.C. Cir. 1990).

484


regions.4 A second development was the establishment of the England & Wales market in 1989. Although the England & Wales transmission grid was consolidated under a single regulated transmission company, which did not take place in the USA on a large scale, the organization and rules applied in the first phase of that market became the touchstone for many of the US ISO market design debates that took place in the early to mid-1990s. For many observers, tight power pools with rules similar to the England & Wales market represented a straightforward platform for the introduction of electricity auction markets of sufficient scope and design specifications to support efficient and reliable markets (e.g. Hogan, 1994). Others argued that for purposes of market development, transmission system operators should remain separate as much as possible from forward market exchanges, which should be subject only to light-handed regulation. 14.2.2. Order 888 and ISOs Order 888 (FERC 1996) established the open access rules for the US transmission system under federal jurisdiction. All transmission-owning utilities were required to set rates for transmission access and possibly also transmission usage (e.g., congestion pricing) that were non-discriminatory and non-preferential and to make the available transmission capacity known on a timely basis.5 Order 888 also allowed for the full recovery of “stranded costs” associated with wholesale contracts.6 Order 888 suggested but did not require that the industry follow a particular type of market organization or adopt specific pricing rules. Instead, the industry and state regulators, and regional organizations (such as power pools) and other market stakeholders were allowed to develop their own approaches, consistent with the open access requirement. Hence, Order 888 was written to encompass both the transmission systems operated to support the organized, or “centralized” ISO-type markets for power then under consideration in California and PJM, and those that remained under the control and operation of individual utilities. To assist the former, Order 888 provided the principles for ISO formation.7

4

Power “pooling” generally referred to utilities that formed a group to co-ordinate and share resources so as to increase reliability and achieve economic savings. In the “tight” power pools that formed in the Northeastern USA, the member utilities typically enforced reserve sharing agreements through penalties and followed a form of centralized least-cost dispatch. 5 Available transmission capacity in the US context refers to the transmission capacity that a utility outside an ISO makes available to other parties after it has set aside sufficient capacity to serve its own retail customers and meet reliability requirements. In the ISO markets, the ISO makes the determination of the available transmission. 6 Also, in the 1990s, the incumbent regulated monopoly utilities in many key states were facing prospects of significant stranded costs in generation, as state regulators were beginning to conduct prudency reviews of their investments, particularly for nuclear plants. In most states that allowed retail competition, the acquiescence of the utilities was a result of allowing them to recover their stranded costs through regulated retail rates. 7 These principles included fair and non-discriminatory governance, that ISO employees had no financial stake in the assets that they would operationally control, that the ISO would offer a non-discriminatory, non-preferential tariff, that the ISO would be responsible for ensuring short-term reliability of grid operations, that the ISO control operations of the transmission system, that it identify constraints on the operation and take actions to relieve those constraints, that the ISO have proper incentives for efficient management and administration, that the ISO’s pricing of spot energy and AS encourage efficiency, that information be made available on a timely basis, that the ISO should co-ordinate with neighboring control areas, and that an alternative dispute resolution should be the first resort.


485

As its name suggests, an ISO is an entity, typically a non-profit regulated organization, that acts as a third party operator of the power system. The ISO owns no physical generation or transmission assets. The ISO’s primary purpose is to separate ownership and operational control of the transmission grid, thus greatly increasing the transparency of the open access regime.8 All suppliers interconnected with the grid in the ISO’s region would be granted equal access to the transmission system, whether they were supplying their own retail customers or offering spot power. The ISO co-ordinates the central dispatch on the basis of spot market offers provided by any eligible generator, subject to reliability constraints. The market design details are explored in the remainder of this chapter. 14.2.3. ISO formation and developments, 1996–1999 The first major steps in organizing ISOs with functioning markets were taken in California (1997) and in the three power pools of the Northeastern USA, PJM (1997), New England (1999) and New York (1999). Importantly, the design of each of these ISO markets was different (as were the state regulatory “bargains” that supported restructuring). The California design established a separate ISO and DA Power Exchange (PX) based on zonal pricing. The PX would clear and send its schedules for approval by the ISO, which in turn would be required to solicit “adjustment bids” from the PX to maintain inter-zonal transmission constraints. Then the ISO would procure AS and operate the system in RT, incorporating the actual inter- and intra-zonal transmission constraints. The inefficiency of this design was well known, but it had strong support from the trading sector of the restructured industry. Until 2000, the California market produced relatively low prices and the occasional instance of market manipulation was absorbed without serious concern (for further details on the California market, see chapter by Sweeney, this volume). In contrast, PJM was first to implement a market design with integrated market and system operations in 1998, allowing the RT energy auction to result in unit commitment schedules and locational marginal prices that satisfied system constraints. PJM also was first to introduce point to point financial transmission rights (FTRs) to hedge locational congestion charges (for details on the PJM market, see chapter by Bowring, this volume). New York followed with a similar market design in 1999, but with a DA locational marginal pricing (LMP) market and bid-based markets for operating reserves and regulation. New England also started a centralized market in 1999, but calculated a single pool-clearing price and out of merit generators on a (mitigated) pay-as-bid basis. 14.2.4. Order 2000 By the late 1990s, while trading of bulk power had expanded, the US markets remained quite balkanized.9 Outside the ISOs, there were still barriers to trade and other types of

8

A largely equivalent arrangement from the perspective of transmission users could be achieved via a regulated transmission company that owned and operated the grid but owned no generation. This was the arrangement established in England & Wales in 1990 through privatization. In the mid- and late 1990s, many parties in the USA lobbied for such transmission companies to form across large regions of the grid. But they never managed to aggregate more than a few utilities in a few states. 9 Quantitative measurements of how much trading has increased are difficult to make. For example, in Order 2000 (1999: p. 15), FERC noted that the reported volume reported by power marketers increased

486


economic inefficiency resulting from the manner in which vertically integrated utilities operated their systems. In particular, there was substantial concern about frequent curtailments of transactions, justified on the basis of reliability but often questioned by parties to the transactions.10 The ISOs themselves, while representing larger aggregations of the network, were still quite varied in terms of geographical scope – some single states, other multi-state. The continued presence of artificial ISO boundaries made some transactions vulnerable to curtailment or unanticipated changes in energy prices and congestion costs (a key such interface was the PJM–New York boundary). And there were the first signs that the market designs in some ISOs, such as California, were problematic. The next policy effort to address the problems of open access and market development took place in 1999, when FERC set forth Order 2000 (FERC, 1999). The centerpiece of Order 2000 was the concept of the RTO, essentially an ISO with sufficient authority and geographical scope to ensure that ISO boundaries were not interfering excessively with high trade areas (the RTO also would have more explicit authority to procure AS and conduct regional planning). The crucial limitation of Order 2000 was that FERC did not make RTO formation mandatory. Instead, it made it voluntary while providing incentives to those entities that joined an RTO. 14.2.5. The California crisis (2000–2001) and consequences In the spring of 2000 and continuing into 2001, the combination of a shortage of hydropower that California depended on during the summer months, market manipulation and market rules that had largely prohibited forward contracting by the major California investorowned utilities while leaving their retail rates frozen in most areas, created the wholesale price spikes that in turn bankrupted the utilities and forced the state to purchase forward power at very high prices (for analysis of the California crisis, see chapter by Sweeney, this volume; also Joskow, 2001; Joskow and Kahn, 2002; FERC, 2003a; Wolak, 2003a, b). The price shock spread to other Western states. The political response across the country was that these events were the result of regulatory and market failure that called restructuring into question. Particularly in the western and southeastern states, plans for ISOs and RTOs largely were set aside and the general momentum toward such markets waned elsewhere as well. The general intellectual assessment was that the California market rules were a major contributor to the susceptibility of the market to gaming and the exposure of the wholesale buyers to spot prices. However, in other operating ISOs, while capacity shortages would similarly raise prices, utilities could hedge themselves through forward purchases and transmission property rights. The regulatory response by FERC included supply offer caps and mandatory must-offer requirements in the California and Western markets and lengthy refund cases based on reconstructions of what a “just and reasonable” wholesale market prices would have been (e.g., FERC 2001a–c, 2003a).

9

(continued) from 1.8 million MWh in the first quarter of 1995 to over 400 million MWh in the first quarter of 1999, but much of the latter growth was through re-trading of power. In the ISO markets, there are measures of spot market volume, which have typically increased each year, but the ISOs do not collect data on how much of this volume is backed by forward contracts. 10 Such curtailments are supposed to follow the North American Electricity Reliability Council’s (NERC) Transmission Loading Relief (TLR) procedures, which provide non-economic criteria for the management of congested transmission facilities. For details, see www.nerc.com


487

14.2.6. The standard market design proposal In 2002, motivated in part by the failure of the California market design, the failure of additional ISOs and RTOs to form, and the continued seams problems between existing ISOs, FERC proposed a standard market design or SMD (FERC, 2002b). Developed through an examination of best practices in the eastern US ISOs, the core of the proposal was that all jurisdictional utilities should be individually or jointly under ISO-type market designs, including DA and RT markets for energy with LMP, FTRs and resource adequacy requirements. It would also have transferred jurisdiction from the states to FERC. This proposal was supported by many parties but strongly resisted by a number of utilities and some state regulatory officials in the northwestern and southeastern states. On reconsideration, FERC stressed that it would honor regional preferences on market design, but the political resistance to the proposal did not subside and it was formally withdrawn in 2005. 14.2.7. ISO formation and developments, 2000–2005 Despite the setbacks to organized markets, there have been several developments subsequent to the California crisis that represent the continued pursuit of regional market co-ordination and efficient market design. In 2003, ISO New England adopted the PJM market design rules and software. In 2004, PJM expanded westward to become the largest single power market operator in the USA. In 2005, following many years of development, the Midwest ISO moved from being a scheduling co-ordinator to operating a centralized dispatch over a region almost as large as PJM. In the same period, the Southwest Power Pool (SPP) advanced plans to form a centralized dispatch, although with DA scheduling based on pre-existing Order 888 transmission rights rather than FTRs. In California, the state began a period of re-examination following the electricity crisis. FERC abolished the California PX in 2001, leaving the California ISO in charge of the energy markets and initiating a market redesign process (FERC, 2002a). Currently, the California ISO has proposed market reforms for implementation in 2007 that will largely follow the designs of the east coast ISOs. As this overview has suggested, the ISO markets in the USA are still in evolution. There are some trends in evidence. Although the prescriptive aspects of the SMD have been withdrawn, about two-thirds of the US power sector has adopted or is moving toward a market design that combines the best features of that proposal and eliminates the failures of the previous market designs. But even in these markets, there remain many unresolved design issues, as will be discussed next.

14.3. Market Design: Principles and Performance 14.3.1. Defining and evaluating efficient spot market design In the USA, economic efficiency has been the major stated reason why government has chosen to restructure and reform the regulation of electricity markets. The political drivers, however, have included stranded cost recovery for owners of private assets and the desire of large users to shop for power, free of obligations to pay for the historic cost of uneconomic generation assets (Chapter 1 explores generally the factors that motivated restructuring across the world). Economic efficiency is defined in several dimensions. Productive efficiency is the provision of a good or service through the least cost mix of inputs (e.g., capital, fuel, labor, emissions allowances). Principally, allocative efficiency means that the good or

488


service is consumed by those who value it most highly. Dynamic efficiency means that as market conditions change over time, production and allocation are efficient (e.g., more efficient technology substitutes for less efficient). In general, efficient markets often require complex market designs or rules (e.g., Coase 1988). In the case of electricity, ISO market rules attempt in particular to provide for: 1. consistency between the reliable, physical operation of the forward and spot market pricing of generation services; 2. locational pricing and price-based congestion management; 3. the allocation and trading of transmission property rights; 4. market power mitigation; 5. development of a spot demand curve (in the absence of market-driven price-responsive demand) based on administered demand-response programs and scarcity pricing during shortages; 6. longer-term planning functions and markets to support reliability and investment. Standard neoclassical economic theory tells us that markets are efficient if they meet certain conditions (e.g., Mas-Collel et al., 1995). As applied to generation markets, if suppliers are induced to offer generation services at marginal opportunity cost, demand bids for electric power reflect the true value of power, transactions costs are not an impediment to efficient trade, and environmental and other externalities are insignificant (or internalized) then the equilibrium, or market-clearing, prices and quantities are both productively and allocatively efficient. In this market model, short-term (for our purposes, spot market) efficiency can lead to longer-term (dynamic) efficiency if entry and exit of generation suppliers are unrestricted and through the addition of forward markets to hedge risk. This standard model, for historical reasons of mathematical convenience, also assumed convexity of production functions and demand valuations; that is, that supply and demand could be smoothly increased or decreased over time (without “non-convexities” such as discrete production decisions) for purposes of market clearing. There are typically several possible design approaches to achieving the objectives of efficient markets. ISO market designs have been evolutionary in the USA, as in other countries undertaking restructuring (the other chapters in this volume offer much additional detail on market design alternatives). Some prominent questions on how to achieve economic efficiency and market completeness through design have included the following: ●

●

●

●

●

●

whether spot generation markets (DA and RT) and transmission system operations should be integrated through central auctions that recognized all relevant system constraints, or if they should be decoupled for purposes of decentralized forward trading of energy under more typical commodity trading rules; how many different generation services should be defined and priced through markets, and how to account for complementarities and substitutions among those services; how differentiated the prices need to be over space for energy, operating reserves and capacity; whether non-convex bids for supply spot offers, including (e.g., start-up and minimum load costs, are necessary); whether in the absence of price-responsive demand, the spot market design should provide administrative scarcity pricing for reliability based on the expected value of lost load or other measures; what limits on the exercise of market power are needed and in which circumstances;


●

● ● ●

489

whether there can be efficient investment in the absence of price-responsive demand and the presence of market power mitigation, and, if not, whether forward reserve markets or capacity markets will lead to such investment; how to price transmission usage, including congestion and line losses; whether to create “physical” (scheduling) or financial transmission property rights; whether to define transmission rights between nodes or zones or on “flowgates.”11

Table 14.2 shows a timeline of the introduction of bid-based markets for various generation services and transmission property rights in the US ISOs. 14.3.2. Consequences of design failure Generation markets (and almost all other product or service markets) depart from this efficient ideal for a number of reasons, the most well known of which are economic externalities due to the physical properties of power flows that cause congestion and losses, the failure of prices to reflect actual demand valuation by consumers of power due to rate regulation and lack of RT metering, the lack of competitiveness for some generation services in some locations, and the presence of economies of scale and scope in production. Left unregulated, a market beset by pervasive failures will result in inefficient quantities produced at inefficient prices, and over time yield the wrong mix of technology at the wrong locations. Government can attempt to correct for market failures though regulation, including oversight and refinement of market design.12 When a market design fails to achieve economic efficiency, purposely or not, this is often called a market design flaw. However, we note that the state of knowledge is often insufficient to determine ex ante which design best achieves short- and long-term efficiency; consequently, it is only after the fact that design flaws are revealed and hopefully corrected. This is why regulatory reform is, and will continue to be, an ongoing process. Looking back over the experience with market design, particularly in the USA but also in other countries, some designs that proved flawed over time were the result of stakeholder compromises to promote alternative design goals, while others were instead due to lack of knowledge or the state of technology. Some of these flaws will be discussed in the next section.

14.4. Design of efficient generation spot markets This section briefly reviews some of the debates over generation spot market design, primarily for energy, regulation and operating reserves. The discussion will address design characteristics and issues associated with typical ISO centrally optimized auction markets, as found in PJM (see chapter by Bowring, this volume), New York and New England. In the DA market, suppliers offer generation services, buyers submit bids for energy and the ISO procures AS on behalf of buyers. These markets are then cleared through a security

11

The term flowgate as used here simply means a transmission facility. In some usages, it may refer to particular transmission facilities that are being monitored for security. 12 We use the term regulation broadly here. Government may also fail to regulate or deregulate efficiently, but this is not our topic here. On market design, among the many surveys, see Chao and Huntington (1998), Hogan (1999), O’Neill et al. (2002), Stoft (2002), and Wilson (2002).

490

Table 14.2. Bid-based US RTO and ISO markets for generation and transmission services and some key design changes by year of introduction and termination.a 1998–1999

2000–2001

2002–2003

DA energy (zonal) HA energy (zonal) RT energy (zonal) Incremental/Decremental Bids (congestion) Regulation up Regulation down 10 minutes spin 10 minutes non-spin Replacement reserves

DA energy terminated HA energy terminated

Replacement reserves suspended AS based on zonal pricing

PJMc

RT energy (LMP) UCAP FTRs

DA energy (LMP) Regulation

10 minutes spin ARRs

New Englandd

RT energy AGC 10 minutes spin 10 minutes non-spin 30 minutes non-spin ICAP OPCAP (terminated 11/99)

ICAP auction terminated (ICAP requirement continued)

DA energy (LMP) RT energy (LMP) AGC supplanted by regulation ARRs/FTRs Operating reserve markets are terminated; future introduction is under consideration

Forward reserves

New York

DA energy (LMP) RT energy (LMP) Regulation (zonal) 10 minutes spin (zonal) 10 minutes non-spin (zonal) 30 minutes non-spin (zonal) ICAP (zonal) TCCs

ICAP demand curve

Regulation and operating reserve demand curves

California

b

2004–2005 (as of June)


Midwest ISO

DA energy (scheduled) RT energy (zonal) Regulation up Regulation down Responsive reserves Non-spin Replacement reserves DA energy (LMP) RT energy (LMP) FTRs

Abbreviations: AGC – Automatic Generation Control; ARRs – auction revenue rights (for transmission rights); AS – Ancillary Services; DA – Day Ahead; FTRs – Fixed or Financial Transmission Rights; HA – Hour Ahead; ICAP – Installed Capacity; LMP – Locational Marginal Pricing; Non-Spin – Non-Spinning Reserves; OPCAP – Operable Capacity (a product offered in New England for a few months that was available capacity on a daily basis); RT – Real Time; Spin – Spinning Reserves; TCCs – Transmission Congestion Contracts; UCAP – Unforced Capacity Credit Market. a For purposes of brevity, each column represents 2 years. Services appear in a column in the period that they were introduced and only reappear in subsequent columns if they were terminated or re-designed substantially. Note that the wholesale organized markets in California started bid-based operations on March 31, 1998, PJM on April 1, 1999, ISO New England on May 1, 1999, New York ISO on November 18, 1999, ERCOT on July 31, 2001 and Midwest ISO on April 1, 2005. If spatial pricing, such as LMP, is used for a product, it is shown in parenthesis; otherwise, there is a single system price. Zonal pricing implies that the clearing price applies only to a sub-zone of the system. Locational pricing of reserves and capacity is usually for an aggregation of nodes and hence is better described as zonal pricing. b The DA and HA energy markets in California were operated by the California PX, which was terminated in January 2001, after which only the ISO markets continued to operate. The remainder of the bid-based markets are operated by the California ISO; while energy is settled in RT, AS procurement is done DA, HA and in RT. c PJM began a zonal energy market with cost based offers on May 1, 1997, followed by an LMP market based on cost-based offers on April 1, 1998. The bid-based LMP market began a year later. d On March 1, 2003, ISO New England began operations under a new market design, most notably including using LMP to price energy, changing its pricing of regulation and suspending all operating reserve markets until a future date.


ERCOT (Texas)

491

492


constrained unit commitment auction (e.g., Hobbs et al., 2001a). The resulting clearing prices are used for financial settlement. The RT, or dispatch, market then prices and financially settles deviations from the DA schedule, based on additional supply offers and demand bids submitted after the close of the DA market. 14.4.1. Representation of energy offers and bids A generation unit’s actual daily variable cost structure, depending on the technology, is composed of a number of possible components, including most notably fuel costs, which can differ for start-up, no-load and incremental energy output; opportunity costs relative to sales at other locations, times, or other generation service markets; fuel storage costs (e.g., if a gas unit changes its daily storage need on short notice); and other operating and maintenance (O&M) and labor costs. For an efficient dispatch, certain physical characteristics of the unit must also be considered, such as maximum and minimum operating levels, ramp rates, and minimum on- and off-times. In the early phase of generation market design, a key debate was over how many spot energy offer components should be required, or even allowed for voluntary representation, especially in the forward (e.g., DA or hour-ahead (HA)) markets operated by an ISO. In the early California design, the prevailing view was that DA offers in the California PX should consist only of a single part: a price for the quantity of energy offered along with separate offers that contained prices for adjustments.13 Other than total quantity, no physical generation constraints, such as ramp rates, were allowed to be specified. The motivation for the single part offer rule was to establish a single (zonal) market-clearing price without additional payments to generators for unit commitment costs, which were to be factored into the single offer price and the adjustment offers. Through these DA offers and the subsequent adjustments made after the PX market settled by the ISO, the market was supposed to result in an efficient dispatch. However, in practice, this offer rule did not result in short-term efficiency, in part because the ISO was not allowed to optimize based on unit characteristics. Another result was that when the federal regulator later determined that the California market was not competitive during the price spikes of 2000–2001, it was not possible to determine ex post whether a generator’s output over a day was due to its inability to express its actual marginal cost (including start-up) or an attempt to exert market power through physical withholding (as defined in Section 14.6). In contrast, the eastern US ISO and RTO spot markets evolved from tight power pools in which generators were more familiar with centralized dispatch. Consequently, in these markets the “multi-part” offer eventually become standard, with a separate price component for start-up ($), no-load ($/h) and incremental energy ($/MWh).14 There are slight variations in the offer components among the ISOs; Table 14.3 shows the required and optional components for the New York ISO (the other ISOs have similar offer specifications). In addition to allowing firms to better approximate their true marginal costs, the multi-part offer provides

13

Generators were also allowed to submit offers to increment (“inc”) or decrement (“dec”) output from the cleared quantity. Such offers could then be used by the California ISO to establish a physically feasible dispatch. 14 Wholesale buyers could also have a multi-part representation of their bids. For example, an industrial customer may have an analog to the generator start-up cost if, for example, it is bidding in all-or-nothing fashion to shut down a particular production line on a hot day when prices are high.


493

Table 14.3. Supply offer components in the New York ISO short-term markets for generation services.a Parameters

Variability

A. Offer prices and quantities Startup price Minimum generation energy block and price Dispatchable energy Regulation capacity availability Regulation capacity price Spinning reserve price 10-minute non-synchronized reserve 30-minute operating reserve

$/hour MW, $/hour # steps, $/MWh, MW/Step MW $/MW $/MW $/MW $/MW

Hourly Hourly Hourly Hourly Hourly Hourly, DA only Hourly, DA only Hourly, DA only

B. Physical generation characteristics Dispatch status Startup time

Whether ISO or self-committed Hours, minutes

May vary May vary per commitment period, DA or RT

Minimum run time Minimum down time Maximum startups per day Normal upper operating limit Emergency upper operating limit Normal response rate Regulation response rate Emergency response rate Reactive power capability Physical minimum generation limit

Hours, minutes Hours, minutes 1–9 MW MW MW/minute MW/minute MW/minute MW plotted against MVARs MW

Static May change over day May change over day May vary

Static Static

Abbreviations: DA – Day-ahead; RT – Real-time. a Additional details are found in the NYISO tariff. Static refers to offer components that remain relatively constant over the life of the offer, but can be changed. Sources: NYISO Market Services Tariff, Att. D; version February 1, 2005; NYISO technical manuals.

a basis for introducing efficient markets for operating reserves, as discussed below. On the other hand, offers with multiple price components and physical parameters create new gaming strategies, and this has indeed been experienced in some US markets, which typically impose limits on changes in each component.15 Also, most ISO markets allow “virtual” supply offers and demand bids in the DA market. Virtual offers are single part offers (i.e., price and quantity only) that are not necessarily backed by a physical asset or real load. Any virtual supply settled financially at the DA price must be “bought-back” at the RT price (similarly, any virtual demand settled at the DA price must be sold-back at the RT price). Entities submit virtual offers and bids to arbitrage differences between the DA and RT market prices for example, as suppliers to sell high (day ahead) and buy-back low (real time). If enough virtual supply enters, this may 15

For example, in PJM, suppliers submitted 24 h minimum run times to take advantage of opportunities to be run out of economic merit order. The ISO market monitor eventually limited such physical inflexibility. In theory, a competitive market should be incentive compatible; that is, they should motivate generators to truthfully state their costs and physical constraints in their offers. However, in practice, the complexities of power markets may require some monitoring and mitigation of all offer components.

494


lead to convergence of the DA and RT prices. In addition, entrance of virtual supply and the price convergence increases the volume of the DA market. However, for reliability purposes, as discussed below, ISOs remove accepted virtual supply offers from the DA schedule prior to determining whether additional physical resources are needed prior to the dispatch market (the so-called reliability unit commitment). ISOs may also apply additional rules to virtual offers to limit market manipulation, such as restricting such offers to certain locations (e.g., PJM) or suspending the rights to participate of entities whose offers consistently cause prices to diverge (e.g., New England), typically as a means to create and garner congestion rents. 14.4.2. Spot energy pricing Two primary types of energy pricing have been considered for the spot auction markets: “uniform market clearing prices,” in which all suppliers at the same locations are paid the same price, typically the marginal accepted offer; and “pay-as-bid” pricing, in which suppliers are paid what they offer. The efficiency implications of auction markets with these two types of pricing have been examined extensively elsewhere (e.g., Ausubel and Cramton, 1996; Kahn et al., 2001; Fabra et al., 2004, Son et al., 2004; Ren and Galiana, 2004a, b); here we simply note the often misunderstood fact that in a competitive “pay-as-bid” market, suppliers’ offers will converge on their estimate of the market clearing price (i.e., suppliers will offer not at their production costs, but through an estimate of their opportunity costs). This requires them consistently to raise their offers above marginal cost. If there are regulatory rules that require monitoring and mitigation of market power as measured in the deviation of price offers from a benchmark competitive price (as there currently are in the US markets), then the offer incentives under pay-as-bid will make the task harder than those under uniform pricing.16 In the ISO markets, the multi-part offer creates a pricing regime that combines these two pricing rules, with uniform prices that clear the market for energy, and “pay-as-bid” prices for the non-convex offer components, start-up and no-load (e.g., Motto and Galiana, 2002; Hogan and Ring, 2003; Elmaghraby et al., 2004; O’Neill et al., 2005a). Start-up and no-load payments through the pool can be part of a set of energy prices and additional payments to generators that supports a competitive equilibrium, in that given those prices and payments, no generator can increase its profit by deviating from the accepted supply schedule (as shown in O’Neill et al., 2005a). We will discuss efficiency properties of this pricing rule further below. 14.4.3. Complementary and substitution properties of energy and short-term AS Short-term generation services, such as energy, regulation and reserves, are complementary or substitute uses of the same machine. Energy and regulation are complementary services, since a unit must provide some energy to provide regulation, although the converse is not true. However, regulation and energy are also partial substitutes in that if a unit is providing regulation, it must deviate from its optimal energy output in response to regulation signals to ramp up or down. Meanwhile, regulation and operating reserves have the property of 16

This is because uniform pricing should induce suppliers to offer at their production cost, in the knowledge that they will receive the price set by the marginal offer (but see discussion in Section 14.6.6 of offer “creep” under some market power mitigation rules).


495

“hierarchical substitution.” A quicker response reserve (sometimes called a “higher quality” reserve) can provide the same or better reliability benefit as any slower response reserve. When these multiple services are offered into the same regional market for electric power, the complementarities and substitutions must be fully recognized to achieve productive efficiency. For example, if an ISO is prevented from substituting higher quality reserves for those of lower quality, price “inversions” can occur in which the lower quality reserve receives a higher price, unnecessarily increasing costs.

14.4.4. Offer representation and pricing of regulation and operating reserves In the absence of demand-side response, regulation and operating reserves are procured by the ISO on behalf of buyers to fulfil reliability requirements (efforts to establish demand curves for reserves will be discussed below). These short-term AS are either explicitly priced (i.e., with market-clearing prices) or alternatively generation units are paid opportunity costs or stand-by costs based on their energy offers (possibly along with additional market payments to stimulate investment). The choice between these designs, and others, remains the subject of debate (e.g., Cramton et al., 2005). All offers into regulation and operating reserve markets, both those cleared sequentially or simultaneously with energy (as discussed in the next section), need to be accompanied by an energy offer that will allow the unit to be dispatched for energy as needed for reliability. The multi-part bid for energy is the basis for an efficient market design for simultaneous clearing of energy, regulation and spinning or quick start non-spinning reserves. This is because a generator whose energy output is “backed down” for the purpose of providing regulation or reserve should be paid the opportunity cost of not providing the energy (price minus its marginal energy cost, if the marginal cost of regulation or reserve is close to zero). This generally makes the generator indifferent between providing energy or these alternate services. In addition, a generator that is dispatched to a minimum operating level to provide reserves has already represented its start-up and no-load costs through the multi-part energy bid and will be paid those costs if standing by to provide reserves. While the three-part energy bid provides the right incentive for a generator to back-down or stand-by to provide regulation or reserves, the argument has been made that generators have other costs associated with providing these services. Hence, regulation and reserve market designs also may allow additional “availability” offer components for efficient market-clearing. Table 14.3 illustrates the offer components in such a market, the New York ISO. These might be used to represent opportunity costs associated with selling in a different market or availability costs that are not fully captured in the multi-part bid. For example, regulation offers typically allow additional bid components to cover the fixed costs of the equipment for automatic generation control (ACG) as well as any additional wear and tear associated with providing regulation. Some designs allow for different offers for different regulating ranges around the generator set point. Yet others separate regulation into two services, “regulation up” and “regulation down”, and allow separate offers for each. Nonspinning reserve offers can be allowed an availability component to account for the costs of standing by to provide reserve energy. Critics of such “availability” pricing for reserves argue that in the spot markets, the marginal cost of providing reserves is negligible, and hence the availability bid will generally only be positive due to market power (e.g., Cramton et al., 2005). As such, competitive short-term reserves markets with mostly close to zero prices will fail to attract investment in reserve technology. They propose that instead of explicit spot reserve pricing, market

496


designs are implemented that provide positive price signals for reserves, such as the scarcity pricing and forward reserve or capacity markets discussed below.

14.4.5. Sequential versus simultaneous market clearing When spot reserves are explicitly priced through reserve offers, there are two basic ways to clear the short-term markets for energy, regulation and reserves, reflecting the complementarities and substitutions among them: sequential and simultaneous. The sequential method first operates a market for energy followed by markets for regulation and reserves, cleared in a sequence. Variations of this approach were adopted initially in California and New England. In both cases, the different markets were initially cleared without considering substitutions with other services. This increased the potential for market power, by fragmenting what was actually a larger market when substitutions were considered, and also resulted in price inversions. There are two major alternatives for introducing hierarchical substitution into sequential reserve auctions (e.g., Kamat and Oren, 2002). In the first, the ISO (which procures reserves for load) minimizes overall production costs, as represented by the supply bids. In this design, the ISO buys all high-quality reserves that are available to substitute for higher cost, lower quality reserves. In the second design, the ISO minimizes procurement expenditures, accounting for the prices of different reserves (the latter is also called the “rational buyer” model). That is, the ISO will make substitutions only to the extent that it lowers total procurement costs. Sequential markets with rules for hierarchical substitution can gain some or all efficiencies resulting from substitution policies. Integrated spot markets achieve productive efficiency through simultaneous optimization (also known as joint or co-optimization) of generation services. In this design, offers for all services are submitted at the same time, and the auction minimizes as-bid production costs or total procurement cost associated with providing energy, regulation and reserves, with the hierarchical substitutions reflected in the optimization constraints. Minimization of procurement cost is a more difficult problem than minimizing total as-bid costs, because the total payment objective is a function of both dual variables (prices) and primal variables (Luh et al., 2004).

14.4.6. Spatial aspects of energy and reserves pricing Under either of the energy pricing rules discussed above (i.e., uniform or pay-as-bid), there are two general ways to account for the spatial price differentiation caused by transmission network effects, such as congestion and losses: zonal pricing and nodal pricing. In theory, a zone is a set of nodes in geographical/electrical proximity whose prices are similar and are positively correlated over time; that is, they are not affected much by congestion or losses between them. Spot energy is settled at the zonal price, with side payments made for generators that are “constrained up” and “constrained down” due to congestion within the zone. Constrained up generation is that energy whose marginal cost exceeds the zonal price, but is required because of transmission constraints. They are normally paid an uplift equal to the difference between their bid and the zonal price. In contrast, constrained down generation is output whose marginal cost is less than the zonal price, but cannot be taken because of, again, transmission constraints. Constrained down generation is sometimes paid the difference between the zonal price and its bid for output that could not be accepted due to transmission congestion.


497

With the goal of simplicity in transmission usage pricing, versions of multi- or singlezone pricing was the early choice in the PJM, England & Wales, California, ERCOT and New England markets. This was because zonal pricing was viewed as supportive of decentralized forward markets and because intra-zonal congestion, as experienced prior to the market start, was considered minimal. In some regions, zonal pricing also maintained the existing congestion, or “re-dispatch,” cost subsidies from low price locations to high-price locations. However, when the transmission congestion within zones does not prove to be minimal, as was typically the case once centralized markets started, then the allocation of zonal re-dispatch costs can quickly become large and inefficient (e.g., Stoft, 1997; Hogan, 1998, chapter by Adib and Zarnikau on the Texas market, this volume). For example, there is an incentive to understate one’s cost and exaggerate one’s potential output to magnify constrained down payments – the so-called “dec” game in California. Moreover, due to its averaging, zonal pricing masks the price signals for within zone location of new generation (or, as discussed below, transmission or demand response) (e.g., Baldick, 2003). Such difficulties have led all US ISO markets to either abandon, or propose abandoning, zonal pricing schemes in favor of LMP. However, zonal pricing remains popular in some other countries (e.g., chapter by von der Fehr, Amundsen and Bergman on the Nordic market, this volume), sometimes because within-zone congestion is relatively unimportant in those markets, or is managed by a regulated transmission company through its transmission rates, or all the generation in the zone is owned by one entity. In contrast, nodal pricing, or LMP, provides the value or cost of the marginal energy produced or consumed at the nodes and eliminates subsidies of re-dispatch costs.17 LMP is defined as the marginal cost of delivering the next increment of power at a network bus. For sellers of spot energy, the locational price is the price that they will be paid for each MWh; for buyers, it is the price that they will pay for each MWh. The difference between the total LMP-based payments by buyers and the total LMP-based payments to sellers is the total congestion rent collected by the transmission system operator (the total amount available to pay holders of FTRs). Critics of nodal pricing often argue that large numbers of locational prices are confusing to market participants and make forward contracting more difficult. The larger US ISO markets can have tens of thousands of LMPs when there is transmission congestion. There are several ways to address this issue. In the ISO markets, any bilateral contract can be “self”scheduled to obtain transmission service regardless of LMPs (any resulting congestion charges can be hedged through transmission rights; see Section 14.7).18 In addition, the ISOs also calculate “hub” prices, reflective of an average of LMPs in a major delivery zone (e.g., PJM West). These hub prices simplify forward contract settlement. Importantly, the LMP allows for making economic decisions once the spot price is known independent of a generator’s forward contract price for energy. In general, a generator should produce energy when its marginal cost is lower than or equal to the LMP (and assuming that the market rules allow for capturing start-up and no-load costs, as noted above). A generator should fulfil the contract through spot market purchases if the LMP at its location is lower than its

17

With the exception of load LMP averaging across multiple utilities, as takes place in some regions. Load in many LMP markets has opted for a zonal price based on a load weighted average of the LMPs. When more than one retail supplier exists within a zone, this means that some cross-subsidy takes place. 18 A self-schedule or bilateral schedule means that the supplier requests that the ISO allow it to inject and withdraw power regardless of the market price. These schedules can remain outside the ISO’s dispatch, except for purposes of reliability.

498


cost (again, as long as either the buyer or seller also holds transmission rights to hedge congestion charges). The simplest method to produce only when it is economic to do so is to submit the generator’s supply offer into the ISO energy markets regardless of whether there is a forward contract for its power. This is indeed what happens in the ISO markets. The logic of locational pricing is now being extended to other generator services. In the USA, operating reserves have generally been priced on a system-wide basis, although some ISOs have defined sub-zones for reserve procurement, with all market participants being responsible for a load-ratio share of the reserve costs. As in the case of energy, an accurate locational marginal price for reserves can provide for a more efficient dispatch, assigning costs of providing reserves more directly and providing a better signal for investment. However, a locational reserves market is not a straightforward extension of the energy market. A fully functional market design for locational reserves must achieve the following objectives simultaneously: energy reserves must be priced separately from energy production at each node; sufficient generation capacity and transmission capacity must be held in reserve and priced so that no matter what contingency occurs, a feasible dispatch is possible; and the costs associated with holding capacity in reserve should be allocated by marginal costing principles to the participants whose demands imply the need for those reserves (O’Neill and Stewart, 2003; Fu et al., 2005). This last goal is best accomplished by establishing locational marginal prices for both production and reserves at each node in the system and establishing prices for transmission elements and transmission reserves that reflect the cost of holding some production and transmission in reserve. 14.4.7. Revenue sufficiency guarantee All the US energy and reserves markets with multi-part supply offers (start-up, no-load, energy) and simultaneously optimized energy and reserve markets have converged on another crucial market rule: the guarantee of offer revenue sufficiency. This rule states that for all generation service offers by a particular unit that are accepted in the spot auction, the total daily revenue from the market must at least equal the offer requirements. If not, the supplier is eligible for a ‘’make-whole’’ payment that is recovered as an uplift charge to all load.19 In supporting multi-part offers, the payment guarantee further reduces the uncertainty associated with a bundled offer (e.g., a one-part offer) and allows for bidding consistent with an efficient dispatch. 14.4.8. Scarcity pricing and demand curves for reserves To promote allocative efficiency and investment in capacity, prices should rise above marginal cost if generation capacity is binding, that is, the requirements for reserves and energy exceed what generation can provide (Schweppe et al., 1988). In electricity markets, then, reserve shortages are the real-time indicators of such market “scarcity” and hence when

19

For example, consider a generator with a $3000 start-up cost and a $40/MWh energy offer for its range of output, which is run for 4 h during which it gets paid a $50/MWh price for output of 20 MWh/h. It thus gets paid $50/MWh ⫻ 20 MWh/h ⫻ 4 h ⫽ $4000. According to its offer, it needs at least $3000 ⫹ ($40/MWh ⫻ 20 MWh/h ⫻ 4 h) ⫽ $6200 to cover operating expenses for that output. Hence, the generator is eligible for a revenue sufficiency payment of $2200.


499

competitive market prices should be increasing. There are two major barriers to connecting short-term reliability and market scarcity. First, very little of the system load is currently price responsive in the short run (as discussed in more detail in Section 14.5). Second, because reserves are modeled as “hard” constraints, the system is declared to be reliable if the required amount of reserves is present. Anything less is declared unreliable. Any additional reserves are not paid for. This approach was satisfactory from a purely reliability perspective under monopoly regulation, but neither of these assumptions provides appropriate market incentives. Reserves should be allowed to be shorted at a price, so long as security is not violated; also, some degree of excess reserves should be priced because reliability increases as more reserves become available. The method for substituting for the lack of buyer price responsiveness is so-called “scarcity pricing”. There are several ways to implement scarcity pricing. One is the now-abandoned England & Wales method of adding to the price a factor equal to the product of the loss of load probability (LOLP) on a given day and an assumed value of unserved energy. Another, which addresses directly the reserves pricing issue, is to employ a demand curve for reserves, ideally based on the expected value of lost load as reserves diminish, that will increase the price the ISO pays if reserves are short and also pay a low but positive price for reserves above the hard requirement (e.g., Stoft, 2002). A high price for operating reserves will then translate into a high cost of energy, as generators factor in the opportunity cost of the high reserves price in their energy bids or through the co-optimization algorithm. Such concepts have been discussed in ISO New England. This artificial, but market based, demand curve has two purposes. It sends high price signals during periods of scarcity to expand supply and restrict demand. Further, it eliminates the artificial all-or-nothing construct of reserves, recognizing that more reserves are worth something. This curve should be calibrated to promote demand-side bids that can eventually substitute for it (e.g., if the artificial shortage price hits $1200/MWh, buyers that value power at less than that price will be prompted to bid and voluntarily reduce consumption). In theory, a competitive market in which shortage prices reflect the value to load of reliability can result in the optimal incentives for investing in new capacity (e.g., Caramanis, 1982). However, the presence of price caps or other market power mitigation measures may mean that new generators cannot earn enough gross margin (revenue minus variable costs) to cover fixed costs. This endangers the reliability of the system, and shows that the design of spot markets should not be done separately from the design of market power mitigation and capacity markets, as will be discussed in the following sections.

14.4.9. Pricing of reactive power Another topic of current interest in the ISO markets is pricing of reactive power (FERC, 2005a). Many systems are dispatched without using a full alternating current (AC) optimal power flow, while imposing overly restricted voltage levels. As a result, the market is incomplete: the essential commodity of reactive power is either inappropriately priced or not priced at all. A basic question is whether it would be more cost-effective for an ISO to sign long-term contracts for reactive power (similar to “reliability must run” contracts for expensive generation in load pockets), or to operate forward and spot markets for reactive power (Hogan, 1993; Kahn and Baldick, 1994; FERC, 2005a). If full markets are to be created for reactive power, work is needed to improve optimal power flow software and its integration with unit commitment models.

500


14.4.10. Sequence of forward and spot generation markets In the most general sense, the ISO markets for generation services should be a sequence with the following properties. For each forward market, the financial position taken by each buyer and seller can be changed in the next market by buying back or selling back some or all of the prior position (e.g., O’Neill et al., 2002). A financial position can be turned into a physical position by acquiring new capacity. In the transition to the final, physical, or dispatch, market, the last adjustment is made and the prices associated with delivery of the actual product are cleared. In electricity market design, this is often called a “multi-settlement system”. In the actual US markets, there are several auxiliary procedures and rules that have arisen to substitute for the lack of demand responsiveness and for situations where prices in the sequence of markets are insufficient to ensure market clearing (e.g., when sufficient capacity is available but offered supply before the dispatch hour or actual output is not sufficient to meet actual demand). First, most markets have added a type of reserve purchase that reflects the difference between the ISO’s next-day forecast and the scheduled and bid in next-day load. This additional reserve is typically cleared by paying “as bid” for start-up and no-load for any generators postured to provide the reserve over and above the DA market-clearing level. Second, several markets have instituted financial penalties for deviations from the market and ISO’s dispatch instructions; that is, for any generator that does not perform as instructed based on its accepted supply offer.

14.5. ISO demand-response programs In the early years, ISO market design focused on generation and transmission issues on the assumption that competitive market prices would lead to wholesale buyers becoming more price responsive. Most ISO designs for energy markets can accommodate detailed (multipart) bids by buyers and with the proper certification, price-responsive demand may also provide reserves and capacity (e.g., FERC, 2002b). The potential economic benefits have been well documented (e.g., Boisvert et al., 2002). However, demand-side participation has been slow to develop at the wholesale level, reflecting persistent barriers to retail customer response. With the exception of some large industrial customers, for many retail customers in the USA to engage in demand-response requires dynamic retail pricing or incentive or subsidy programs by their retail supplier (and hence has more to do with state regulation than with FERC rules). This is because most retail customers currently see electricity rates that are averaged over a month or more. Their consumption decisions would be unaffected by price spikes in the wholesale market, unless they are responding to public service announcements or they are among the minority that participate in utility interruptible rate or load control programs. Consequently, during peak periods consumers pay far less for power than it costs to generate. In addition, at the wholesale level, all ISOs impose supply offer caps that create effective limits on spot wholesale prices, typically at around $1000/MWh. These price limits dampen the incentive to invest in demand-response capability. With FERC and state regulatory encouragement, ISO market designs have addressed this lack of demand response in two general ways (Barbose et al., 2004). First, as discussed above, ISOs have sought to institute design solutions for wholesale pricing, such as the scarcity pricing discussed above. Second, ISOs have established demand-response programs that attempt to stimulate price responsiveness by providing incentives and/or payments for participation (e.g., participants may get the higher of the market-price or a pre-determined price). These programs are categorized as emergency programs and market-price-driven programs, with the former requiring the ISO to request load response directly. The northeastern ISO demand-response programs are summarized in Table 14.4, and programs in

Table 14.4. Demand-response programs in the northeast US ISOs, 2004. Program type

ISO-NE Real-Time Price Response Program (RTPR) Real-Time Demand-Response RTDR Real-Time Profiled Response Program (Profile)

NYISO DA Demand-Response Program (DADRP) ICAP SCR (EDRP)

Program enrolment

Enrolment as percentage of Regional 2004 peak load variations

9,216 MWh 0 0

108 MW 165 MW 83 MW

0.4% 0.7% 0.3%

3,535 MWh Contractual 0 Emergency 0

377 MW 981 MW 581 MW

1.3% 3.9% 2.0%

48,622 MWh 724 MW 179 MWh 46,561 MWh

0.7%

0 0

1.3% 1.7%

Price taker Emergency and contractual

Bid based

PJM Economic Load Response Programs (ELRP) ● Day Ahead Option (ELRP-DA) Bid based ● RT option (ELRPRT) Price taker ● Non-hourly, metered program (Pilot) Varies Emergency Load Response Program (Emergency) ALM

1.881 MWh Emergency Emergency

1385 MW 1806 MW

Enrolment varies by zone Combustion turbine (CT) zone accounts for 89% of Program (RTDR) and 29% of RTPR NEMA makes up majority 37%) of RTDR and ME is majority (91%) of profile program Enrolment varies by zone NYC and Long Island (Zones J&K) account for 53% of Emergency DemandResponse Program (EDRP) load, 28% ICAP-special case resources (SCR) and 4% DADRP Western superzone (Zones A–E) is 29% of EDRP. 64% of ICAP-SCR and 73% of DADRP

Economic program enrolment varies significantly by PJM control zones. Eighty-two percent of MWh reductions occurred in single zone (AP). Two zones had no DR activity (PEPCO and RECO)

501

Source: FERC 2005c, p. 213.

Achieved reduction

Reliability based


Price based

2004 program usage

502


California are described by Sweeney (chapter this volume). ISO demand-response programs and studies of non-ISO regions in 2004 found that on average in the USA, about 3% of peak load was available for demand-response (summarized in FERC, 2005c: 212–213). The ISO programs have proven significant in preventing load curtailments during some peak demand periods, most notably in New York in 2002 and in parts of New England in 2005. However, demand participation continues to face barriers and has not been fully exploited in any market (e.g., Borenstein, 2004). Furthermore, there are cases in which traditional active load management (ALM) programs (such as residential water heater or air conditioner controllers) are being abandoned by some utilities on the basis that energy prices in the early 2000s have not justified the cost of maintaining them (PEPCO, 2004). Thus, even though increased demand response continues to be a FERC policy objective and is a goal of the Energy Policy Act of 2005, progress on this front remains halting.

14.6. Design of markets for generation capacity In normal commodity markets, funding for the capacity and storage required to meet peak demands is provided by higher than normal prices during those times. But in several US power markets, there are separate capacity markets for electricity or other resource adequacy mechanisms to ensure that “enough” generation capacity is built.20 Several reasons are offered for the prevalence of such capacity or resource adequacy mechanisms, all pointing to unique characteristics of power markets or to one failure or another of such markets to conform to the assumptions of the perfect competition ideal (e.g., Jaffe and Felder, 1996; Hirst and Hadley, 1999). One factor often presented is the combination of capital intensiveness and absence of significant storage (e.g., reservoir hydropower); as a result, meeting peak demands that only occur a few hours per year is very expensive. Spot prices in the thousands of dollars per megawatt hour may be needed for recovery of capital costs, depending on the type of unit and hours run (such prices could be market-driven or via scarcity pricing such as described above). In contrast, other industries that are less capital intensive have more ability to store and transfer commodities from one period to another. They can also charge high prices to meet demand and recover costs during peak periods. As a result, the swing of marginal cost from off-peak to peak periods is not nearly as extreme as in the power industry. But high peak marginal costs do not by themselves explain the need for capacity markets. The other consideration is inefficiencies in the demand side of the market. One failure of the demand side is the presence of regulatory price caps, typically justified on the basis of market power mitigation, or other sources of price rigidity that prevent prices from climbing anywhere near the level of scarcity prices during peak periods. A further demand-side failure noted above is that price fluctuations in the bulk power market are not communicated to most retail customers, who pay a rate that is either constant

20

Capacity requirements are typically distinguished from operating reserve requirements in several ways. First, the former is typically a multiple of a LSEs’ peak load, whereas the latter is a fraction of peak load or a load-ratio share of the largest contingency (multiple contingencies). Second, the set of generators that can fulfil capacity requirements is different from those that can fulfil operating reserve requirements of various time frames. Hence, some generators may be eligible for capacity payments that are not competitive in operating reserves. For these and other reasons, the addition of a capacity market will produce a different revenue stream for generators than the operating reserve markets.


503

or just seasonally adjusted (possibly under regulated retail rate caps). In contrast, when consumers are subject to prices that fluctuate in RT, they do respond by decreasing loads in peak periods, or shifting uses to off-peak periods (e.g., Boisvert et al., 2002). This lessens the need for expensive capacity. In theory, RT prices faced by all market participants will result in the optimal amount of system capacity and reliability, as the market prices will express the consumers’ willingness to pay for power during peak times, just as in other commodity markets. This was, of course, the original vision of Schweppe et al. (1988) for a power market in which decisions are co-ordinated by price. But price cap regulation and lack of hourly meters for most customers mean that this goal is unattainable, at least in the near future. As a result of these demand-side failures, generation capacity becomes a public good. That is, the benefits of adding capacity are received by all consumers in the market to whom its power can be delivered, and are not captured by the owner of the capacity in the form of higher revenues. Economic theory says that public goods tend to be undersupplied in markets, so therefore too little capacity would likely be built in the face of the demand-side failures. Besides the demand-side market shortcomings, another market problem provides a rationale for capacity markets: that of market power. In the extreme, unresponsive demand means that suppliers can raise prices at will. Even if no individual supplier is “pivotal,” prices can be above marginal cost.21 However, this will be less of a problem if there is more generation capacity, particularly in spot markets if loads forward contract the bulk of their needs. If the output of most capacity is already committed to be sold at a fixed price, there is little advantage to manipulating spot prices (Green, 2000). Several types of fixes are proposed to correct the market failures and ensure adequate generation capacity. As of this writing, FERC has generally been supportive of imposing resource adequacy requirements (for example, pressing California to establish them) and is evaluating alternative design proposals (as summarized in Table 14.5). 14.6.1. Energy-only market The first approach is to forgo capacity requirements and to rely on scarcity pricing of energy or operating reserves to provide enough gross margin to generators. As a result, there are no or very high caps on energy offers or prices. This was the course taken by California, although its offer caps were progressively tightened during the crisis period.22 The major concern with energy-only markets is distinguishing genuine scarcity pricing (when prices rise in order to ration demand to available capacity (ACAP)), which is economically efficient, from exercise of market power. In the absence of demand elasticity or extensive market power mitigation such as price caps, bid caps and mandated forward contracting, there are significant incentives for suppliers to remove capacity from the market so as to push energy prices up during times of high load. However, some observers argue for a transition to energy-only markets because of the efficiency benefits of having prices actually reflect opportunity costs when supply is tight, and suggest market power mitigation or required contracting as interim measures that lend themselves more to that transition than permanent ISO-administered capacity markets (e.g., Hogan, 2005; Oren, 2005). 21

A supplier is “pivotal” in a period if the removal of its operable capacity from the market would put that market into deficit. Demand is assumed to be inelastic. It is thus inferred that any pivotal supplier has the ability to raise prices substantially in that period. 22 Energy only is also the approach taken by the Australian and some European systems (although there may be resource adequacy or financial hedging requirements at the local level).

Table 14.5. Status of resource adequacy market designs in US ISOs as of October 2005. Proposal under consideration

ISO-New England

Each LSE has a monthly ICAP requirement, which is a load-weighted share of ISO-NE’s objective capability (target for pool capacity). Central auction abandoned after high clearing prices experienced in 2000.

New York ISO

LSEs required to obtain ICAP, differentiated by location. Demand curve used to set price in monthly spot auctions of ICAP. Curve is linear, starting at level where ICAP credits are valued at 150% of the cost of a new CT, and sloping linearly until it reaches 0 at 132% of peak load. ICAP requirement: all LSEs responsible for meeting their share of annual IRM requirement set by PJM a year in advance. Own assets, contracts, demand response and tradable ICAP credits can be used to meet capacity requirement. No locational component. Short-term resource adequacy requirement in place.

The expense of extensive reliability must run contracts in load pockets motivated development of the ISO-NE Locational Installed Capacity (LICAP) proposal, currently in hearings before FERC. Includes locational demand curves for capacity with market clearing determined by a transmission constrained linear program. Unique features include forfeiture of LICAP payments if a generator is unavailable during times of system stress, and LICAP payments adjusted downwards on an annual basis by amount of gross margin earned by a reference CT in that year. However, LICAP has been extremely contentious politically because of concerns that it will unduly increase prices in load pockets. The NYISO has submitted a hybrid proposal to its stakeholders for consideration which incorporates a voluntary forward capacity market for procurement of a portion of its future resource. The NYISO, ISO-NE and PJM continue to discuss making their capacity markets compatible to encourage capacity trading.

PJM

Midwest ISO

ERCOT

California

No resource adequacy requirement in place. In 2001, Public Utilities Commission of Texas (PUCT) said in 2001 that it would not adopt resource adequacy requirement until reserves were forecast to fall below 12.5%. No resource adequacy requirement in place.

In response to concerns about inadequate capacity in eastern PJM, RPM filed at FERC August 2005. Unique feature: a 4-year ahead auction for capacity. Price in auction to be set by a downward sloping administrative demand curve determined in part by cost of CT capacity, net of expected gross margin of baseline CT. Will have locational component. LSEs and generators can also bilaterally contract. MISO has stated that it will implement a more permanent resource adequacy plan. Midwest Independent System Operator (MISO) not presently intending to develop ICAP-like markets, but rather rely on bilateral contracts for capacity and forward energy to promote resource adequacy. In 2005, PUCT initiated a rulemaking on resource adequacy, and is anticipated to make a ruling. There is the likelihood that it will propose to keep the energy only market, with $5000/MWh rather than $1000/MWh price cap.

California Public Utilities Commission (CPUC) proceedings underway considering alternative resource adequacy mechanisms. Likely to feature a forward contracting requirement for LSEs, possibly with a demand curve-based “backstop” administered by the CAISO.


Present status

504

ISO


505

14.6.2. Long-term contracts or options for energy A second approach is a regulatory requirement that retail suppliers hold long-term contracts or options for energy, perhaps with a stipulation that the options be backed up by physical generation assets (Vazquez et al., 2002; Oren, 2005). This is the proposal made in 2003 by the Public Utility Commission of California (CPUC, 2003). Another version of such a design is the forward reserve market implemented in New England (Cramton et al., 2005). Such a requirement could be complemented by a control system (and political will) that in the event of a shortage would first curtail consumers who lacked such contracts; this would partially correct the demand-side market failure by converting the public good of capacity into a private good that consumers would be willing to buy and generators would be paid for. 14.6.3. Payment mechanisms for capacity A third type is payment-based mechanisms, where the system operator provides a fixed or variable payment per MW for capacity, subject perhaps to performance penalties. This approach has not been proposed in the USA, but is presented here for completeness. The payments can take two forms. The first is a payment for installed capacity (ICAP) separate from payments for energy, as was done in Argentina until March 2000. In Spain, the capacity payments are similar to stranded investment compensation (Oren, 2005). The second form of capacity payment is an uplift in the energy payment that depends on the state of the system and the capacity availability. For example, the England & Wales market included such a payment based on LOLP before March 2001. 14.6.4. Quantity requirements for capacity A fourth approach is quantity-based methods, in which either a market operator procures reserve capacity directly (as Sweden has) or sets up a capacity market.23 There are several flavors of capacity markets, but each has the following basic features. First, there is a target level of system generating reserves (commonly based on a probabilistic adequacy criterion of capacity deficits occurring only once every decade). In the USA, this is often a state regulatory decision. Second, the ISO must allocate responsibility for meeting that target by creating an obligation (either on the part of load-serving entities (LSEs) or the system operator itself) to acquire capacity or capacity credits.24 Third, there is a system to assign credits to generators, based on their capacity and reliability, and perhaps to demand-side programs

23

As each large electricity consumer or company that serves end use consumers must contract sufficient capacity resources to meet his peak demand plus a reserve margin, the sum of all generating capacity should exceed total peak demand by the same reserve margin. This margin is calculated by the regulator or ISO to obtain a certain level of reliability. For instance, an “over-under” analysis (Cazalet et al., 1978) might be undertaken to determine what level of reserves yields the minimum sum of capital, operating, and outage costs, based on an assumed value of lost load. A regulator might specify that the reserve margin achieving the minimum cost be chosen. But since the relationship of capacity additions and expected outage costs is asymmetric, this suggests that in the presence of uncertain load growth, it is better to have too much capacity than to have too little. The market power problems accompanying low reserve margins only reinforces that point. Hence, the best reserve margin might be somewhat to the right of minimum. 24 The LSE is the entity that has retail load under contract. Hence, if a distribution utility buys power under contract from an independent generator, the utility is the LSE.

506


such as load controls that can substitute for capacity. Fourth, there is trading of such credits so that those with credits beyond their needs can sell them to those who are short. Fifth, there is a set of requirements defining how far ahead of time (days, months or years) those responsible for obtaining capacity must contract for it. And sixth, there is a system of incentives to encourage availability of capacity when needed, and for penalizing LSEs who have insufficient credits. Quantity-based systems include the ICAP markets of the northeastern US ISOs, in which the traded commodity represents “iron in the ground;” ACAP markets, in which capacity is given credit on a day-by-day basis only if it is available on that day; and scarcity pricing of operating reserves, discussed earlier, in which the system operator states a maximum willingness to pay if it is short of spinning or non-spinning reserves. In the latter case, if the maximum willingness to pay is sufficiently high, then generators will receive enough extra revenues to pay for its capital costs from either the operating reserves or energy markets when reserves are short. High operating reserves prices spill over to the energy market during shortage periods because most generators can choose to sell in either market, and so there will be an opportunity cost to selling in the energy market. We turn briefly to examine the ICAP market model in more detail, because it is an important feature of several US markets that evolved from cost-based power pools. The system of ICAP requirements uses market incentives to implement the chosen installed reserve margin (IRM), so that the reserve capacity is acquired at the lowest cost. The primary market incentive consists of the fact that rights to ICAP resources are tradable. Thus the provision of ICAP resources to loads occurs within a competitive market. In the US markets, most ICAP credits are either self provided by vertically integrated utilities serving their own load, or bilaterally contracted; a small percentage is traded through the centralized ICAP auctions. In states with retail competition, a daily ICAP auction is desirable to allow for daily adjustments of competitive suppliers. An important characteristic of ICAP systems is that these resources do not necessarily have to be available or on-line at particular times, such as the system peak; rather, they are physically available and in operable condition for most of the year. The fact that they may not necessarily be available at moments of supply scarcity is one of the key weaknesses of ICAP. Furthermore, the market need is for an instrument that is more forward looking that can be used to finance construction in the year to several year-ahead horizon. Finally, ICAP markets typically have no locational requirements inside ISOs.

14.6.5. Demand curves for capacity A fifth approach is a hybrid of the price- and quantity-based approaches. It involves the market operator creating a downward sloping demand curve that pays more for capacity if reserves are short, and provides some payment even when there is significantly more capacity than the amount needed to attain a given reliability standard. The motivation for using a demand curve is that capacity prices will be less volatile, providing a more predictable stream of revenues for generators that they “can take to the bank.” In contrast, a pure ICAP system will, in theory, bounce between two extremes, depending on whether there is too little capacity or too much relative to the target. The upper extreme is the penalty that LSEs pay if they have insufficient credits, while the lower extreme is zero. Use of such a demand curve may also help moderate market power, since a pivotal supplier would no longer be able to force the ICAP price up to its effective ceiling simply by withholding just enough capacity so that the market has fewer credits than the target reserve margin requires. A variant of the fifth approach is the operating reserves markets we described earlier in which a market operator has a downward demand curve for reserves (Stoft, 2002).


507

There is presently no consensus as to which approach to capacity market design among those tried or proposed (including none) is best. New design elements are being devised at this writing to address perceived shortcomings of existing approaches. For instance, the New York ISO has a locational ICAP system, so that capacity in different locations represents separate commodities and has separate prices. Meanwhile, the New England ISO is proposing a locational requirement and a set of penalties to address the criticism that ICAP payments reward “iron in the ground” and not performance during times that capacity is truly needed. To encourage those who receive ICAP payments to increase their availability, New England proposes reducing ICAP payments in each time period by the proportion of time that a generating unit is unavailable when energy prices exceed a certain threshold (Cramton and Stoft, 2005). Finally, PJM has considered differentiating capacity in another way, according to its flexibility. The argument is that capacity with quick start times or fast ramp rates is more valuable to the system than inflexible capacity, and should be rewarded (for further description of the PJM approach, see Table 14.5 and chapter by Bowring on PJM, this volume). The reason why the energy market might not yield the right amount of flexible capacity is price caps, which mean that the ability of a generator to quickly respond to a price spike will not be as rewarded as it would be in an uncapped market. A counterargument is that appropriately designed AS markets would be as or more efficient a means to reward flexibility, while some (e.g., the apparent majority of participants in the Australian market) say that removal of energy market price caps is the right response to this need. 14.6.6. Present status of capacity markets in the USA In response to some of the issues described above, several of the ISOs in the USA are in the process of reforming their capacity market structures. Several of the proposed designs are controversial and FERC continues to assess them. Table 14.5 summarizes the status as of this writing. 14.7. Market Power Monitoring and Mitigation Market power is the ability of a seller or buyer to alter the market price – that is, to raise it or lower it from the competitive level. Through analysis of the market and repeated interaction with other sellers and buyers, market participants learn how much market power they have and attempt to increase their profits accordingly. We have discussed market power as an issue in generation market design several times in passing, first as one reason why otherwise well-designed short-term markets can yield inefficient (or inequitable) outcomes, and then as a reason why capacity markets are sometimes used (in concert with market power mitigation) so that the market does not have to rely on high energy prices to elicit investment in generation. While market power is a matter of degree, and is difficult to quantify accurately, all governments have laws and regulations (or the ability otherwise) to limit the exercise of market power in the electricity sector, for purposes of improving efficiency and equity and also for addressing the interests of political constituencies. In this section, we will discuss the sources of market power, methods of mitigation in the USA, and design requirements for effective market power mitigation. 14.7.1. The sources of market power in generation markets Suppliers in generation markets attempt to exercise market power either by reducing the physical availability or output of a generator (from its true operable capability) or by changing

508


its offer price from its marginal cost such that it produces less (or more) that it would otherwise. In US regulatory parlance, the former is sometimes called “physical withholding,” while the latter is called “economic withholding.” In a market with fewer suppliers, or in which there are very large firms, there is a greater capability to withhold profitably. Further, transmission congestion creates bottlenecks that magnify market power in certain locations (and at least some generation was built in part to provide transmission support and thus “must run” for the sake of reliability). Suppliers may actively congest or decongest transmission constraints to enhance their locational market power in energy (i.e., collect congestion rents) and to affect revenues from transmission rights (Joskow and Tirole, 2000). In addition to market concentration, market power in electricity is greatly exacerbated by the lack of storability of electric power and the requirement of second-by-second balancing. As noted, short-term demand for electricity is largely inelastic and, at least in the USA, there are regulatory barriers to increasing price responsiveness. Hence, when there is market scarcity, and all suppliers are needed to make available or run most or all of their units to meet demand, then the ability to exercise market power becomes more acute.

14.7.2. Mitigation of market power The regulatory approach to market power monitoring and mitigation is somewhat different in each country, reflecting the prevailing law and regulation, historical evolution of the industry and other factors. For example, in the Australian national market, which is an energy-only wholesale market, suppliers are subject to a AUS $10,000/MWh price cap until a cumulative price threshold is reached, after which the offer cap becomes much more restrictive.25 In contrast, as shown in Table 14.6, the US ISO markets are typically subject to lower offer caps. This then requires that market design provide revenues that would otherwise be obtained at a competitive market price during shortages (or at other times), such as through administrative scarcity pricing (of energy and reserves) or capacity markets. A market power mitigation regime that results in both short- and long-term economic efficiency has its foundation in the efficient market designs that we have discussed elsewhere in this chapter. FERC has established essentially four primary types of market power mitigation, which will be explained in the following sections. The first is ex ante market-wide concentration analysis, applicable to the “destination markets” of all wholesale market sellers. The second is ex ante screening and mitigation of ISO spot markets. The third are behavioral rules, again applicable to all market participants. The fourth are ex post refund proceedings for participants found to have sold power at market-based prices that are not deemed just-andreasonable.

14.7.3. Long-term ex ante screening for market-based rates authorization Prior to allowing suppliers to sell at market prices, FERC requires a market concentration analysis of the supplier’s destination markets under various market conditions, using various market metrics (percentage market share, sum of squared market shares or pivotal supplier determination), to infer whether its pricing will be sufficiently competitive. If the 25

Specifically, the market’s “cumulative price threshold” is reached if the sum of market prices reaches $150,000 in any 7-day period. At that point, the market operator imposes an administered price cap of $100/MWh between 7 AM and 11 PM on business days and $50/MWh otherwise.

Table 14.6. Rules for screening and mitigating offers into US ISO energy markets.

Safety net offer cap

Rules outside load pockets Triggering condition for offer screening

Offer conduct test

Market price impact test

Offer mitigation

RP ⫹ 8760 ⫻ average price in RT Market over prior 12 months ⫻ (2% ÷ total constrained hours over prior 12 months) Net annual fixed cost/ expected run hours

LMPs ⭓ $150/MWh; All suppliers

Lower of 300% increase or an increase of $100/MWh over RP

LMP increases by 200% or $100/MWh

Market RP

Supplier is pivotal


Market RP

Binding transmission constraint and suppliers with threshold generator shift factor on that constraint MCP must be ⭓$91.87/MWh

Lower of 50% increase or an increase of $25/MWh over RP Lower of 300% increase or an increase of $100/MWh over RP


Market RP

Lower of 200% increase or an increase of $100/MWh over RP

Lower of a $50/MWh or 200% increase in the MCP compared with a reference MCP in which all bids failing the conduct test are replaced

Market RP

Offer caps

New York

$1000/MWh

New England

$1000/MWh

Midwest

$1000/MWh

Net annual fixed cost of a new peaker ÷ total constrained hours over prior 12 months

California

$250/MWh

PJM

$1000/MWh

Currently managed through RMR contracts; lower of $50/MWh or 200% greater than the MCP OOM generators capped at marginal cost plus 10%

No mitigation of in-merit generation; OOM generators capped at marginal cost plus 10% (i.e., cost-based RP)

509

Abbreviations: OOM – Out-of-merit; MCP – Market-clearing price; RP – Reference price. Source: This table is based on the table in Helman (2006).


Rules inside load pockets

510


supplier fails these screens and cannot take steps to mitigate its market power (e.g., by proposing a mitigation plan that FERC accepts or by joining an ISO market and being subject to its mitigation rules, as described next), its application is denied and its wholesale energy must be sold at cost-based rates (e.g., FERC, 2004a, c).

14.7.4. Ex ante mitigation of ISO spot markets With the advent of ISOs in 1998, a new type of market power screening began to evolve. On the one hand, centralized ISO auctions allowed for the instantaneous participation of potential wholesale suppliers from larger regions than were previously possible. As such, there was the initial expectation that if the generation capacity within the ISO boundaries was considered as a “geographic market,” then market concentration would be low and hence market power would be greatly diluted (e.g., FERC, 1996a). But on the other hand, the locational spot pricing introduced by most ISOs highlighted another factor: the marketdividing effect of transmission congestion and the lack of price-responsive spot demand. In combination, these factors could result in large spot price jumps when system supply was tight and transmission capacity limited the import of power to a location. Those locations, often cities or towns, that faced persistent binding transmission constraints on imports are commonly known as “load pockets” (although transmission constraints can emerge anywhere on the network). In load pockets, for some number of hours in the year, suppliers could, if they chose, raise prices enormously by exerting market power. This fact drove some ISOs to establish the two-tiered approach to generation market power summarized in Table 14.6: looser spot offer restrictions outside the load pockets and tighter offer restrictions or cost-based contracts in the load pockets (or more generally for any generator whose output was affected by transmission congestion). Outside the load pockets and other locations affected by transmission congestion, the ISO market could be likened to a large “pool” of power that is relatively un-concentrated. In these areas, the standard rule is to impose a $1000/MWh “safety net” spot offer cap. New York ISO later established a method called a “conduct-impact” test: if the supply offer exceeds a marketbased average reference price (RP) by a certain amount and changes the market-clearing price by a certain amount, then it is mitigated to the RP. This approach has been taken up by other ISOs as well (see Table 14.6). Other ISOs determine whether suppliers are “pivotal” prior to mitigating their offers. The ISOs re-run the auction until clearing prices are calculated that meet the market power mitigation rules. But in fact, because the sellers know their RP and the bounds established by the conduct-impact test, they rarely trigger mitigation.

14.7.5. Behavioral rules Following the California and Western US price spikes of 2000–2001, suppliers are subjected to new behavioral restrictions in the forward markets, such as not violating market rules and not misrepresenting fuel or contract prices (FERC, 2003a, b).

14.7.6. Ex post refunds and settlements If one or more of the prior methods of market power screening and mitigation fails, and FERC determines that tariff rules have been violated, then the offers submitted by firms that exercised market power could be reexamined and prices recalculated as a basis for refunds.


511

The major recent example of this was the refund proceedings for the California and Western markets in 2000–2001 (e.g., FERC, 2001a-c). 14.7.7. Lessons learned and performance evaluation Each of these methods of market power mitigation involves regulatory determination of the appropriate level of market shares, market offers, market prices or other types of market behavior to reduce the potential for, or exercise of market power. As such, regulatory errors are likely: either over-mitigating in some cases, driving the price paid to suppliers below long-run competitive levels, or under-mitigating, and allowing some generators to make profits well above competitive levels. Particular difficulties are presented when estimating a competitive benchmark price by factors such as recovery of unit commitment costs, intertemporal constraints (e.g., due to emissions restrictions and energy (fuel) limitations) and the economies of scale in generation investment mentioned above (e.g., Harvey and Hogan, 2002). The focus of US regulation of market power in recent years has been to try to reduce such errors by reducing market uncertainty about mitigation rules (e.g., clarifying what types of market behavior are allowed and making spot market screening more transparent and ex ante) and through modification of market design. Regulators must also be cognizant of the offer incentives created by mitigation rules. For example, when market-based RP are used as a benchmark for allowable ISO energy market offer price increases, the supplier may seek to increase the RP over time to create more latitude for affecting market prices within the rules. Strict offer caps (e.g., $1000/MWh) also become “focal points” for offer prices in repeated auctions, that is, prices to which suppliers converge over time under certain conditions (e.g., reserve shortage) because they have greater confidence that other suppliers will also offer at around that price. In general, empirical analysis of price-cost margins in the US ISO markets outside the congested areas has only found significant evidence of market power during the California price spikes (see Table 14.7). This suggests that the high caps in these parts of the markets have been sufficient, as long as market design is efficient. In many ISO markets, where market power mitigation can be developed in close relationship to market design, the current focus is on improving locational market power mitigation, because the presence of transmission constraints makes it difficult not to make regulatory errors for specific-generating units. The more restrictive the mitigation, the harder it is, all other things equal, to attract investment to locations where the spot price is suppressed below competitive levels. One example was that some suppliers in New England load pockets failed to recover fixed costs even when they were allowed to factor those costs into their spot offers.26 Similar issues arose in PJM, where in 2004 suppliers complained to FERC about the strictness of offer capping in transmission constrained locations (by the rules shown in Table 14.6). FERC determined that this offer capping was not just and reasonable for generators that are capped for 80% or more of their run hours and are not recovering their costs 26

The New England ISO’s Peaking Unit Safe Harbor (PUSH) rules allowed owners of low capacityfactor units (less than 10% annual capacity factor) in load pockets to include fixed costs in their energy offers without risk of mitigation in order to increase opportunities for fixed cost recovery and to produce signals for investment through higher prices in these areas during periods of scarcity. The evidence shows that PUSH did not achieve its intended purposed of adequate cost recovery (ISO New England, 2003b).


512

Table 14.7. Estimated price-cost margins in US ISO and RTO energy markets, 1999–2004 (all values load-weighted averages unless otherwise indicated). 1999

2000

2001

2002

2003

2004

California ⫺0.02–0.31 PX ⫺0.16–0.63 PX 0.2–0.4 ISO 0.14–0.23 ISO 0.01–0.23 ISO New England 0.10a ⫺0.03–0.06 0.06–0.11 ⫺0.04–0.09 ⫺0.06–0.03 PJM 0.02b 0.04 0.02–0.09 0.02–0.11 0.03–0.12 0.03–0.08 a

May–December. bApril–December. Source: The California analysis is found in Bushnell et al. (2002) and California ISO (2003–2005); an average for the year 2001 is not available in the public literature. New England analysis is found in Bushnell and Saravia (2002) and ISO New England (2002, 2003a, 2004, 2005). PJM analysis is from PJM (2000–2005). The ranges for PJM are the difference in the Lerner index between using marginal cost estimates (lower estimate) and marginal cost plus 10% estimates (higher estimate). The ranges for California before 2004 and New England reflect the difference in the Lerner index calculated using the same marginal cost offer stack but based on either the mean bid price submitted that would clear the market (lower estimate) or the market clearing price (higher estimate). Both of these clearing prices might be lower than the marginal cost of the marginal unit. The California ISO estimates for 2004 are the range of monthly average markups calculated using a “single resource portfolio” methodology (California ISO, 2005: 2–15).

(FERC, 2004b). PJM subsequently relaxed the offer capping by applying a market concentration test: if the generator that is being re-dispatched due to a binding transmission constraint is “pivotal” with respect to relieving the constraint, or if it and one or two other firms are jointly pivotal, then it is offer capped. Otherwise, it is not offer capped. This was considered sufficient to provide the needed generator revenues while not compromising market power mitigation. Upon review of the PJM case and similar ones, FERC determined that there was no standard method to ensure fair cost recovery for generation units possessing locational market power, especially those needed for reliability, and to promote entry (FERC, 2004b). The market design solution that is appropriate to ensure fair generator revenue recovery and promote entry – administrative scarcity pricing, locational reserve or capacity pricing, forward reserves, entry/exit auctions, and so on – has varied from region to region. Indeed, different market design solutions could lead to similar results. Hence, there is currently a proliferation of design schemes to address the market impact of locational market power mitigation and provide long-term generator revenue sufficiency. 14.7.8. Alternatives to behavioral restrictions There are other regulatory methods to reduce market power: forced divestiture of generation assets (which is not a regulatory option in the USA, although voluntary divestiture has taken place under regulatory agreements), increased investment in transmission to relieve constraints and thus increase the scope of the generation market, and investments in demand response. Each has its advantages and limitations. For example, changing the structure of the market to reduce concentration will certainly result in less need for regulatory controls. However, divestiture may not work appropriately the first time, and may need to be repeated, as was the case in England & Wales market. Moreover, divestiture can diminish some benefits of economies of scale and scope. Similarly, if the regulator provides incentives for increases in regulated transmission investments, this may create uncertainty among market-driven generation investors, given that spot prices are often affected by transmission constraints.


513

14.8. Design of ISO transmission markets ISO transmission markets did not emerge in a vacuum. Much of the US transmission grid is privately owned and there were a number of transmission contracts in place before the ISO markets were created. Prior to open access, transmission-owning utilities calculated how much transmission capacity was available on their grid after they had served their (regulated) retail customers and had also set aside some transmission for reliability purposes. They then sold at their discretion the remaining transmission capacity to other parties, typically other utilities, on the basis of average costs. FERC approved these transmission rates. Generally, the transmission selling utility internalized the costs of managing congestion resulting from its transmission sales (although a few contracts had an incremental rate for redispatch or losses). When a threat to reliability emerged that the transmission selling utility could not manage through re-dispatch of its own generators, such as a potential overload of a transmission line (i.e., congestion) due to power flows from outside its system, there was a “last in-first out” system of transaction curtailments that proceeded to cut other parties’ transmission service roughly based on a priority ranking (less expensive “non-firm” transmission service was curtailed before “firm” transmission service). This administrative method of congestion management was inefficient. Moreover, transmission contracts were typically specified along a simplified “contract path” between the end points of the contract that did not recognize the externalities associated with the actual power flows, which would affect other parties’ transmission systems and have to be managed accordingly. Power pools sometimes had more efficient methods of cost-based re-dispatch for their system, but the cost assignment could sometimes embody cross-subsidies, charging less congested parts of the system on the same basis as more congested parts. As noted above, Orders 888 and 889 (FERC, 1996a, b) established the terms for open access to electricity transmission, limited to utilities under the FERC jurisdiction. Two types of transmission service were created: ●

●

“network” transmission service for a transmission customer embedded in the network of a transmission-owning utility that provides for economic dispatch from multiple generators and delivery to multiple customers; “point-to-point” transmission service, for a customer seeking to inject power at a particular generator bus or to cross the utility’s network on the way to its own load.

ISOs essentially had to meet the basic requirements of these types of transmission service or provide comparable service. Moreover, they had to map the pre-existing rights into the new markets with locational pricing and transmission property rights that provided efficient dispatch incentives. How they would do so pre-occupied market designers for much of the first decade of regulatory reform and continues to present challenging issues. 14.8.1. Market design debates Pre-ISO transmission rights were essentially “physical” scheduling rights and some initial observers of the proposed ISO markets argued that a scheduling rights model should be continued. Although the design of such rights in an ISO context was never fully elaborated, physical rights were often defined as rights that the party using the grid would have to hold to physically schedule. That is, if the party wanted to inject at location A and withdraw at location B, they would have to hold capacity rights to the transmission paths joining those locations. Two refinements of this model were common. First, it was evident that if physical

514


rights were required for all transmission paths between a point of injection and point of withdrawal, then the physical rights model would quickly become unwieldy, as parties would have to hold rights to large number of paths. Thus, most advocates of physical rights argued that zonal aggregations were necessary and that rights would be defined between large zones. Empirical support was offered that with such zones, intra-zonal congestion would be minimal and thus would not cause significant cost shifts (e.g., Walton and Tabors, 1996). However, in practice the desire to have relatively few zones for commercial simplicity has typically led to violation of those assumptions (Baldick, 2003). Second, there was concern that physical transmission rights could be used to exert market power in the generation market, by restricting transmission capacity and therefore entrenching a right to “physically withhold” transmission capacity. Hence, a “use-or-lose” rule was often proposed that would require holders of such rights to release them into the transmission market prior to RT energy market if they were not used. The rejoinder to the physical rights model was the financial rights model (Harvey et al., 1997). This approach began from the observation that hour-to-hour transmission system operations typically required frequent adjustments in available transmission capacity that made a forward physical allocation of that capacity subject to possibly significant error. Parties would thus be subject to constant misalignments between the physical capacity that they owned rights to and their actual physical use of the system. That error would be compounded if the rights were zonal, and thus largely ignored transmission constraints within the zone, resulting in potentially high intra-zonal congestion charges. Moreover, the zonal price would provide only a weak locational signal for transmission expansion decisions and generation location. Instead, a financial transmission right would not be required for physical scheduling but would rather be used to hedge congestion charges incurred when scheduling in the ISO markets. These financial rights would be “nodal”, corresponding to the LMP of energy discussed above. As long as the set of nodal injections and withdrawals were feasible in the dispatch, with the financial right it did not matter which transmission lines the power flowed on: the right would re-pay sufficient revenues to cover exactly the congestion charges, within the specifications of the right. Heretofore, all the operational ISOs have implemented financial point-to-point rights. In several regards, this decision has been demonstrated correct through experience. Most importantly, the location of congested paths in the ISO markets has proven difficult to predict as power markets expand and market conditions change. The California ISO’s experience is instructive, where intra-zonal congestion costs generally exceeded inter-zonal costs. During 2004, for instance, intra-zonal congestion in the southern California zone arose when new plants were sited in Mexico behind transmission constraints that greatly restricted exports in that generation pocket. That generation would be fully scheduled day ahead in the zonal market, and then would have to be (expensively) “dec-ed” in RT. The cost of RT dispatch to clear intra-zonal congestion was $103 million in 2004, while inter-zonal congestion was half that amount (CAISO, 2005). Zonal rights are thus clearly inefficient. One argument set forth by the advocates of physical zonal rights has merit: for forward power trading, point-to-point rights may not be the most tradable instrument for a supplier that owns a portfolio of generators in one area and may serve a number of different buyers in another. Such a supplier would prefer a right sourced from with its portfolio that hedges the bulk of the congestion charges when transferring power between the areas. A later attempt to capture this desirable feature of the zonal right was the “flowgate” right (Chao et al., 2000). Rather than fixed zones, the flowgate right returned to the idea of hedging specific transmission paths. The key empirical claim was that in most ISO territories, there would be few such “significantly congested” paths. Hence, the rights to those paths could


515

be traded more readily. However, advocates of the flowgate right typically suggested that the awards of the flowgate rights should carry ISO-backed “insurance” to guarantee that the holder would never face congestion costs if the system topology changed or if its transmission usage encountered congested flowgates for which it did not hold rights. A synthesis proposal suggested that transmission users could voluntarily nominate financial flowgate rights without any ISO insurance and that such rights should be offered simultaneously with point-to-point rights (O’Neill et al., 2002). That concept was later taken up in FERC’s SMD proposal (FERC, 2002b). Other concerns about financial point-to-point rights have been raised: that they are not sufficiently long term, that they are difficult to value, and that they do not provide the right incentives for transmission investment. We examine the properties of such rights next. 14.8.2. Definition and attributes of FTRs A point-to-point right specifies a source node, where power is injected, a sink node, where power is withdrawn, and the quantity (MW) injected and withdrawn. The source and sink nodes can be a single bus on the transmission network or a zone or hub.27 The right also specifies the period of time that the right is active: most ISOs offer rights differentiated by month, season and daily peak and off-peak hours. Point-to-point rights are settled against the locational marginal prices in the ISO DA market (or the RT market if there is no DA market). Rights settled DA do not cover congestion charges accruing for deviations settled in the RT market. The primary reason for this is that it creates an incentive to submit accurate DA schedules (which assists the ISO’s next day planning and reduces the need for additional reserves). These rights may also exclude the cost of transmission losses, as in the California ISO’s market reforms proposed for 2007. Point-to-point rights have two further characteristics: obligations and options. An obligation right confers the right to collect positive congestion revenues, when the locational marginal price at the source node is lower than the price at the sink node, but also the obligation to pay negative, or “counterflow”, congestion revenues, when the price gradient reverses. In contrast, an option right confers the right to collect positive congestion revenues, but not the obligation to pay negative congestion revenues. As such, the ISO cannot assume any counterflows when auctioning or allocating option rights, as discussed below. 14.8.3. Allocation and trade in FTRs Heretofore, ISOs have awarded 1-year transmission rights to the existing grid on an annual basis, followed by monthly auctions for re-sale or reconfiguration or for release of any additional transmission capacity (incremental long-term rights are available for transmission expansion). Transmission users are generally eligible each year for rights up to their peak load or to fulfil point-to-point contracts. There has been one exception to this annual process: the New York ISO offered 2- and 5-year obligation rights in one auction in 2000. However, due to difficulties in valuing the rights, New York reverted to annual rights thereafter. 27

A “zonal” or “hub” point-to-point right can be approximated by assigning the quantities withdrawn to a number of source or sink nodes. For example, if the right sources 100 MW in a zone, but the exact source generators may vary each hour or season, then the right can distribute the 100 MW over the possible sources so as to improve its hedging quality.

516


Under the Energy Policy Act of 2005, FERC will require ISOs to define longer-term rights to the existing grid in order to provide a more consistent hedge and possibly to make financing of remote generation siting more predictable. Each allocation or auction of transmission rights must ensure that the auctioneer (the ISO) is revenue adequate. This is done through a “simultaneous feasibility test”, which is an analysis of the physical feasibility of the injections and withdrawals assumed in a set of transmission rights. The test is conducted using a power flow model of an assumed test system.28 For any set of simultaneously feasible FTRs, as long as the system topology remains unchanged from the test system, the ISO will remain revenue adequate (Harvey et al., 1997). That is, the ISO will collect sufficient congestion charges in each hour of the DA market to pay all holders of FTRs settled in that hour. In reality, however, revenue insufficiency often occurs, as we discuss below. There are two basic methods for allocating financial rights: direct allocation of the rights themselves or direct allocation of auction revenue rights (ARRs) followed by an auction of the FTRs. In the direct allocation model, followed by PJM for the first few years of ISO operation and then by the Midwest ISO and proposed in the California market re-design, parties nominate FTRs between their generation resources and the location of their load. There are typically several rounds of the allocation, allowing for less uncertainty over coverage of certain transmission paths, usually to cover baseload resources. If more rights are requested than are feasible, then some pro-rata adjustment is made of each party’s requests. In some systems, such as in the first year of the California market re-design (scheduled for 2007), the ISO is responsible for verifying that the requested rights actually are consistent with contracts held by the requesting LSE. In subsequent years, the California rules propose that portions of previous rights requests can be “grandfathered”; this substitutes for a long-term transmission right. However, it is important that such prioritization of requests not be dependent on going forward decisions (such as location of new generation), because this has the potential for distorting investment location and contracting choices. The advantage of the allocation model is that it removes most of the uncertainty that market participants may face when entering an auction; a disadvantage is that it does not require valuation of the rights and thus may encourage inefficient transmission use. In the auction model, implemented first by the New York ISO and then by PJM and ISO New England, the market participants are first allocated ARRs, which will return to them auction revenues. They can then enter the auction and bid sufficiently high for the transmission rights that they want knowing that they have claim to the resulting auction revenues. Alternatively, they can choose not to buy through the auction and instead collect what others are willing to pay for the rights. 14.8.4. Revenue shortfalls When the topology of the transmission system changes, due to line deratings or outages or due to changes in generator status that could affect transmission transfer capability, the set of FTRs may become revenue inadequate. This has occurred often in PJM, for instance, where transmission rights achieved 98% revenue adequacy in 2003–2004 (PJM, 2004; for more details, see chapter by Bowring on PJM, this volume). Inadequacy can also occur

28

PJM initially used an DC load flow model while New York ISO used a AC load flow model. The trend seems to be toward using the DC model for solution tractability, computing speed and to allow for representation of options.


517

because of the non-convexities of AC load flow that are not considered in the linearized simultaneous feasibility tests used by ISOs (Lesieutre and Hiskens, 2005). When inadequacy occurs, ISOs must then determine how to cover their revenue obligations to the holders of financial rights. There are several approaches. PJM created a fund of surplus congestion payments (payments made by transmission users that did not hold transmission rights) and taps into that fund whenever transmission rights are revenue inadequate. However, when the fund is empty, PJM reduces payments to financial rights holders on a load-ratio share basis. In contrast, for a number of years, New York ISO would keep holders of financial rights revenue sufficient by charging any revenue shortfalls to the transmission-owning utilities (Potomac Economics, 2005: 46–47). 14.8.5. Performance of financial transmission right markets Experience in the eastern ISOs is that liquidity in the secondary market for rights is greater for auction systems than for allocation systems. Especially when regulated LSEs are allocated rights, there is a danger that owners will just hold them whether or not they are useful for hedging risks; this can put other market parties at a disadvantage. Experience also shows that auction valuations of rights can be much greater than the actual congestion revenues that are yielded, and that the correlation between auction prices and congestion payments is poor (Bartholomew et al., 2004). This was the case in the NYISO, but not in PJM. Such occurrences may be because regulated LSEs play large roles in the market and have objectives other than profit maximization. Recently, concerns have been raised that the annual allocation process, with the uncertainty that it creates over the year-to-year availability of rights, is problematic for some market participants who seek long-term property rights for planning, contracting or financing (FERC, 2005b). As noted, New York ISO is the only market to have offered rights of greater than 1 year duration and only in one auction. Providing such long-term rights in a financial rights framework is not a simple matter: long-term obligation rights could be financially risky and longterm option rights could foreclose system capacity for other parties’ transmission rights. Although the scale of the problem is not clear, it is likely to require a careful design response. Another issue concerns the role of FTRs in the business model of merchant transmission. In general, ISOs grant point-to-point financial rights to merchant transmission in amounts related to the extra capacity they provide to the system. This provides a revenue stream in the form of congestion revenues. However, except for radial or direct current (DC) lines, the amount of extra capacity is ambiguous, and there may be negative impacts on other rights holders. Furthermore, there are strong scale economies at least for expansions up to lower voltage and smaller scale expansions (Dixit and Baldick, 2003) and lumpiness in transmission construction. As a result, building a line to relieve congestion might eliminate most or all congestion, obliterating the value of the transmission rights that the line builder would receive. It is yet to be seen whether the FTRs revenue model can be a significant incentive for line construction with the exception of DC circuits between otherwise isolated or widely disconnected systems (Joskow and Tirole, 2005). Other methods for market-based pricing of transmission usage, such as “dispatchable transmission,” may provide additional revenues (O’Neill et al., 2005b). 14.8.6. Loss pricing methods Surprisingly, even though the value of transmission losses is typically of the same order of magnitude as congestion costs, their financial consequences and management have

518


received much less attention in the design of ISO markets. PJM has simply retained average loss pricing and its locational marginal prices for energy thus reflect only system energy and congestion components. Other systems, such as New York, New England and the California market re-design proposal, have adopted marginal loss pricing.29 Marginal loss pricing results in the ISO collecting a loss charge surplus. This phenomenon arises from the quadratic nature of losses, which implies that the average cost of losses are roughly one-half the marginal cost. As a result, an ISO, such as California, might collect on the order of $100 million per year in surpluses from charging for marginal losses in LMPs. Simplicity seems to be preferred in allocating that surplus; in the case of the California market re-design, it will be allocated back to LSEs in proportion to their loads. As a result, the omission of losses from FTRs in the California market re-design proposal has been entirely uncontroversial, and there has been relatively little debate over how to allocate the surplus that the ISO will earn from charging marginal losses. In contrast, debates over the allocation of congestion revenues have been complicated and intense, leading to very complex allocation and auction systems, as described earlier in this section.

14.9. ISOs and Reliability ISOs are not only market operators, but also reliability and planning authorities. As such, they are responsible for maintaining short-term reliability through contingency constrained scheduling, procurement of AS, market price co-ordination of generators and administrative measures, such as emergency procedures. Longer-term reliability is enhanced through the centralized forward markets for simultaneously feasible transmission rights, auctions for forward reserves and capacity, and regional expansion planning to co-ordinate marketdriven and regulated investments in generation, transmission and demand-response capability. Most of these topics have been addressed above; in this section we briefly turn to the functional reliability role of the ISO and to whether ISOs enhance reliability. There have been three general ways in which ISOs have taken over regional reliability functions (both NERC standards and procedures and regional procedures). In the first, where the ISO replaced a pre-existing power pool (PJM, New York, New England), it essentially inherited the regional reliability functions and procedures of the power pool. This transfer has typically taken place fairly smoothly. In the second, utilities that were not previously part of a power pool ceded all control area functions to the new ISO. This was the case in California. In the third, the ISO and the control area operators have established a functional division of labor that intrudes to some extent on market operations. For example, in the Midwest ISO, there is a transitional phase in which the control areas and the ISO share balancing authority functions. Day ahead, the control areas provide operating reserve and regulation settings for generators, which are then incorporated into the ISO’s DA schedule. In RT, the ISO sends financially binding dispatch signals to the generators in its territory and its calculation of net scheduled interchange between the control areas. The control areas are required to maintain balance within their territories. In general, FERC has encouraged all ISOs and RTOs to merge into one control area. The August 14, 2003, blackout was the first major failure of short-term reliability in the USA (and Canada) under electricity restructuring. Although it took place in the territory of the 29

Midwest ISO also calculates marginal loss prices, but for the first 5 years of operations will refund the marginal loss surplus so as to more or less recreate an average loss charge.


519

Midwest ISO, under the rules prevailing at the time, the Midwest ISO did not have a centralized dispatch (which it began in 2005) and while it had emergency authority over the region, it lacked the RT metering data to track the event and had only weak capability to conduct a response in the necessary time frame. As a result, the localized failures of a particular utility could cascade into an enormous regional event. The lessons drawn from the governmental investigation of the blackout were focused, not on failures of restructuring, but on making voluntary industry reliability standards mandatory with government oversight (later mandated in the Energy Policy Act of 2005) and on improving training and maintenance (US-Canada Power System Outage Task Force 2003, 2004). From the perspective of ISO operations, others have pointed out that with increased trading across regions and with the improvements in monitoring and control technology, rather than markets undermining reliability, it is the vertical control area structure that is no longer sufficient. The suggestion has been made that if the events preceding the blackout had begun within an ISO with centralized dispatch, it would have been contained or prevented (e.g., Alvarado and Rajaraman, 2003; Landrieu, 2004).

14.10. ISO Performance: Evaluation and Improvement ISO performance – including most notably operational costs, but also management skill, ability to correct design and operational problems and customer responsiveness – has been a contentious issue in some parts of the USA. Typically, there has been a learning curve in market operations, including discovery of design flaws and software errors, that lasts for at least 2–3 years after the market start. Performance then improves, but only if the market design is not flawed to begin with or is corrected in time. There is no question that costs have been too high in some cases and that better incentives are needed to keep costs down. However, in this section and in the prior one, we contend that a long-term view is taken of how the initial monetary investments in ISOs will affect future economic benefits, including reliability. Moreover, at least some of the investments that ISOs are making, particularly in information technology, reflect areas of relative underinvestment by the regulated utilities. Finally, in this section we note the role of quantitative tools for design decision-making. Such tools can help reduce design errors. 14.10.1. Costs and benefits of ISOs Cost–benefit analyses of national or regional restructuring have been conducted either to fulfil legislative requirements or as instruments in the debate over whether to implement ISOs or RTOs (e.g., USDOE, 2003). We will not examine these analyses in detail. They typically require deployment of large-scale electricity models and forecasts over 20 years or more. The suppliers in the models are assumed to be “competitive,” supplying at marginal cost. The benefits are derived from optimizing over a larger area (i.e., removing barriers to trade between utilities, typically represented as a “hurdle” rate on transmission) and through assumptions that generation will be operated more efficiently. The costs stem from the operational costs of ISOs discussed below. The cost–benefit ISO studies to date have remarkably similar findings.30 The short-term benefit is about $0.20/MWh, the long-term benefit is

30

Steven Henderson, “RTO Cost Benefit Analysis,” Harvard Electricity Policy Group, May 22, 2003. Available at http://www.ksg.harvard.edu/hepg/RTOs.htm

520


about $0.35–$1.00/MWh. As discussed next, the incremental cost of operating an RTO is in a similar range to the short-term benefits and on the lower end of the long-term benefits. Hence, long-term benefits are the key, although more speculative. There are also some preliminary empirical results on improved efficiency in the USA since restructuring. Markiewicz et al. (2004) find that investor-owned utility plants in restructured states reduced their labor expenses by about 5% and non-fuel expenses by about 10% compared to such plants in states that did not restructure. Douglas (2004) examines changes in the utilization of coal power plants in the eastern USA since in 1996 through 2000. He finds that utilization rates of low-cost plants increased relative to those of highcost plants after 1996, but only in regions with ISOs. Cost savings identified were on the order of 2–3%. Doing comparative analysis of ISO costs is difficult for several reasons. First, FERC has approved different designs as ISOs and RTOs. A now standard distinction in the USA is between a “Day One” ISO, in which the system operator is essentially a centralized transmission scheduler for the utilities in the region (and possibly takes on some emergency coordination functions), and a “Day Two” ISO, which would become the reliability authority and operate markets such as those we have described in this chapter. Clearly, Day One ISOs are cheaper to organize and maintain than the Day Two ISOs, unless the latter was able to convert existing centralized operations (e.g., a tight power pool) to the market functions – as was the case, for example, in PJM. FERC has estimated the annual operational costs of a Day One ISO as between $0.16/MWh and $0.22/MWh (FERC, 2004d). Estimates of the range of Day Two annual operational costs lie between $0.40/MWh and $0.58/MWh, although the higher range consists of outliers, such as the California ISO and PX, that had much higher than average costs (various studies, cited in FERC, 2004d, USDOE, 2003). The chapter on PJM in this volume offers additional data on each ISOs cost per MWh.

14.10.2. Advances in information technology and large-scale computation ISOs have been making large investments in information technology and improved computational capability, sometimes to great criticism. FERC (2002b: 194–198) recognized that, particularly in the early years of ISO formation, “software and data systems inherited from the [regulated utilities and power pools] are often idiosyncratic, making changes and seams issues more difficult than they should be. Market participants often find software to be impenetrable ‘black boxes.’ Software development and modifications have become expensive and software ‘wheels’ are being reinvented.” At that time, FERC urged that software become transparent, testable and modular, and there has been some evidence of that in the ISOs. In particular, after an expensive multi-year redesign effort, in 2003, ISO New England largely adopted the PJM market software and market rules. Although there have clearly been some wasteful expenditures, ISOs are also compensating for the under-investment in software under regulated vertically integrated utilities. Significant computer hardware and software advances in market design and reliability management have been made in the last decade (e.g., Hobbs et al., 2001a). Large-scale unit commitment and dispatch problems, once considered intractable, are now solved in several minutes.31 These

31

In 1996, a 300-node network market model in New Zealand was the state of the art. In 2004, a 30,000-node model with greater network detail is being solved faster than the 300-node model in 1996. Topology estimators have been introduced into state estimators.


521

advances allow reliability to be more fully integrated with market software. PJM has recently introduced new market software for unit commitment that promises savings of $50 million/ year.32 This level of saving can justify a significant amount of software investment. 14.10.3. Incentives for efficient management and innovation In the start-up phase, the ISOs have operated under a cost pass-through approach, which has weak efficiency incentives. ISOs should have incentives to be efficient, but not through capital-asset-based cost-of-service regulation. They are market operators and information conduits, not electrical equipment owner/operators. The asset base is human capital and mostly short-term information technology assets. Regulating by setting return on capital investment makes little sense. For example, a budget cap based on a reasonable start year, with adjustments for any geographical expansions, could be fixed and any subsequent cost reductions or efficiency enhancements could be linked to ISO employee bonuses and refunds to market participants. Performance measures (e.g., through benchmarking) are needed to maintain and reward quality services. 14.10.4. Quantitative tools for design decisions Market designs for spot markets and capacity markets, as well as detection of market power in those markets, can be tested with models. Although not a regulatory requirement, these sorts of investigations are increasingly used by ISOs for the fuller range of market design proposals that are now being considered. Anticipating and preventing market design problems is likely to be cheaper than correcting them after the fact, as California has learned. Models can be used in two modes. One mode uses models to communicate and explore differences among alternative market designs. This usually involves simple models, often spreadsheet based, that capture the essence of the alternatives and are applied to very small examples (e.g., the classic three-node model of an electric system). Such models have the purpose of documenting proposals in an unambiguous and clear way to members of the market design team and to stakeholders by allowing readers to work through and check all calculations themselves. This mode of using models has been used effectively by the California ISO and its consultants to explore the efficiency, market power and income distribution aspects of alternatives for example, market power mitigation, congestion rights distribution and averaging of LMPs within demand zones (e.g., Harvey et al., 2005; Rahimi, 2005). Sensitivity runs of simple models can also document possible problems with problems in a much more convincing way than descriptive text. A risk in using models in this way is that in illustrating alleged weaknesses of particular proposals, the examples chosen may be highly contrived and entirely unrepresentative of actual conditions that are likely to prevail in the market. A proposal that exhibits a weakness in one simple numerical example may actually perform well under nearly all conditions in the real, much more complex market. Or, a proposal that does well for a simple case may do badly when actually implemented. Thus, considerable judgment is needed when developing the models and reviewing their results. However, informed use of simple models, even if they are sometimes misleading in their simplicity, is better than qualitative argument because the ambiguities in the latter

32

See “MIP (Mixed Integer Programming) Based Unit Commitment”, PJM MIC Meeting, March 31, 2004, www.pjm.com/committees/mic/downloads/20040331-mip-based-unit-commitment.pdf

522


can hide contradictions or implications that would quickly become apparent if applied to a simple quantitative example. The second mode of using models is to commission the use of more complex models to explore particular market design issues in depth. This mode has the disadvantages of taking time and significant resources, while the complexity prevents verification of model results by design team members and stakeholders. On the other hand, credible testing of proposals may require detailed modeling of system features that do not lend themselves to simple models (such as the interactions of multiple players on a complex transmission system topology, or unit commitment decisions for multiple types of plants in the face of time varying demand) if those aspects are crucial to a proposal’s performance. Or specific estimates of the benefits and costs of particular changes in market design or structure might be required by a regulatory process, in which case simulations of the actual power system over a wide range of system conditions may be required. Models used in this mode are of several types , including equilibrium models (e.g., Hobbs et al., 2001b, Hobbs and Helman, 2004), more sophisticated dynamic models (e.g., Botterud et al., 2005, Ford, 2001) and agent-based techniques (Bunn and Oliveira, 2001). For example, Hobbs et al. (2001b) calculated the equilibrium values of the ICAP price in PJM as well as probabilities of alternative energy price regimes, and base and peak load-generating capacities under each regime.33 Market modeling incorporating transmission constraints has allowed for more detailed consideration of locational market power (Day et al., 2002, Hobbs and Helman, 2004, Neuhoff et al., 2005). Laboratory experiments with live subjects, although expensive, allow for exploration of market design subtleties that models often omit. This was the argument of PJM stakeholders who argued for experimental testing of the PJM Resource Pricing Model (RPM) reform of the PJM capacity market (see Table 14.5), being dissatisfied with a dynamic modeling analysis (Hobbs et al., 2005). Finally, empirical comparisons of existing systems provide irreplaceable evidence of how designs work in practice, although the lack of experimental controls often implies that there are several possible explanations for market outcomes (e.g., Harvey and Hogan, 2002; Joskow and Kahn, 2002).

14.11. Conclusions The design of ISO markets has been an evolutionary process in the USA as in most other countries. While many design experiments have been run, and some have failed at high cost, there is increasing understanding of the relationship between design elements. However, the transition to well-functioning ISO markets is still underway. Designs remain incomplete and many significant decisions remain on the table. As stressed in this chapter and by previous authors (e.g., Stoft, 2002), the various components of spot market design for energy and AS, the specification of transmission property rights, the decision as to whether to include a capacity or resource adequacy market and the design of such a market, and the approach to market power mitigation are policy choices that must be made in concert and then finely tuned to ensure that efficient market pricing and efficient investment both result. Experience in ISO markets has lead to a fair amount of 33

Hobbs et al. (2001b) found that the equilibrium ICAP price is 64,000$/MW/yr, although other researchers would suggest lower or higher values. Each generator’s expected revenue was divided between energy and ICAP sales; ICAP accounted for 40% of the gross margin (revenue minus operating expenses) for a baseload unit and 97% for a peaking plant. Under assumptions that entry occurs until profits are zero, each plant’s gross margin precisely equaled its levelized capital cost.


523

consensus that the core design elements should include LMP of energy, a sequence of DA and RT markets, and FTRs (e.g., FERC, 2002b). The inclusion of a resource adequacy requirement and market have also become the norm in the USA, but there is still ferment over whether and how to implement locational pricing of capacity or employ a capacity demand curve in the absence of demand price responsiveness. Moreover, artificial operational boundaries remain between ISOs, creating inefficiency, although there is great interest in co-ordination along with advances in metering and communication technology, paving the way to lower barriers to competition across large regions. Also, while efficient pricing of marginal transmission usage over large regions is a significant achievement, there has been less satisfaction with the rules for FTRs, especially in establishing a relatively stable long-term hedge. And, as discussed extensively in other chapters in this volume, there is concern that transmission investment has not followed short-term spatial price signals. Finally, ISOs must demonstrate a capacity for efficiency and innovation themselves, or risk the departure of market participants.

Acknowledgment The research assistance of Javier Iñón in preparing the section on capacity markets is gratefully acknowledged. Mario DePillis, Partha Malvadkar, David Mead and Kelly Perl provided useful comments. Sections of this paper including several of the tables appeared earlier as Ross Baldick, Udi Helman, Benjamin F. Hobbs and Richard P. O’Neill, “Design of Efficient Generation Markets,” Proceedings of the IEEE, Special Issue on Electric Power Systems: Engineering and Policy, November 2005. Partial support for the participation of B.F.H. was provided by NSF Grant ECS-0224817.

References Alvarado, F. and Rajaraman, R. (2003). The 2003 Blackout: Did The System Operator Have Enough Power? August 18, Christensen Associates, Madison, WI. Ausubel, L. and Cramton, P. (1996). Demand reduction and inefficiency in multi-unit auctions. Working Paper No. 96-07, Department of Economics, University of Maryland, College Park, MD, 1996. Baldick, R. (2003). Shift Factors in ERCOT Congestion Pricing. Technical Report, Department of Electrical and Computer Engineering, The University of Texas at Austin, March, Available online at http://www.ece.utexas.edu/⬃baldick/papers/shiftfactors.pdf Barbose, G., Goldman, C. and Neenan, B. (2004). A Survey of Utility Experience with Real Time Pricing. Environmental Energy Technologies Division, Lawrence Berkeley Laboratory, December, Available at http://eetd.lbl.gov/ea/EMS/EMS_pubs.html Bartholomew, E.S., Siddiqui, A.S., Marnay, C. and Oren, S.S. (2003). The New York transmission congestion contract market: is it truly working efficiently? The Electricity Journal, 16(9), 14–24. Boisvert, R.N., Cappers, P.A. and Neenan, B. (2002). The benefits of customer participation in wholesale electricity markets. Electricity Journal, 15(3), 41–51. Borenstein, S. (2004). The long-run effects of real-time electricity pricing. Center for the Study of Energy Markets (CSEM), Working Paper No. 133, University of California, Berkeley, California, June. Bunn, D. and Oliveira, F. (2001). Agent-based simulation: an application to the new electricity trading arrangements of England and Wales. IEEE Transactions on Evolutionary Computation, 5(5), 493–503. California ISO, 2002 Annual Report on Market Issues and Performance. Department of Market Analysis, California Independent System Operator, April 2003. California ISO, 2003 Annual Report on Market Issues and Performance. Department of Market Analysis, California Independent System Operator, April 2004.

524


California ISO, 2004 Annual Report on Market Issues and Performance. Department of Market Analysis, California Independent System Operator, April 2005. California Public Utilities Commission (CPUC) (2003). Order instituting rulemaking to promote policy and program coordination and integration in electric utility resource planning. R.04-04-003. Caramanis, M. (1982). Investment decisions and long-term planning under electricity spot pricing. IEEE Transaction on Power Apparatus System, PAS-101(12), 4640–4648. Cazalet, E.G., Clark, C.E. and Keelin, T.W. (1978). Costs and Benefits of Over/Under Capacity in Electric Power System Planning. EPRI, Palo Alto, CA. Chao, Hung-po and Huntington, H.G. (eds.) (1998). Designing Competitive Electricity Markets. Kluwer Academic Press, Boston, MA. Chao, Hung-po, Peck, S., Oren, S. and Wilson, R. (2000). Flow-based transmission rights and congestion management. Electricity Journal, 13(8), 38–58. Coase, R.H. (1988). The Firm, the Market and the Law, University of Chicago Press, Chicago. Cramton, P., Chao, Hung-po and Wilson, R. (2005). Review of the Proposed Reserve Markets in New England. Prepared for ISO New England by University of Maryland, EPRI and Stanford University, January 18, (Online) Available at www.iso-ne.com Cramton, P. and Stoft, S. (2005). A capacity market that makes sense. Electricity Journal, 18(August/ September) 43–54. Day, C.-J., Hobbs, B.F. and Pang, J.-S. (2002). Oligopolistic competition in power networks: a conjectured supply function approach, IEEE Transaction on Power System, 17(3), 597–607. Dixit, K. and Baldick, R. (2003). An empirical study of the economies of scale in AC transmission line construction costs. Working Paper, Department of Electrical and Computer Engineering, The University of Texas at Austin, December. Douglas, S. (2004). Utilization Rates of Coal-Fired Power Plants in the Eastern US and the Efficiency of Electricity Market Reforms, Department of Economics, West Virginia University, Morgantown, WV, August. Elmaghraby, W., O’Neill, R.P., Rothkopf, M.H. and Stewart, Jr., W.R. (2004). Pricing and efficiency in “Lumpy” energy markets. Electricity Journal, 17(5), 54–64, June. Fabra, N., von der Fehr, N. and Harbord, D. (2004). Designing electricity auctions. Center for the Study of Energy Markets, University of California, Berkeley, CSEM WP-122, February. Federal Energy Regulatory Commission (FERC) (1996a). Promoting Wholesale Competition Through Open Access Non-Discriminatory Transmission Services by Public Utilities; Recovery of Stranded Costs by Public Utilities and Transmitting Utilities, Order No. 888, FERC Stats. and Regs. ¶ 31,036, April 24. Federal Energy Regulatory Commission (FERC) (1996b). Open Access Same-Time Information System (formerly Real-Time Information Networks) and Standards of Conduct, Order No. 889, 75 FERC ¶ 61,078, April 24. Federal Energy Regulatory Commission (FERC) (1999). Regional Transmission Organizations, Order No. 2000, 89 FERC ¶ 61,285, December 20. Federal Energy Regulatory Commission (FERC) (2000a). Order accepting for filing in part and rejecting in part proposed tariff amendment and directing re-evaluation of approach to addressing intrazonal congestion, 90 FERC ¶ 61,006, January 7. Federal Energy Regulatory Commission (FERC) (2000b). Order initiating hearing proceedings to investigate justness and reasonableness of rates of public utility sellers in California ISO and PX markets and to investigate ISO and PX tariffs, contracts, institutional structures and bylaws; and providing further guidance to California entities, 92 FERC ¶ 61,172, August 23. Federal Energy Regulatory Commission (FERC) (2001a). Order establishing prospective mitigation and monitoring plan for the California Wholesale Electric Markets and establishing an investigation of public utility rates in the wholesale Western energy markets, 95 FERC ¶ 61,115, April 26. Federal Energy Regulatory Commission (FERC) (2001b). Order on rehearing of monitoring and mitigation plan for the California wholesale electric markets, establishing west-wide mitigation, and establishing settlement conference, 95 FERC ¶ 61,418, June 19. Federal Energy Regulatory Commission (FERC) (2001c). Order establishing evidentiary hearing procedures, granting rehearing in part, and denying rehearing in part, 96 FERC ¶ 61,120, July 25. Federal Energy Regulatory Commission (FERC) (2002a). Order on the California comprehensive market redesign proposal, 100 FERC 61, 060, July.


525

Federal Energy Regulatory Commission (FERC) (2002b). Standard Market Design and Structure, Notice of Proposed Rulemaking, 18 CFR Part 35, Docket Number RM01-12-000, September. Federal Energy Regulatory Commission (FERC) (2003a). Final report on price manipulation in western markets fact-finding investigation of potential manipulation of electric and natural gas prices, Prepared by the staff of the Federal Energy Regulatory Commission, Docket No. PA02-2-000, March. Federal Energy Regulatory Commission (FERC) (2003b). Order establishing evidentiary hearing procedures, granting rehearing in part, and denying rehearing in part, 102 FERC ¶ 61,317 March 26. Federal Energy Regulatory Commission (FERC) (2003c). Investigation of terms and conditions of public utility market-based rate authorizations, Order amending market based rate tariffs and authorizations, 105 FERC ¶ 61,218, November 17. Federal Energy Regulatory Commission (FERC) (2004a). Order on Rehearing and Modifying Interim Generation Market Power Analysis and Mitigation Policy, 107 FERC ¶ 61,018, April 14. Federal Energy Regulatory Commission (FERC) (2004b). Order on Tariff Filing, 107 FERC ¶ 61,112, May 6. Federal Energy Regulatory Commission (FERC) (2004c). Order on Rehearing, 108 FERC ¶ 61,026, July 8. Federal Energy Regulatory Commission (FERC) (2004d). Staff Report on Cost Ranges for the Development and Operation of a Day One Regional Transmission Organization, Docket No. PL04-16-000, October. Federal Energy Regulatory Commission (FERC) (2005a). Principles for Efficient and Reliable Reactive Power Supply and Consumption, FERC Staff Report, AD05-1-000, February 4. Federal Energy Regulatory Commission (FERC) (2005b). Notice Inviting Comments On Establishing Long-Term Transmission Rights In Markets With Locational Pricing, Docket No. AD05-7-000, May 11. Federal Energy Regulatory Commission (FERC) (2005c). 2004 State of the Markets Report, Staff Report by the Office of Market Oversight and Investigations, June. Ford, A. (2001). Waiting for the boom: a simulation study of power plant construction in California. Energy Policy, 29, 847–869. Fu, Y., Shahidehpour, M. and Li, Z. (2005). Security-constrained unit commitment with AC constraints. IEEE Transactions on Power Systems, 20(2), 1001–1013. Green, Richard (2000). Competition in generation: the economic foundations. Proceedings of the IEEE, 88(2), 128–139. Harvey, S.M., Hogan, W.W. and Pope, S. (1997). Transmission Capacity Reservations and Transmission Congestion Contracts, John F. Kennedy School of Government, Harvard University, Cambridge, MA, June 6, 1996; revised March 8. Harvey, S.M. and Hogan, W.W. (2002). Market power and market simulation, John F. Kennedy School of Government, Harvard University, Cambridge, MA, July 16. Harvey, S.M., Hogan, W.W. and Pope, S. (2005). Comments on the California ISO MRTU LMP Market Design, LECG, LLC, Cambridge, MA, February 23, Available at http://ksghome.harvard.edu/ ⬃whogan/ Helman, U. (2006). Market power monitoring and mitigation in the US wholesale power markets, Energy, 31(6–7), 877–904. Hirst, E. and Hadley, S. (1999). Generation adequacy: who decides? Electricity Journal, 12(8), 11–21. Hobbs, B.F., Rothkopf, M.H., O’Neill, R. and Chao, Hung-po (eds.) (2001a). The Next Generation of Electric Power Unit Commitment Models, Kluwer Academic Press, Norwell, Massachusetts. Hobbs, B.F., Iñón, J. and Stoft, S.E. (2001b). Installed capacity requirements and price caps: oil on the water, or fuel on the fire? Electricity Journal, 14(6), 23–34. Hobbs, B.F. and Helman, U. (2004). Complementarity-based equilibrium modeling for electric power markets. In D. Bunn, (ed.), Modeling Prices in Competitive Electricity Markets, J. Wiley, Chichester, pp. 69–98. Hobbs, B.F., Iñón, J.G., Hu, Ming-Che and Stoft, S.E. (2005). Capacity markets: review and a dynamic assessment of demand-curve approaches. Proceedings, IEEE Power Engineering Society General Meeting, June 12–16, pp. 2792–2800. Hogan, W.W. (1993). Markets in real electric networks require reactive prices. Energy Journal, 14(3), 171–200. Hogan, W.W. (1994). An Efficient Bilateral Market Needs a Pool–Comments on the “Blue Book” Regarding Competitive Wholesale Electric Markets: Role, Structure and Efficacy, California Public Utilities

526


Commission Hearings, San Francisco, CA, August 4, Available at http://ksghome.harvard.edu/ ⬃whogan/ Hogan, W.W. (1998). Nodes and zones in electricity markets: seeking simplified congestion pricing. In H. Chao and H.G. Huntington, (eds.), Designing Competitive Electricity Markets, Kluwer, Boston, pp. 33–62. Hogan, W.W. (1999). Restructuring the Electricity Market: Institutions for Network Systems, John F. Kennedy School of Government, Harvard University, Cambridge, MA, April, Available at http://ksghome. harvard.edu/⬃whogan/ Hogan, W.W. (2005). On an “Energy Only” Electricity Market Design for Resource Adequacy, John F. Kennedy School of Government, Harvard University, Cambridge, MA, September 23, Available at http:// ksghome.harvard.edu/⬃whogan/ Hogan, W.W. and Ring, B.J. (2003). On Minimum-Uplift Pricing for Electricity Markets, John F. Kennedy School of Government, Harvard University, Cambridge, MA, March, Available at http://ksghome. harvard.edu/⬃whogan/ ISO New England, Inc., (2002). Annual Markets Report, May 2001–April 2002, ISO New England, Holyoke, MA, September 12. ISO New England, Inc., (2003a). Annual Markets Report, May–December 2002, ISO New England, Holyoke, MA, August 13. ISO New England, Inc., (2003b). A Review of Peaking Unit Safe Harbor (PUSH) Implementation and Results, Market Monitoring Department, December 3. ISO New England, Inc., (2004). Annual Markets Report 2003, ISO New England, Holyoke, MA. ISO New England, Inc., (2005). Annual Markets Report 2004, ISO New England, Holyoke, MA. Jaffe, A.B. and Felder, F.A. (1996). Should electricity markets have a capacity requirement? If so, how should it be priced? Electricity Journal, 9(10), 52–60. Joskow, P.L. (2001). California’s electricity crisis. Oxford Review of Economic Policy, 17(3), 365–388. Joskow, P.L. and Kahn, E. (2002). A quantitative analysis of pricing behavior in California’s wholesale electricity market during Summer 2000. Energy Journal, 23(4), 1–35. Joskow, P.L. and Tirole, J. (2000). Transmission rights and market power on electric power networks. Rand Journal of Economics, 31(3), 450–487. Joskow, P.L. and Tirole, J. (2005). Merchant transmission investment. Journal of Industrial Economics, 53(2), 233–264. Kahn, A.P., Cramton, P., Porter, R. and Tabors, R. (2001). Pricing in the California Power Exchange Electricity Market: Should California Switch from Uniform Pricing to Pay-as-Bid Pricing? Blue Ribbon Panel Report, Prepared for the California Power Exchange, January. Kahn, E. and Baldick, R. (1994). Reactive power is a cheap constraint. Energy Journal, 15(4), 191–202. Kamat, R. and Oren, S.S. (2002). Rational buyer meets rational seller: reserves market equilibrium under alternative auction designs. Journal of Regulatory Economics, 21(3), 247–288. Kirschen, D.-S. (2003). Demand-side view of electricity markets. IEEE Transactions on Power Systems, 18(2), May, 520–527. Landrieu, P. (Chairman, MAAC) (2004). August 14, 2003 Lessons Learned. Harvard Electric Policy Group, on the HEPG Web site, March 1. Lesieutre, B.C. and Hiskens, I.A. (2005). Convexity of the set of feasible injections and revenue adequacy in FTR markets. IEEE Transaction on Power Systems, 20(4), 1790–1798. Luh, P.B., Blankson, W.E., Chen, Y., Yan, J.H., Stern, G.A., Chang, S.-C. and Zhao, F. (2004). Optimal auction for the deregulated electricity market using augmented lagrangian and surrogate optimization. Paper presented at the IEEE Power Systems Conference and Exposition, New York, October. Markiewicz, K., Rose, N.L. and Wolfram, C. (2004). Has restructuring improved operating efficiency at US electricity generating plants? Center for the Study of Energy Markets (CSEM). Working Paper 135, University of California, Berkeley, July. Mas-Collel, A., Whinston, M.D. and Green, J.R. (1995). Microeconomic Theory. Oxford University Press, New York. Motto, A.L. and Galiana, F.D. (2002). Equilibrium of auction markets with unit commitment: the need for augmented pricing. IEEE Transaction on Power System, 17, 798–805.


527

Neuhoff, K., Barquin, J., Boots, M.G., Ehrenmann, A., Hobbs, B., Rijkers, F.A.M. and Vazquez, M. (2005). Network-constrained Cournot models of liberalized electricity markets: the devil is in the details. Energy Economics, 27(3), 495–525. O’Neill, R.P., Helman, U., Hobbs, Benjamin F., Stewart, Jr., W.R. and Rothkopf, M.H. (2002). A joint energy and transmission rights auction: proposal and properties. IEEE Transactions on Power Systems, 17(4), 1058–1067. O’Neill, R.P. and Stewart, W. (2003). Locational marginal pricing of reserves. Draft Working Paper, Washington, DC, June 28. O’Neill, R.P., Sotkiewicz, P.M., Hobbs, B.F., Rothkopf, M.H. and Stewart, Jr., W.R. (2005a). Efficient market-clearing prices in markets with nonconvexities. European Journal of Operations Research, 164, 269–285. O’Neill, R.P., Baldick, R., Helman, U., Rothkopf, M.H. and Stewart, W. (2005b). Dispatchable transmission in RTO markets. IEEE Transactions on Power Systems, 20(1), 171–179. Oren, S.S. (2005). Generation adequacy via call options obligations: Safe passage to the promised land, Energy policy and economics. Working Paper, University of California Energy Institute, October. PJM (2000). State of the Market Report, 1999, Market Monitoring Unit, PJM Interconnection, L.L.C., June, Available at www.pjm.com PJM (2001). State of the Market Report, 2000, Market Monitoring Unit, PJM Interconnection, L.L.C., June, Available at www.pjm.com PJM (2002). State of the Market Report, 2001, Market Monitoring Unit, PJM Interconnection, L.L.C., June, Available at www.pjm.com PJM (2003). State of the Market Report, 2002, Market Monitoring Unit, PJM Interconnection, L.L.C., March 5, Available at www.pjm.com PJM (2004). State of the Market Report, 2003, Market Monitoring Unit, PJM Interconnection, L.L.C., March 4, Available at www.pjm.com PJM (2005). State of the Market Report, 2004, Market Monitoring Unit, PJM Interconnection, L.L.C., March 8, Available at www.pjm.com Potomac Econocics Ltd. (2005). 2004 State of the Market Report, New York ISO, Report by Independent Advisor to the New York ISO, July, Available at www. nyiso.org. Potomac Electric Power Company, PEPCo (2004). Before the Public Service Commission of Maryland in the Matter of the Potomac Electric Power Company’s Report on the Future of its Maryland Energy Use Management Programs – Case No. 8796, Phase II, Baltimore, MD, Potomac Electric Power Company, Available at http://www.psc.state.md.us/ Rahimi, F. (2005). The Feasibility Index Method for Competitive Path Assessment, MRTU Proposal, www.caiso.com/docs/2005/08/10/200508101637549859.pdf, California ISO, Folsom, CA, August 10. Ren, Y. and Galiana, F.D. (2004a). Pay-as-bid versus marginal pricing-part I: strategic generator offers. IEEE Transactions on Power Systems, 19 (November), 1771–1776. Ren, Y. and Galiana, F.D. (2004b). Pay-as-bid versus marginal pricing-part II: market behavior under strategic generator offers. IEEE Transactions on Power Systems, 19 (November), 1777–1783. Schweppe, F.C., Caramanis, M.C., Tabors, R.D. and Bohn, R.E. (1998) Spot Pricing of Electricity. Kluwer Academic Publishers, Boston Son, Y.S., Baldick, R., Lee, K.-H. and Siddiqi, S. (2004). Short-term electricity market auction game analysis: uniform and pay-as-bid pricing. IEEE Transactions on Power Systems, 19(4), 1990–1998. Stoft, S. (1997). Transmission pricing in zones: simple or complex? Electricity Journal, 10(1), 24–31. Stoft, S. (2002). Power System Economics. IEEE Press, New York. US Department of Energy (USDOE) (2003). Report to Congress: Impacts of the Federal Energy Regulatory Commission’s Proposal for Standard Market Design, US Department of Energy, DOE/S-0138, April 30. US-Canada Power System Outage Task Force (2003). Interim Report: Causes of the August 14 Blackout in the United States and Canada, November. US-Canada Power System Outage Task Force (2004). Final Report on the August 13, 2003 Blackout in the United States and Canada: Causes and Recommendations, April.

528


Vazquez, C., Rivier, M. and Perez-Arriaga, I.J. (2002). A market approach to long-term security of supply. IEEE Transactions on Power Systems, 17(2), 349–357. Walton, S. and Tabors, R.D. (1996). Zonal transmission pricing: methodology and preliminary results from the WSCC. The Electricity Journal, 9(9), 34–41. Wilson, R. (2002). Architecture of power markets. Econometrica, 70(4), 1299–1340. Winston, C. (1993). Economic deregulation: days of reckoning for microeconomists. Journal of Economic Literature, 31(3), 1263–1289. Wolak, F.A. (2003a). Diagnosing the California electricity crisis. The Electricity Journal, 16(7), 11–37. Wolak, F.A. (2003b). Measuring unilateral market power in wholesale electricity markets: the California market. American Economic Review, 93(2), 425–430.

Chapter 15 Competitive Retail Power Markets and Default Service: The US Experience TAFF TSCHAMLER Retail Energy Practice, KEMA, Inc., Englewood, CO, USA

Chapter Summary This chapter provides a review of the US experience with competitive retail power market design with focus on default generation service (i.e. incumbent utility service). It sets forth the author’s assessment of effective default service policy, assuming an objective of vigorous competition and customer participation in retail electric markets. It also provides a summary of the various default service approaches or models adopted by states that allow customer choice. This chapter also provides historical data on the level of market activity by jurisdiction.

15.1. Introduction In the USA, the development of well functioning, competitive retail power markets continues to be a work in progress. Although the difficult transition to retail choice has slowed reforms among a number of states, 20 states and the District of Columbia offer some form of customer choice to all or part of their electric customer bases. Many of these jurisdictions have enacted substantial policy changes since the initial competitive structures were put in place. Several others are expected to enact new policies over the coming few years – transition periods are expiring, wholesale markets are evolving and pressure from all sides to change the current state of partial regulation are strong forces of change (Fig. 15.1). Consequently, a number of states will reach a cross roads of retail electric policy in the coming years as initial transition periods expire. Policy-makers must decide what form of retail generation service their state will adopt. During the transition to fully open retail markets, the primary driver of competition and customer choice has been default service policy. Default service refers to the generation supply and related customer services provided to customers that do not receive electricity from a competitive supplier. An important distinction made throughout this chapter is the existence of two types of default service customers: (1) those that have not selected competitive service (Status Quo Service) and (2) those that did choose competitive service and then voluntarily or involuntarily left competitive service (Last Resort Service). 529

530


Choice for all customers Choice for large customers only No choice

Fig. 15.1. Eligibility of electricity customer choice by state.

Although customer choice and competition levels have increased rapidly among some customer groups in some states, it is generally expected that large numbers of customers will voluntarily remain on default service for several years to come, especially small customers. Given this inertia and the lack of competitive suppliers in several open markets, the central question for those seeking to advance competition becomes what are the appropriate policies for structuring default service, particularly for Status Quo customers? This chapter specifies the principles and corresponding models of default service that will lead to robust retail competition and meaningful customer choice. It provides a review of the experience to date across US competitive retail markets. The principles guide recommendations contained in a road map for achieving robust retail competition.

15.2. The Purpose and Components of Default Service Default service is the single most important driver of competition in retail electric markets.1 Its general purpose has been to ensure generation service for customers that do not purchase or cannot purchase power from the competitive market. It is a regulatory construct used in most jurisdictions to insure a smooth transition to competitive markets, and in a few cases a long-term alternative to the competitive market. Its specific legislated purpose and resulting structure varies widely from jurisdiction to jurisdiction. There are generally two types of customers that purchase default service: (1) those do not choose a competitive service (Status Quo Service) and (2) those that did choose competitive service and then voluntarily or involuntarily left competitive service (Last Resort Service). These two types of service are described in more detail in the subsequent section. Default service is critically important to competition because it is the basis of comparison for consumers shopping for competitively priced retail generation service and in large part, 1

A wide body of theory, empirical research and opinion has discussed the importance of default service to market development. For example see Tschamler, Taff (2000); Joskow, Paul (2003); Reitzes et al. (2002) and Graves, Frank and Joseph Wharton (2003).

Competitive retail power markets and default service

531

it determines the attractiveness of the market for both competitive suppliers and customers. It is the surrogate market price determined through a regulatory or legislative process that may be set administratively or through a competitive process, such as an auction or a wholesale spot market. Default service policies have a variety of rationales that reflect the many economic and equity impacts as well as the specific stakeholder interests associated with the transition to retail electric competition. As we discuss later, certain rationales and resulting policies can be inconsistent with the objective of developing competitive retail markets. The most common rationales fall under three general categories: ●

●

●

Continuity of service: As electricity is continuously consumed and is generally viewed as an essential service, default service is intended to ensure continuity of service when a supplier fails. Stability of prices and consumer protection: Prices subject to regulation can be structured so that they are stable or fixed over some period of time. In addition, prices subject to regulation presumably protect customers from adverse market impacts through regulatory oversight. This rationale generally implies that competitive markets yield more volatile prices and more risk. Smooth transition for customers: The movement to unregulated electric service can be disruptive to customers, particularly under customer assignment policies. Default service can avoid large-scale disruptions and economic change relative to structures that move directly to full competition.

15.2.1. Dimensions of default service Default service policies encompass a variety of pricing, contractual and service components. Characterizes nine dimensions of default service and more broadly, retail market design. For each dimension a set of categories or typical approaches are provided that characterize market structure. The most common criterion for defining a default service model is the form of pricing, which comprise three of the nine dimensions: price frequency, price setting mechanism and price components (Fig. 15.2). We describe each dimension as follows: ●

●

● ●

●

●

●

Supply obligation refers to the entity (and associated contractual requirements) responsible for providing default service whether Status Quo and/or Last Resort. Utility separation (organizational models) refers to the form of separation between regulated and unregulated organizations and retail and generation organizations of a utility in a restructured market. Price frequency refers to how frequent the retail price is set for default service customers. Price setting mechanism refers to the approach taken to determine retail prices, including negotiated rate caps, competitive procurement, pass through of spot market prices, etc. Price components refer to the elements, whether presented as bundled or unbundled charges to the customer, that are included in the total default service price. Customer Ts and Cs (CTC) mechanism refers to the approaches used to: (1) determine stranded cost levels, (2) recover approved stranded costs and (3) the approach used to reconcile allowed recovery levels with actual recovery levels. Supplier Ts and Cs refers to the terms and conditions between retailers and utility delivery companies.

532

Electricity Market Reform Categories and Approaches Utility: not divested

Supply obligation

Utility: divested

Utility separation

Integrated w/ Functionally code of conduct separate

Price frequency Price setting mechanism Price components CTC mechanism

Non-utility

Legally separate

Separate ownership Annual

⬎Annual

Variable

Spot

None

Fuel adjustment

Retail service (billing, care)

Adders/credits

POLR profit

Code of conduct

Hourly

Monthly

Multi-month

Fixed

Auction

Administrative

Wholesale supply Rate design

Supplier Ts and Cs

Fees

Customer Ts and Cs

Fees

Wholesale design

Utility affiliate

RTO/ISO

Procurement Retail supply Rate level

Term

Risk

Credit

Transaction requirements

Payment protocols

Licensing

Enrollment/ termination

Usage history

Switch back Billing policies rules

Spot and Supply Transmission requirements requirements forward markets

Scheduling/ settlement

Liquidity/ hedgability

Fig. 15.2. Dimensions of default service.

●

●

Customer Ts and Cs refers to the terms and conditions between end use customers and default service providers. Wholesale design refers to the components of the wholesale market structure that have an impact on default service and retail market design.

Combinations across dimensions result in the two model categories that define the retail market structure: pricing models and organizational models. Models under each category are addressed in subsequent sections. 15.2.1.1. Last Resort Service versus Status Quo Service We define and assess specific policies associated with the above dimensions in terms of the two general purposes for which default service has been established, Status Quo Service and Last Resort Service: ●

●

Status Quo Service: Provided at regulated prices to customers that do not switch to competitive supply. It is designed to provide customers with a longer-term alternative to competitive offers. As a regulated service, it generally results in an attractive alternative to services offered in the competitive market. Last Resort Service: Provided at regulated prices to customers (1) that have sought, but cannot purchase competitive service for some reason (e.g. due to customer’s bad credit, lack of offers in the market) or (2) whose selected competitive supplier terminated service for some reason (e.g. bankruptcy, left the market, customer breach of contract, customer non-payment). As a regulated service, Last Resort Service is generally not an attractive alternative to the competitive market.

In contrast to Status Quo Service, Last Resort Service is designed so that customers have a backstop or safety net that protects them from disconnection in the event of a supplier failure or some other event. Given its purpose, Last Resort Service is less attractive to customers purchasing electricity than what is available in the competitive market. Unlike Status Quo


533

Service, Last Resort Service is generally not a tool used by policy-makers to provide customers with price protection or comparable service to the competitive market. To achieve sustainable and robust retail competition, Status Quo Service should be either eliminated or transformed into a service that is not subject to price regulation. Furthermore, default service policies, whether Last Resort or Status Quo, should be designed so that few customers purchase the service and those customers that do, do not purchase it for very long. The consequence of this principle is a set of policies that lead to decreasing sales of priceregulated electricity and increasing sales of market-priced electricity. Another outcome associated with this principle is the general migration of default customers from a Status Quo Service to a Last Resort Service. This principal can be adhered to through various market-based pricing policies and restrictions on service attributes that attract customers to default service. Where default service is structured as a Status Quo Service, one approach to transitioning toward a Last Resort Service is to permit customers to move to and from default service at their will, subject to competitive retail service contract requirements. Restrictions or penalties for moving on and off default service, such as minimum stay requirements, should not be imposed on customers. Other provisions might include prohibitions on default suppliers from providing on-line information services, green power products and any other service that is in demand from customers and that would be best provided through the competitive market. Another transitional approach to eliminating Status Quo Service is to phase in the policy changes starting with large power buyers first. In general, large buyers are more knowledgeable and engaged in competitive energy buying, and therefore represent the best candidates to impose initial policy changes upon. An alternative to fully eliminating Status Quo Service is to change it to a service that is not subject to price regulation. That is, customers that do not switch to a competitive supplier would remain with an incumbent provider, but the price charged would be set at whatever level the provider chooses. Texas has taken such an approach. This requires that prices be unregulated and regulated utilities be structurally separated from the default retail supplier. These principals are discussed later in the report. 15.3. Default Service Pricing Models Price is the most important aspect of default service in terms of its significance to competitive market development. Although there are many variations, most default service pricing approaches can be categorized into six general models: ● ● ● ● ● ●

Negotiated rate models (standard offer model). Auction or RFP price models (bidding model). Hub or forward market price models (formula model). Spot market price models. Fuel factor models. Unregulated price models.

The models are primarily defined in terms of the price setting mechanism and the price frequency for default service. In terms of developing sustainable, robust competition, each has its pros and cons, which are summarized following the model description. In identifying the pros and cons we apply the general criterion that the closer the default service pricing model approximates the competitive retail market price, the better the model is for market development.

534


The models described below do not capture all aspects of the default service rate structures, but summarize key price relationships. Each model can have several important variations. Specifically, the rate design of stranded costs and the use of rate caps and deferral mechanisms can create substantially different outcomes under the same pricing model. Across all models, rate caps, residual stranded cost mechanisms and deferral mechanisms distort retail supply prices and are barriers to market development.

15.3.1. Negotiated rate model (standard offer models) The negotiated rate or standard offer model requires the incumbent utility to provide default generation service, typically over a multi-year transition period, at fixed rates. This model was used by a group of states that were some of the first to implement customer choice, including Massachusetts, Rhode Island and Pennsylvania. Often the standard offer model has included a mandatory rate reduction for customers. Stakeholders generally negotiated the unbundled rate levels, apportioning the transmission and distribution (T&D), CTC and “shopping credit” components under settlement agreements. Some standard offer models permit the use of rate adjustments due to fuel or CTC over/under recoveries, but the rates are generally not tied to underlying wholesale power market prices.

15.3.1.1. Pros None. Negotiated models do not set prices based on any market mechanisms. Relative to the other models, it has no positive attributes with respect to market development.

15.3.1.2. Cons As negotiated models do not use market mechanisms, they do no approximate prices that the retail market would charge. The model results in distorted price signals to the market and often prevent the development of robust retail customer choices. These models also often place substantial risk on incumbent utilities and their rate payers since a long-term supply obligation that is at a below-market rate will adversely impact utility financial performance and possibly result in deferred balances that rate payers must pay for. Figure 15.3 illustrates the negotiated rate model, assuming a declining stranded cost charge and increasing shopping credit, which is similar to the Massachusetts “standard offer” model.

15.3.2. Auction or RFP model (bidding models) The auction or RFP model generally requires the incumbent utility to solicit bids for generation service from the wholesale market using either an auction or an RFP. The procurement process is overseen and approved by state utility commissioners, and in some jurisdictions, such as Maine, the Commission administers the process. Default service under these programs is generally equivalent to full requirements wholesale service and in some cases a retail adder is included in the rate as a proxy for the avoided retail service costs. To date, this model has not been adopted at the retail level. That is, generation service has been limited to wholesale supply and not functions that directly involve customers, such as billing and call centers.


535

Fig. 15.3. Negotiated rate model.

15.3.2.1. Pros Depending on the term, pricing structure and contract provisions of the bid, this approach will reflect wholesale generation market prices (at the time of the bid) more consistently than negotiated models and spot market models. Fixed retail prices more closely align with fixed price auction structures and perhaps hub-based structures. Compared to the hub approach (see above), this model sets an actual market price according to the results of a bid, rather than a synthetic calculation.

15.3.2.2. Cons To date, auction or RFP models have been limited to bids for wholesale generation service. Price levels, although they tend to track market-based retail prices, are structurally below the retail market prices. Under wholesale bids, the utility default provider usually embeds its retail costs in T&D charges. One common solution to the structural difference is to apply an adder to the default price or a credit to utility rates of migrated customers. However, the default customer is effectively paying twice for retail service with an adder. Also, auction or RFP models generally do not eliminate or transfer Status Quo Service. Integrated incumbent utilities are the most common provider. Figure 15.4 illustrates the auction model, assuming a fixed stranded cost charge and a fixed delivery charge across three auction periods.

15.3.3. Hub or forward market price model (formula models) The hub or forward market model utilizes public pricing data at a wholesale market hub(s) to calculate a default service price. Utilizing hub price data, the default prices are offered over some time period, typically 6 months or a year, and then recalculated. To reflect full requirements power service and perhaps retail service, the hub price is generally adjusted via formula to account for various other components beyond wholesale block power, such as shaping premiums, ancillary services, transmission, credit and other adders. The default price is generally not tied to the cost incurred by the default provider. If this model is used in conjunction with a rate cap, the stranded cost charge is calculated as a residual, as illustrated in Figure 15.4. The hub model could also be used without a rate cap, by setting the CTC at a level unrelated to the default price, or if there is not a CTC and rate cap, as illustrated in Figure 15.5. The only US market that currently utilizes hub pricing model is Illinois

536


Fig. 15.4. Auction or RFP model without rate cap.

Fig. 15.5. Hub or forward market model with rate cap.

through its power purchase option among the state’s three largest utilities. Cinergy has proposed a hub-based model as its post transition default service (Fig. 15.6).2 15.3.3.1. Pros Like the auction model, the hub model is designed to set fixed prices based on forward markets and therefore it approximates a fixed price generation service. Depending on the specific factors included, the formula and hub used, the default price may approximate a retail market price (at the time of price determination). 15.3.3.2. Cons The hub model is only as good as the formula and the data it uses. Hub models are artificial in the sense that the price is a formula, rather than reflecting actual market-based transactions.

2

See CG&E application to modify non-residential rates, January 10, 2003, Docket 03-0093-EL-ATA.


537

Fig. 15.6. Hub or forward market model without rate cap.

Fig. 15.7. Spot market price model without rate cap.

15.3.4. Spot market model Spot price models utilize an hourly day ahead or real time wholesale energy market price, typically administered through an independent system operator (ISO), to set retail default service rates. The model requires a functioning electricity spot market and typically includes additional rate components, such as capacity, ancillary services and perhaps a retail adder, as in New York and New Jersey. As in California’s original retail market design, the model may also include a rate cap that is implemented through the use of a residual CTC mechanism. If the spot price combined with the other rate components exceeds the total capped rate, a deferral mechanism is used. In Figures 15.7 and 15.8 two spot price models are illustrated one without a rate cap and deferral mechanism and one with the mechanism. 15.3.4.1. Pros Without rate caps in place, price volatility will encourage customers to sign with competitive suppliers and attract new entrants to the market to offer less volatile-priced products.

538


Fig. 15.8. Spot market price model with rate cap.

15.3.4.2. Cons Spot market prices do not reflect long-run marginal costs for generators, nor do they represent the type of price structure that most electricity consumers would purchase. Therefore, spot market prices do not approximate market-based retail prices as well as the hub, RFP and unregulated models. 15.3.5. Fuel factor model The fuel factor model represents a hybrid of a cost of service-based model and a marketbased pricing model. The leading example of this model is the Texas Price to Beat (PTB) service, the price of which can be adjusted based on a simple formula tied to natural gas futures. Massachusetts also had a type of fuel factor adjustment as part of the standard offer, but it is not as mechanized as the Texas approach. The standard offer base rate in Massachusetts was a negotiated price. In general, a fixed base rate is established, either on a bundled basis or for generation service over some time period. The base rate is based on historical cost of service rates or negotiated unbundled rates. A fuel charge is then applied that accounts for changes in the prices of underlying generation fuels, primarily natural gas. The actual fuel costs incurred by the generating entity is generally not the basis for changes in the fuel factor. Rather, the changes are based on market-based fuel prices. Figure 15.9 illustrates the fuel factor model. 15.3.5.1. Pros The fuel factor model allows for price variation with one of the most important drivers of retail market price levels: fuel costs. Depending on the structure, it can be an effective option for separating the cost of power procured by the default provider and the price default customers pay. This is a principal we refer to a decoupling and is discussed later. 15.3.5.2. Cons The model does not account for variations in other factors that impact retail market prices. Also, if the structure is based on historical prices or fuel costs, it will not approximate retail market prices, since market prices are based on current and future fuel prices, rather than historical levels.


539

Fig. 15.9. Fuel factor model.

15.3.6. Unregulated price model The unregulated price model does not have any specific mechanism that sets prices and is not subject to regulatory control of prices in either level or term. In the US, only Texas has adopted such a model for its over 1 MW customers. In April 2002, Great Britain lifted price controls for all customer classes. In those markets, prices are set according to the competitive dynamics of the market and generally follow the wholesale market forward curve. The Status Quo–Last Resort Service distinction is important under unregulated models. Although not a requirement, unregulated price models have typically applied only to Status Quo Service – customers that do not choose a provider are subject to whatever price the default provider wishes to charge. To ensure continuity of service, typically some form of Last Resort Service is provided in addition to unregulated service for customers that are subject to supplier default, as is the case in Texas with the Provider of Last Resort Service. Figure 15.10 illustrates the general relationship between wholesale spot prices, wholesale forward prices and full requirements retail prices. 15.3.6.1. Pros Unregulated prices are not approximations of retail market prices, rather they are, by definition, retail market prices. Competition against regulated pricing is eliminated. 15.3.6.2. Cons From the perspective of competitive market development and maximizing customer choice, this model is the best approach. Therefore, there are no cons, relative to the other models. 15.3.7. Principles for designing pricing models What is the most appropriate pricing model to achieve sustained, robust retail competition? Although specific designs can vary to achieve the stated objective, the following principles should be adopted as default service evolves.

540


Fig. 15.10. Unregulated price model.

Default service prices, to the degree they are price regulated, should reflect retail market prices as accurately as possible. Rate caps, residual stranded cost mechanisms and deferral mechanisms distort retail supply prices and should not be utilized in retail market designs. Whether fixed or variable, stranded cost recovery charges, if appropriate, should be a component of T&D service rates and applied independently from generation rates. CTC charges should be applied to all customers, rather than only those that switch to a competitive supplier. A switching penalty or restriction is not equitable, nor does it encourage competition. As a rate design principle, above-market generation investments and regulatory assets intended for all customers should be recovered through CTC mechanisms that apply to all customers. 15.3.8. The price–cost relationship An important attribute not captured in the pricing models above is the price–cost relationship. That is, do retail default prices reflect the actual/prospective costs incurred by the default supplier or are the default prices unrelated to the underlying costs incurred? The price–cost impact on market economics varies depending on specific retail pricing approaches adopted. However, in general a default supplier that has a pricing mechanism unrelated to its procurement costs has a significantly different incentive structure from a default supplier that passes or largely passes through its generation supply costs. The profit function is different as are the supply risks associated with the default load obligation. Some jurisdictions have implemented or proposed price mechanisms that are disconnected or decoupled from the cost to the default service provider. In Texas, for example, the PTB fuel factor permits utility affiliates to charge a price based in part on the market price of natural gas.3 Although the base generation rate is tied to historical cost of service levels, the fuel factor permits total price levels to be unrelated to the actual cost of service. Furthermore, the revenues generated from the PTB are not reconciled with costs incurred by the affiliates. Consequently, some affiliates, in their first year of competition, have been able to substantially increase earnings by effectively hedging supply obligations and increasing prices as wholesale natural gas prices increased. However, one affiliate, First Choice Power, did not effectively manage its supply obligation and incurred a significant loss of earnings during late 2002 and early 2003.4 The decoupling of costs with prices, in this case, resulted in default 3

Substantive rules, Chapter 25, § 25.41, Public Utility Commission of Texas, adopted April 10, 2001. see “TNP Enterprises Reports First Quarter Loss,” TNP Enterprises Press Release, May 15, 2003.

4


541

supply customers avoiding payment of First Choice Power’s losses. So, the Texas PTB model requires the affiliate default suppliers and their shareholders to bear most of the supply risks associated with the default load obligation. In contrast to the Texas PTB approach, auction or RFP models require the default supplier to pass through the procurement costs, with perhaps a profit margin included. As examples, a non-profit pass through model is used in Maine and a for-profit pass through model is currently used by Duquense Light (POLR III) in Pennsylvania. For these models the average costs of procured power is closely tied to the retail rate charged. These pass through mechanisms insulate the utilities from the various risks created by the default load obligation. The winning wholesale providers, their shareholders and to a certain degree end use customers (through price changes), are subject to the various risks associated with the load obligation in this case. The utility is insulated from the risks of the default service obligation. As the utility is obligated to provide generation service at fixed rates, the negotiated rate model allocates some level of risk to the utility and its shareholders. In Massachusetts and Pennsylvania, for example, the utilities agreed to serve default load for 4–10 years at set shopping credits. In Massachusetts, there is a fuel factor mechanism and a deferral mechanism that relieves some supply risk for the utilities that largely divested of their generation assets. In Pennsylvania, supply risks are born in large part by the utilities and their shareholders. The failure to manage those risks effectively led to substantial losses by GPU energy and ultimately its acquisition by First Energy. In contrast, PPL Corporation’s successful management of the supply risks has led to lucrative returns for its generation affiliate.

15.3.9. Principles for addressing the price–cost relationship Given the various price–cost relationships possible under default service, which approach is most likely to lead to meaningful customer choice and sustainable, robust competition? As mentioned, ultimately, Status Quo default service should be considered a competitive service, not subject to price regulation, or it should be eliminated. However, this may take some time to fully implement. As an interim step to fully unregulated prices, policy-makers should consider decoupling the costs of supply from the regulated price of supply. This can be conducted through a variety of approaches, including wholesale procurement mechanisms, fuel factor mechanisms and various rate cap mechanisms. Whatever the specific mechanism, the approach should seek to allocate the risks (and rewards) of the default provider’s generation obligation to shareholders and away from customers, thereby moving toward a more market-based structure, rather than a heavily regulated structure. Texas has adopted such an approach. To reiterate, policies that decouple provider costs from regulated default rates should be viewed only as an interim step toward fully unregulated prices. The end state market design will not have price controls for Status Quo Service. To the degree a Last Resort Service is required to ensure continuity of service and protect customers from retailer failure, a pass through price set at frequent intervals (e.g. hourly, monthly) would be appropriate, which would not necessarily decouple price from cost.

15.3.10. Compensation for default service providers Another important dimension not captured by the pricing models and closely tied to the price–cost relationship is default provider compensation. That is, should default providers be compensated for their default service obligation? If so, how and at what level?

542


In practice, default suppliers are subject to a wide array of compensation mechanisms and levels, depending on the market structure. In general, compensation is subject to either (1) traditional rate of return regulation under utility-provided default service models (2) default service compensation mechanisms that are separate and independent of regulated delivery company returns or (3) no compensation (e.g. certain pass through pricing approaches, such a spot market model and an auction model). Often in cases where default service providers receive an authorized return, the price is based on a market price (e.g. bid price, hub price, etc.) that is then marked up at some level equal to the authorized return. For default service providers that utilize their own generation supply and price according to a market price, costs are typically disconnected from price. In these cases, returns are often not explicitly authorized or isolated for the supply obligation (e.g. the integrated utility model). The returns are bundled into the overall regulated return. For default service providers that have divested generation assets, procurement costs are frequently tied to the price, although not always. Default suppliers, if they are chosen through a competitive process, should be compensated for their obligation. If default service prices are subject to regulation, compensation should reflect only the embedded costs of the obligation. If default service is not subject to regulation, compensation should reflect the risk and rewards the competitive market dictates.

15.3.11. Non-price regulation As summarized in the discussion of default service dimensions, retail market design includes several other components beyond default service pricing models. Non-price regulation comprises the various retail-related rules, regulations and business process requirements that are authorized by a regulatory body. These include retailer licensing and certification requirements, customer protection provisions, credit and transaction requirements, data exchange, billing and metering policies, customer enrollment requirements, payment protocols, bad debt and disconnection policies and minimum stay requirements. Non-price regulations establish the “rules of the game” and are critical to realizing robust competition and sustained customer benefits. In general, the broad set of retail rules and regulations should be specified with competition as one of their primary aims. Consequently, the non-price regulation should not unduly hinder competition and ensure a level playing field among new entrants and incumbent providers. In conjunction with this principle, the following positions on non-price rules related to default service are consistent with developing well-functioning competitive retail markets: ●

●

●

●

Where applicable, the default service provider and T&D provider should be subject to non-price regulations that are comparable to those faced by competitive providers. State regulators should adopt market monitoring functions that prevent abuse by market participants, including default providers, and work to improve rules in ways that promote efficient markets. State regulators should adopt customer protection rules that prevent abusive business practices, instill confidence in markets among electricity buyers and provide fair and reasonable business rules to facilitate transactions. Non-price regulations should seek to advance technologies that will enable more efficient, cost effective transactions among market participants.


543

15.4. Default Service Organizational Models In competitive markets, the organizational structure of utilities varies widely from jurisdiction to jurisdiction and can have a substantial impact on levels of competition. Central to restructuring is the issue of where the “line is drawn” among regulated and unregulated utility functions, and the corresponding ownership and cost recovery implications of the line drawing exercise. As policy-makers can and have drawn the restructured lines differently, we have varying organizational models. Furthermore, these organizational models have important implications for the supply obligation, rate design and pricing mechanisms and market power. The organizational functions can be broadly categorized as transmission (T), distribution (D), generation (G) and retail (R).5 In terms of rate determination in competitive markets, T and D are regulated, G is unregulated and R, although primarily regulated, reflects a contentious and often unclear set of policies with respect to ownership, cost recovery and where the line is drawn. As of the historical integrated nature of electric service, retail services have been provided as part of combined T, D and G. Consequently, in competitive markets the de-integration of R results in multiple R categories: those associated with T&D services, those associated with G services and potentially those that are either unrelated or span both T&D and G. To illustrate the variation across jurisdictions, we summarize the “organizational models” of utilities in restructured markets into five classes: ● ● ● ● ●

Integrated Separated G Separated G&R Wholesale bid Retail bid

The organizational models are defined in terms of (1) where the line is drawn on generation and distribution ownership and (2) in terms of the supply obligation. Specifically, the integrated class of firms include companies that maintain bundled electric service and operate generation assets under a regulated corporate entity or are subject to tight regulation of a generation affiliate (e.g. utility and generation affiliate contract subject to regulatory oversight). In contrast, the separated G and separated G&R are comprised of firms that have either divested generation assets, split into two separately owned companies (e.g. regulated wires company and unregulated supply company) or there is loose regulation over a generation affiliate that is supplying the utility default load. The separated firms are distinguished from the wholesale and retail bid models by the supply obligation. That is, the utility or its affiliate maintains the supply obligation under separated models while the bid models assign the supply obligation to the winning bidders (which could be the local utility affiliate). Figure 15.11 illustrates the five models and the jurisdictions that fall under each. 15.4.1. Principles for designing organizational models Determining what entity has the default supply obligation is a key component of the organizational model. The obligation can be split into the retail service obligation, which includes 5

Retail encompasses the many customer-specific processes of providing electric service, including customer contact center services, key accounts, economic development programs, billing, credit and collections, communications, low-income programs and metering. In addition to these traditional retail services, restructuring has created a new set of customer-specific processes such as EDI/XML transactions among market participants, load profiling, scheduling and settlement.

544


Separated G

Integrated T&D

G

Rr

T&D

Separated G&R T&D

R

T&D

G

Ohio Michigan Virginia

Customer

Illinois Pennsylvania New York

Generation supply (electrons) Customer contact (bills, telephone, e-mail)

Retail bid

R

T&D

G

G

Customer

Wholesale bid

G

R

Customer

Customer

Texas

Maine Massachusetts New Jersey

R

Customer

PECO (MST) JCP&L (green pilot)

T&D Regulated delivery service G

Generation service

R

Retail service (customer care, billing, etc.)

Fig. 15.11. Organizational models of restructured utilities.

responding to calls, rendering bills, administering collections, etc. and the generation service obligation, which includes the procurement and management of generation supply to meet the full requirements load obligation of individual or aggregated default service customers. As illustrated in the models above, most models assign the utility the retail service obligation, while the generation obligation varies from independent merchant companies, integrated utility generators and utility affiliate generators. Any number of structures for designating the default supply obligation could lead to meaningful customer choice robust retail competition, from models that bid out default supply, to assigning the obligation to an incumbent utility business unit, to assigning customers to competitive suppliers in some form. Whatever entity is designated the default provider, the critical requirements are that the price of Status Quo Service is unregulated (or eliminated) and the delivery business, including the retail service function is completely separated from the default generation function, including the retail service associated with generation supply. In addition, to default supply obligation, the models also illustrate various organizational structures. The diagrams simplify the potentially complex accounting, legal and personnel dimensions of a restructured utility, but capture the key organizational components. In general, most models require the separation of the regulated delivery business from the unregulated generation business. Few jurisdictions fully separate the retail organization from the delivery business;6 even fewer bid out the retail business function in the form of a competitive retail default service obligation.7 To ensure competition and resulting consumer benefits, existing organizational models must move to a de-integrated set of utility functions starting with fully separated T&D operations and tariffs. For example, the use of a single bundled tariff that designates customer

6

Reliant Energy and CenterPoint Energy voluntarily separated ownership and Rochester gas and electric partially exited retail functions through its single retailer model. 7 Only in PECO’s and JCP&L’s markets has the retail load and service obligation been assigned to other companies and that was on a partial and pilot basis, respectively.


545

choice as a rider or allocates participating customers into what is effectively a separate rate class must be eliminated in favor of separate delivery and default service tariffs. Furthermore, all assets, external resources, personnel and systems used to jointly provide delivery and default services should be fully separated into a regulated delivery services organization and a default service organization. The two organizations should be managed independently of one another and maintain fully separated accounting records. More specifically, the following principles represent a minimum standard for restructuring incumbent utilities and could be considered short-term design principles for moving to a fully competitive market: ●

●

●

Utilities with default service obligations should administer a fully separated T&D service tariff(s) separately from the default service tariff. Shared services of regulated and unregulated utility functions should be limited to corporate functions, such as accounting, executive offices and perhaps human resources. Utilities with default service obligations should strictly adhere to a code of conduct which prohibits any cross-subsidization, market obstruction and ensures adequate disclosure of cross-unit interactions.

The treatment of regulated retail services is a threshold issue for competitive retail market development. Currently, almost all utilities in the USA that act as a default supplier provide retail services to both default customers and to delivery-only customers (i.e. customer that switch).8 In this environment, the costs associated with call centers, billing, customer choice implementation and all other aspects of customer transactions and interactions are recovered under regulated rates. Where both the regulated delivery organization and the organization providing generation service (e.g. default supplier and competitive supplier) have retail service responsibilities – referred to as dual retailer approaches – the delivery company or integrated utility rates recover in part or in full the retailing costs of the default supply, but not the retailing costs of the competitive company. For the dual retailer model, a number of jurisdictions have attempted to address the issue of comparable treatment of retailing costs through the use of retail adders or credits that are generally set based on negotiation.9 These credits are generally arbitrary and not based on embedded costs (i.e. cost of service study) nor on market-driven price components (i.e. estimates of the retail service component in a competitively offered price). The advantage of a negotiated credit is that the administrative burden is significantly less than that of a cost of service or retail price component analysis. In New York regulators are conducting an embedded cost of service unbundling proceeding10 for retail and other services deemed competitive (e.g. power procurement). The outcome of the proceeding will be a set of credits that are backed out of utility default rates when customers switch to a supplier. However, the proceeding has been highly contentious, 8

In Texas, competitive suppliers provide almost all retail services. All customer contact must go through retail suppliers and not regulated delivery companies, although outage notification can go directly to the delivery company if the retailers choose to do so. Rochester gas and electric is the only other default supplier that has required retailers to be responsible for all retail services for customers that switch, but its design is being phased out in favor of a dual retailer model. 9 New York, Maryland and New Jersey all currently or will have in the near future some form of retail credit that approximates the cost of retail services for default service and removes these default supplier charges when customers switch. 10 New York Public Service Commission docket number Case 00-M-0504.

546


costly and is likely to take several more years to conclude. For a number of reasons, this is a difficult approach. In contrast to dual retailer models that use credits or adders, the Texas approach is a single retailer model that separates wires company retail services and their cost recovery from the retail services of default service and competitive providers. That is, the affiliate retail electric providers (REPs) are legally separated from the wires company, have separate rates from the wires company and are responsible for almost all retail services. As a result, most all of the costs of retail services are borne by retail suppliers, including the affiliate default suppliers. Retail service cost recovery is determined through market forces for affiliate REPs, the same as for competitive retailers. Those retail costs that remain with the delivery companies (e.g. metering, outage communication, etc.) are recovered through T&D service rates that are applied equally to customers of default providers and competitive retailers. Therefore, the complex and contentious rate design issues associated with unbundling the retail component from an integrated utility are more easily solved by separating the utility into a wires company and a default retail company and establishing separate and independent rate structures. To achieve robust and fair competition, incumbent utilities should ultimately be subject to the following principles of organizational restructuring. Given the potential impact on market structure, these represent design principles that may need to be implemented over a relatively long period of time compared to some of the other principles set forth: ●

●

●

Retail functions should be deemed competitive and therefore default service should be a retail service, rather than a competitive wholesale or integrated regulated utility service. Retail functions for regulated T&D service should be fully separated and operate independently from any retail service provided by a utility default service unit, including staff, systems, processes, costs and cost recovery, etc. Depending on the retail service obligations of the T&D utility, retail services may be duplicated across the delivery company and the default service company.

Requiring default service suppliers to provide the full set of generation-related retail services separately from T&D retail services will fundamentally change the competitive landscape and the cost structures of incumbent utilities. Requiring default service to effectively compete on retail functions as well as generation functions is the most rational means to achieve retail competition. By contrast, organizational models that define default service as a wholesale generation supply service and utilizes the regulated retail services of the utility is effectively wholesale competition. We realize this full separation of regulated retail from unregulated retail could lead to some duplication of costs and possibly claims from utilities of stranded retail assets, depending on the specific set of retail services and cost recovery mechanisms provided by the delivery company. However, if the stated policy objective is to achieve robust retail competition and the various customer benefits it creates, the path to retail competition must include the separate and subsequent independent management of regulated and unregulated retail service. 15.5. Experience to Date The wide variety of default service policies used in the US has led to varied outcomes in the development of customer choice and competition in retail markets. We summarize the experience to date by the following dimensions: price impacts, innovation and customer choice and allocation of risk.


547

In reviewing the experience to date, it is important to note that price controls and extensive regulation exist in all restructured US markets, with the exception of the Texas large customer market. Consequently, any presentation of the benefits and costs of restructuring must be couched in terms of partial restructuring, rather than full restructuring. Also, this summary does not present an analysis of the full societal costs and benefits of electric customer choice, nor, to our knowledge, has such a study been conducted. Consequently, in summarizing the experience to data, we present anecdotal information to characterize market development and to support our assertion that customer choice and competition are closely correlated with default service policies and that certain policies in certain markets have resulted in customer benefits. Default service policies correlate highly with market development. Several jurisdictions in the US have established default service polices for the largest customer classes that approximate retail market prices, such as Texas, New York, New Jersey and Illinois. The result has been a rapid increase in customer migration levels – roughly 69,000 MW out of a total 275,000 MW eligible to switch. The aggregate amount of load switched is about 69 GW through mid-2005, up from 63 GW at the end of 2004 and 57 GW in 2003. Since the beginning of the year the primary growth markets have been Texas, New York, Maryland, DC and Massachusetts, where roughly 7000 MW have collectively migrated. Declines in migration since the beginning of the year have occurred in four states: IL, CA, MI and NJ. As expected, a large majority of customer switching has occurred in business or nonresidential markets. Roughly 59,000 MW of non-residential peak load migrated by mid2005 out of a total of the 175,000 MW eligible and roughly 375,000 total. This compares to roughly 35,000 MW in mid-2002 and 16,000 in mid-2001. By contrast, of the 1250,000 MWs of estimated residential peak load eligible to choose, roughly 10,000 MW is being served by competitive suppliers. In general, the default prices being charged by the residential default service provider are lower than those that would be charged in a competitive retail market. Aggregate switched levels across combined residential and non-residential markets are displayed in Figure 15.12. The net level of competitive load migrated in 2005 has already exceeded the net level for full year 2004. As the figure illustrates, the past 3 years have demonstrated persistent, but declining growth in competitive retail markets. Figure 15.13 illustrates switching over time in volumes (terawatthours), rather than by peak load (MW). Volumes generally provide a better gauge for competitors of the size of the market since they are more closely correlated with revenue compared to peak load. The figure illustrates both the switched levels over time and the amount of volumes by state. As the figure shows, there is a high level of concentration of switched volumes in Texas, Ohio, New York and Illinois. These top four states comprise approximately 70% of the total migrated volume. Texas alone accounts for 38% of migrated volumes or roughly 138 TWh. In terms of accounts switched, the total number has seen steady, but modest increases over the past 3 years. The total number of customers switched is approximately 3.7 million, with residential customers accounting for 2.9 million. Table 15.1 on the following page provides state level numbers by year of accounts served by competitive or non-incumbent suppliers. Table 15.2 provides preliminary estimates of migrated peak load. Texas is the most robust and mature competitive retail power market in the USA by most measures. Please refer to the chapter on the Texas market for more detail. In Texas the market has rapidly evolved since fully opening in January 2002. The primary driver of the early

11 Adapted from “The 2003 Restructuring Report Card,” KEMA’s retail energy markets service, January 2004.

548

Electricity Market Reform 80,000 70,000 6.3 MW

MW migrated

60,000

6.2 MW 11.9 MW

50,000 40,000 22.3 MW 30,000 20,000 22.2 MW

10,000 0

2001 & prior

2002

2003

2004

2005 YTD

Fig. 15.12. Peak load served by competitive suppliers, GW (not MW as indicated).

TX ERCOT OH PA 2001 2002 2003 2004 2005 2005 default

NY IL MI VA NJ MD MA Other 0

50

100

150

200

250

TWh Fig. 15.13. Energy volumes migrated by state relative to default service. Source: KEMA and state governments.

Table 15.1. Numbers of customers served by alternative suppliers. 2000 YE

2003 YE

Non-residential

TX NY OH PA MA IL CA MI MD DC NJ ME RI CT MT OR VA DE AZ NH NV

103,455 65,410 92,288 26,655 30,869 28,278 22,621 5761 13,045 5257 638 2739 2079 524 2157 0 22 6 0 0 0

322,921 319,707 756,387 251,774 56,894 0 49,670 0 61,825 23,604 1836 5974 53 19,495 71 0 2301 0 0 0 0

171,470 78,858 114,385 75,195 26,848 21,256 23,630 13,168 12,309 4389 7314 3389 2053 583 252 0 22 6 0 0 0

Total

401,804

1,872,512

555,127

Residential

Non-residential

Residential

2005 YTD Residential

Non-residential

Residential

Non-residential

706,966 276,314 830,587 416,939 58,233 0 41,005 0 58,023 21,006 113,263 5225 182 24,490 2 0 2301 0 0 0 0

239,848 121,328 122,407 66,122 25,882 22,926 23,593 18,714 13,510 2747 8430 3199 2634 425 367 7 22 9 0 0 0

1,031,143 303,799 915,626 180,273 59,445 0 35,299 0 42,676 13,224 1544 5579 170 20,444 0 0 2301 0 0 0 0

287,591 149,448 124,238 57,361 46,861 23,718 22,538 18,442 15,397 5652 4491 3148 2654 566 377 29 22 5 0 0 0

1,301,169 319,361 884,608 153,918 159,689 0 30,075 0 29,775 5958 1447 2324 172 26,078 0 0 2301 0 0 0 0

2,554,536

672,170

2,611,523

762,538

2,916,875


State

2004 YE

549

550

Table 15.2. Peak load served by alternative suppliers. 2000 YE State

Non-residential

2003 YE

2004 YE

2005 YTD

Residential

Non-residential

Residential

Non-residential

Residential

Non-residential

Residential

12,971 4194 6142 2826 3882 1994 1801 1611 243 1925 1041 550 336 154 0 147 12 6 0 0 0

1615 574 0 1385 87 249 88 0 5 572 65 13 0 1 0 0 50 12 0 0 0

15,897 5774 6033 3291 4061 1800 2021 2798 2881 2163 911 679 316 151 0 137 15 6 0 0 0

3535 531 0 1595 79 240 95 0 357 964 53 10 0 1 0 0 54 12 0 0 0

18,448 6872 6497 3398 4219 2672 2039 3304 3070 2266 176 728 301 155 86 96 10 6 0 0 0

5156 524 0 1793 70 187 96 0 4 404 39 8 0 1 0 0 48 12 0 0 0

20,217 7866 6364 3575 3978 3605 2695 2853 2811 2333 1384 733 299 167 155 94 15 6 0 0 0

6506 685 0 1769 59 120 262 0 4 344 18 6 0 1 0 0 61 12 0 0 0

Total

39,835

4716

48,934

7526

54,343

8342

59,150

9847

Source: KEMA estimates based on state agency data; energy usage (MWh) from state agencies was converted to MW demand assuming a load factors appropriate for each customer class reported. Most recent data vary from October 2003 to January 2004. Mid-2001, -2002 and -2003 estimates use state agency data equivalent periods.


TX NY IL OH CA MD MA MI NJ PA DC ME MT RI OR DE CT VA NV NH AZ


551

progress in the Texas market is the approach taken to default service, which has three types: PTB (Status Quo Service for under 1 MW customers), Provider of Last Resort and unregulated default customers (customers over 1 MW). For PTB customers, a fuel factor mechanism enables providers to approximate the retail market price. For over 1 MW customers, there are no price controls. Consequently, the Texas market has over 60 active providers, with over 15 new entrants in 2003 alone. Roughly 10 providers have offered services to residential customers in Texas since the market opened. As a result, large numbers of customers have actively participated in the Texas market: over 90% of large business customers, 40% of small business customers and 14% of residential customers have selected a new provider at least once.12 Both an active competitive market and a mandated 6% rate cut has led to substantial savings for customers compared to what rates would have been under traditional regulation. In contrast to markets such as Texas, a number of US markets have adopted default service policies that do not approximate competitive retail market prices. For business customers, states such as Virginia and Arizona have devised rate unbundling policies that result in prices that make it difficult or impossible for new entrants to compete. In most residential and small business markets in the USA, new entrants cannot compete against default service. 15.5.1. Price impacts A primary benefit expected from electric deregulation is price reductions. As we operate in a state of partial regulation and in many cases rate decreases were mandated as part of restructuring, estimating the price reductions (or increases) attributable to deregulation is problematic. That is, price impacts under a partially deregulated US markets would not necessarily reflect price impacts in a “real” market. In addition, it is difficult to isolate retail price impacts of “deregulation” from changes in input costs (e.g. fuel prices), technology change, cost of capital, politics (e.g. settlements, incentives, etc.), etc. In the absence of any definitive analysis, we summarize anecdotal evidence of the impacts of partial deregulation on prices. Stepping back from electric markets, price reductions for customers and efficiency improvements in other network industries have clearly been achieved as a result of restructuring or deregulation. “Experience in a variety of other deregulated industries shows that competition and deregulation tend to produce price reductions of between 10% and 25%, along with service quality improvement whose value to consumers sometimes exceeds the value of price reductions.”14 In deregulated electric markets where default service policies have led to competition, customers have experienced savings over historical electric rates, particularly among large power users. Below is a summary of various studies and anecdotal data that summarizes price impacts of retail electric restructuring. 15.5.1.1. United Kingdom Generally considered the most advanced competitive retail power market, the UK, more precisely England and Wales, has evolved into an intensely competitive market across all customer

12

Public Utility Commission of Texas and Electric Reliability Commission of Texas. ”Scope of Competition in Electric Markets in Texas,” Public Utility Commission of Texas, January 2003. 14 Dr. Jerry Ellig, Comments to Federal Trade Commission’s Notice Requesting Comments on Retail Electricity Competition Plans Mercatus Center at George Mason University. 13

552


groups. The English market has been restructured several times since initially undergoing privatization in 1989. The key elements of the current English market design are: ●

●

● ●

●

Market was initially opened to customers over 1 MW as of April 1990; market was fully open to choice as of May 1999. Twelve incumbent monopoly distributors and retail suppliers inherited customers, but were required to strictly separate delivery and supply functions. Transmission system was separated and transferred to National Grid. Generation assets were separated into three competing businesses: National Power, PowerGen and British Energy. Initially retail prices were capped, but all price controls on incumbents were lifted on retail energy supply in April 2002.

The results of the English approach to electric competition have generally been favorable to customers. As mentioned, switching levels are high. In fact, customer participation levels in the UK are higher than in any country in the world. Savings for customers have been substantial. Through the first 4 years of competition, the average annual residential electricity bill has fallen from £268 to £238 (11%). Furthermore, “research suggests that consumers who switch suppliers are able to make significant savings, with discounts of up to 23% offered to direct debit customers, up to 19% offered to credit customers and up to 12% offered to prepayment meter customers.”15 In a January 2003 interview with the chief UK regulator Callum McCarthy, he states, “Since reforms of the electricity market began 4 years ago, wholesale prices have dropped 40%. Ofgem says this has led to savings of £1.5 billion for industrial and commercial customers and £0.5 billion for domestic customers. Falling wholesale prices have delivered direct savings to users of 34% since 1998.”16 Referring to the lifting of price controls in February 2002, Ofgem concluded, “The evidence … suggests that, over the ensuing year, competition has become an even more powerful influence on the behavior of companies in the market, and is effective in creating a range of consumer benefits. Ofgem’s view remains that competition is sufficiently advanced that price controls would be more harmful than helpful.”17 More recently the UK regulator stated “Ofgem continues to believe that domestic customers’ interests are best protected by a competitive supply market, where customers’ ability to switch places competitive pressures on suppliers in terms of prices and standards of service.”18 Savings attributable to restructuring have been estimated for a number of jurisdictions in the USA as well, although some estimates include mandated rate reductions that accompanied customer choice and even where mandated reductions are absent, default prices remain regulated. 15.5.1.2. Texas According to the Public Utility Commission of Texas (PUCT), “The Commission’s estimates show that retail customers have saved, at minimum, over $1.5 billion in electricity costs during the first year of competition as compared to the regulated rates in effect during 2001. 15

”Making markets work for consumers – the regulation of gas and electricity sales and marketing: a review of standard license condition,” Ofgem, Report 87/03, August 19, 2003. 16 Get used to it. Interview: Callum McCarthy. Petroleum Economist, January 2003, p. 27. 17 Domestic Gas and Electricity Supply Competition, Office of Gas and Electricity Markets, June 2003, p. 4. 18 Domestic Competitive Market Review 2004, Office of Gas and Electricity Markets, April 2004.


553

Table 15.3. Anecdotal savings levels for customers compared to PTB rates in Texas. Aggregation group Whataburger’s 352 restaurants Texas Medical Center Group Texas BOMA Public Power Pool Group of 107 local governments Group of 145 commercial customers Group of 21 commercial customers Cities Aggregation Group of seven multifamily properties

Average annual savings (%) 26* 24* 24* 24 16.5 15.6 14 13.7 6

Winning retailer Reliant Energy Solutions Constellation New Energy Constellation New Energy NewEnergy, TXU ES and Reliant ES TXU Energy n/a First Choice Power First Choice Power Constellation New Energy

*Savings calculated over 2001 rates, not PTB rates. Source: Aggregator annual reports, as filed with Public Utility Commission of Texas, Project Number 6280.

Commercial and industrial customers … paid approximately $645 million less in 2002, compared to 2001 bills.”19 Table 15.3 above presents a selection of actual savings levels of customers reported to the PUCT in its annual aggregator filing. These savings levels are calculated at the account level and rolled up into average savings across accounts for the particular contract. 15.5.1.3. PJM A 2003 study by the Center for Advancement of Energy Markets reported that competition has resulted in “More than $3 billion in total savings in 2002 in the Mid-Atlantic (PJM) region, with individual states and jurisdictions saving in 2002: New Jersey, $1.46 billion; Pennsylvania, $993 million; Maryland, $662 million; Delaware, $97 million; and the District of Columbia, $74 million.”20 15.5.1.4. Nationally In a September 2002 analysis of electricity rates across all US markets – opened and closed – former Pennsylvania Commissioner John Hanger reported substantial rate decreases for customers since restructuring began. “For residential customers between 1996 and 2001 where retail generation monopolies were ended, residential rates declined on average 15.9% in constant dollars. Residential rates in states that maintained traditional retail monopolies declined 11.6%.” And “in constant dollars, 11 states cut commercial rates by 20% or more (between 1996 and 2001). Of these, four are non-retail restructured and seven are retail restructured. Industrial rates are the same or down in constant dollars in 23 states and in nominal dollars in 15 of the 29 non-retail restructured retail market states.”21

19

Scope of Competition in Electric Markets in Texas, Public Utility Commission of Texas, January 2003. Sutherland, Dr. Ron, Estimating the Benefits From Restructuring Electricity Markets: An Application to the PJM Region, Center for the Advancement of Energy Markets, September 2003, www.caem.org 21 Electricity Competition: The Story Behind the Headlines, Citizens for Pennsylvania’s Future, September 16, 2002, www.pennfuture.org 20

554


15.5.1.5. California By contrast, not all restructuring approaches led to price reductions. The California experience led to substantial price increases – on the order of 33% between 1998 and 2002.22 Residential prices increased approximately 14% while commercial and industrial customers saw increases of 40%. If the California Energy Commission’s 2003 price estimates are included in the calculation, commercial prices will have risen 47% and industrial prices by 51%. Prior to the crisis, the California market design did not permit prices to fluctuate with the underlying wholesale market, which led, in part, to the dramatic increase in prices. Consequently, the price increases that occurred in California are not the result of decontrolled prices. 15.5.2. Innovation and customer choice In addition to price impacts, electric restructuring has also had substantial impacts from a product innovation and flexibility perspective. 15.5.2.1. Pricing attributes not available in regulated markets Although a regulated industry can attempt to emulate what might happen in a competitive market through service improvements, new product offers and other customer-driven improvements, regulated utilities will not be able to emulate certain product attributes that may be valued by customers, such as the desire for a single provider, a single bill and a single relationship for generation service across multiple regulated service areas. This is a common selling point among competitive retail providers who serve multiple US markets. In addition, the competitive market offers substantial flexibility and customization of pricing not available under regulated tariffs. In an analysis of products in large customer markets,23 KEMA summarized a wide range of pricing structures that are being offered in competitive retail markets based on interviews with several retail providers. Table 15.4 provides a summary of the primary methods of commodity pricing being deployed in US electric markets. Transactions are often structured to allow for multiple forms of pricing and there are numerous variations on the identified pricing types. We classify retailer pricing into three general categories: (1) simple, (2) custom or complex and (3) option and risk management products. In terms of number of customers, the simple category comprises the vast majority of products provided in the market, while the others are typically specialty products used to serve large and sophisticated buyers that may comprise 5–10% of the customer base and 20–50% of the load. 15.5.2.2. Product customization and flexibility not available under regulated regime To date, competitive power markets have resulted in substantial levels of product customization and flexibility to both meet customer needs and allocate risk to the party willing to bear it. Although some of the contracting flexibility demonstrated in competitive power markets can be emulated by default providers and integrated utilities, many cannot. The following 22

According to the California Energy Commission website, the California state wide weighted average retail electricity price including municipal utilities, went from 10.09 c/kWh in 1998 to 13.41 c/kWh in 2002. see www.energy.ca.gov/electricity/statewide_weightavg_rates.html 23 Tschamler, Taff, Product Success in Power Markets, KEMA’s Retail Energy Markets advisory service, December 2002.


555

Table 15.4. Retail electric commodity products in competitive markets. Pricing types Simple Percent off tariff or guaranteed savings

Description Provides for guaranteed savings off utility default service. Price will vary if utility price varies. Price will remain constant if utility price is constant.

Shared savings

Retail price usually tied to default service pricing approach that results in savings calculation and sharing between retailer and customer of savings.

Fixed energy price

Cents per kWh price for generation only with no demand or fixed standing charges. Delivery of wires charges are either passed through or billed separately.

Fixed bundled price

Single cents per kWh price for both generation and wires service. May or may not have a fixed monthly charge or a demand charge.

Fixed energy and capacity

Fixed cents per kWh energy charge plus fixed monthly charge or variable peak demand charge.

Custom or complex products Electric indexed price Gas indexed price

Price subject to variation based on index, usually a nearby liquid wholesale electric hub. Heat rate formula used to set price level results in price that varies by price of natural gas.

Wholesale blocks

Fixed prices for power supply that is NOT full requirements. Customer buys shaped energy, including ancillary services and transmission, separately.

Fixed-variable combination price

Price is fixed up until some strike price (based on usage, demand, wholesale price, etc.), then index price is applied.

Time differentiated fixed price

Time-of-use pricing varies by time of day or type of day. This is often termed multi-part pricing.

Real time or dynamic price

Price varies in short-time intervals (e.g. hourly spot market price). Products typically structured in ISO markets using day-ahead or hour-ahead spot market prices.

Contracts for differences

Price is fixed under financial contract, but physical supply is purchased through spot markets. Variations between spot market price and retail contract price are then settled.

Block or usage-tiered price

Price level varies by amount of usage; could be tiered usage or other mechanism.

Consumption-adjusted price (collars, caps, etc.)

Fixed price based on specified consumption parameters (load factor, load shape or periodic volumes). Variation outside specified parameters results in change in price level, usually through a market-based price.

Option and risk management products Trigger pricing (form of option)

At specified trigger price, retailer makes wholesale purchase for customer. Customer typically has load reduction capability or on-site generation. (continued)

556


Table 15.4. (continued) Pricing types

Description

Curtailable/interruptible price

Fixed term or indexed price that is discounted to allow retailer to exercise option to reduce load under certain conditions.

Interruptible load and on-site generation options

Retailer pays customer for right to use customer asset (interruptible load, on-site generation) under certain conditions. Arrangement can be in form of revenue-sharing agreement.

Swaps

Customer trades a product (e.g. natural gas) for electricity at an agreed-upon price or trades a type of price (e.g. variable) for another type of price (e.g. fixed).

Supplemental supply options

Provides holder (retailer or customer) with right, but not obligation, to buy (“call”) or sell (“put”) power at a set price. (Could be structured as a trigger price.)

Insurance products

Customer pays a premium to retailer for reliability or power quality. In case of reliability event, retailer (most likely through an insurance company) pays customer ‘damages.’

Integrated pricing

Price of energy supply, interruptible load, energy services, etc., bundled into a single, integrated price that may be in price per square feet or some other non-commodity unit.

Unit of output pricing

Price indexed to price of customer’s product (customer sells product at higher price, power price goes up and vice versa). Retailer must hedge through relevant commodity futures markets. Enron offered this prior to bankruptcy.

provides several examples of the flexibility of existing competitive power products and services: ●

●

●

Allocation of volume risk: Customers have choices with respect to the volume or consumption risk they bear. The use of specified quantities through bandwidth provisions on quantities, stipulated quantities over a time horizon, block power purchases and unlimited quantity contracts permit customers to pay prices that vary depending on their risk tolerance. The more tolerance for risk, the smaller the premium and lower the price paid. Customer service: Competitive markets have led to distinctions among providers through variations in billing and customer services offered, including varying billing cycles, payment and credit requirements, key account management and contact center technology and staffing. Segment-driven price variation: Under regulated regimes, the primary customer segment is the rate class. Although some variation in products and services are provided through this structure, it limits the variation and specialization the regulated provider can offer. In competitive markets, customer segmentation drives varying product offers. There are many examples of product variation based on customer attributes and needs, including: ● Customers that do not want to pay a demand charge and are willing to pay a premium to avoid them can pay a simple volumetric charge in open markets such as Texas. ● In the UK a primary differentiator of electricity offers for small customers is how each customer pays their bill: direct debit customers receive the lowest price.


●

●

●

557

Similarly, business customers across US open markets that are high credit risks generally pay a higher price than lower credit risk customers. Customers that have the ability to self-generate or reduce their load, can participate purchase market-driven demand response products that effectively lower their overall energy price because of the flexibility provided to their suppliers. Contract provisions, such as extensions and “blend and extend,” are offered to certain segments that value a price discount for the obligation to have their contract extended.

15.5.2.3. Customer choice The ability to choose the provider of the products and services we purchase is a fundamental tenant of a democratic and capitalist society. Beyond, price, product and service improvements, the ability to choose an electric provider is a benefit in and of itself. 15.5.3. Shifting risk from customers to suppliers Finally, shifting capital and operational risk away from consumers and toward suppliers and their shareholders is arguably the clearest and most significant benefit of electric restructuring. By shifting these risks to decision-makers, suppliers and their shareholders will be rewarded for superior investment and operational decisions, while the financial impacts to individual firms of poor decisions or adverse events will not be borne by their customers, but rather their respective shareholders. Under traditional rate of return or cost-plus regulation, the many risks (and realized costs) associated with providing retail generation service to customers are born primarily by customers, rather than the utilities. Although utilities have been subject to prudency reviews during rate cases, the vast majority of costs, including billions of dollars of uneconomic investments (i.e. stranded costs), incurred by utilities have been passed on to rate payers. Stated more broadly, the entities that are investing capital and managing operations do not fully bear the associated risks. Consequently, utilities generally face poor incentives to manage risks, optimize capital investments and minimize costs. A 1997 report addresses the issue of utility incentives: When the local utility acts as the “designated” purchaser of power, its decisions, no matter how competitive wholesale power may be, become largely immune from market discipline and, instead, subject to the judgment of regulators. This means that retail customers would continue to bear the brunt of bad decisions, thereby at most only marginally affecting the incentive of the utility to make better decisions.24 The competitive market does not provide a rate base to shareholders to ensure recovery (plus a return) of cost overruns, nor recovery of investments made in technologies that are not competitive, nor recovery of the costs incurred by firms that do not meet the expectations and needs of their customers. At least 50 retail competitors have either gone bankrupt or exited the retail power business since US markets first opened. Collectively, hundreds of millions of dollars were invested by these firms to develop the infrastructure required to serve customers. Customers have not had to bear the burden of the poor performance and 24 Costello, Kenneth, Kenneth Rose and John Hoag. An Assessment of Retail Competition in Kansas’ Electric Power Industry, September 1997.

558


Business model components Target customers

Example Retailers

Product portfolio

Supply approach

Footprint

Niche independents

Small and mid-market customers

Simple set of fixed priced offers

Typically outsourced or longterm partnership

1 or 2 service territories

Advantage Energy Ecoenergy ElectricAmerica Texas Commercial Energy

Home Market affiliates

Targeted segments usually by business type

Primarily fixed price offers, but some index or custom deals

Typically driven by legacy assets with modest trading capabilities

1 or 2 service territories

Exelon Energy PPL EnergyPlus Ameren Energy

Full Service full segment affiliate

All customer groups targeted

National supply shop

Generally larger customers or targeted segments

Full range of deal structures – commodity supply only

Large industrial specialist

Large industrials

Integrated super regional

Generally mid and large customer segments

Typically regional Full range of deal integrated model with structures plus value contracts, legacy added services assets and merchant assets

1 or 2 states, several service territories

ConEd Solutions Key Span Energy Reliant Energy Solutions TXU Energy Services

Contract driven with skills and positions in most major supply markets

Most or all open markets

Constellation New Energy Strategic Energy Tractebel Energy Services

Full range of deal structures, plus sophisticated risk management services

Contract and asset driven with substantial risk management skills

Varies

Full range of deal structures plus value added services

Contract and asset driven, with regional market expertise

Several states, usually contiguous, many service territories

BP Calpine Coral Power First Energy Solutions Pepco Energy Services Select Energy Sempra Energy Solutions

Fig. 15.14. Business models of US retail suppliers.

investments these firms have made. Over time, this realignment of risk will improve economic efficiency and innovation, and ultimately decrease electricity costs.

15.6. An Overview of the Players The business models of US retailers are diverse, from small independent retailers with targeted customer segments to national retailers with a full suite of services and customer segments (Fig. 15.14). We define retailer business models according to four critical strategic and operational elements of the retail business: (1) target customers, (2) product portfolio, (3) supply approach and (4) geographic footprint. As shown in the figure, we have identified six primary business models: 1. 2. 3. 4. 5. 6.

Niche independents. Segment-focused, local affiliate. Full service, full segment affiliate. National supply shop. Large industrial specialist. Integrated super regional.

Although variation within models exists, the vast majority of retailers fall into these categories. The business models have evolved substantially over time, generally reflecting improvements in retailer capabilities and a much more disciplined approach to managing cost


559

Table 15.5. Top five non-residential retailer load served, Summer 2005. Top five suppliers

Peak MW served

Constellation New Energy Reliant Energy TXU Energy Strategic Energy Suez Energy Resources

15,600 8,600 5,700 3,500 3,200

structures and cash flow. The primary shifts we have identified over the past 2–3 years include: ●

● ●

●

● ● ●

more targeted and well-defined value propositions resulting in focused customer segment specialists; more incremental, rather than large scale, geographic rollouts; higher levels of product flexibility and customization in certain markets for certain customer segments; move toward more strict volume-matched retail businesses, rather than a mix of wholesale and retail margin extraction under single retail business units; significantly improved portfolio management techniques and risk management expertise; more wholesale merchants are moving into retail, serving large customers; certain niche startups are achieving short-term profitability.

With over 200 suppliers in 18 states, the US retail market is characterized by a wide diversity of firms. We estimate affiliates of local utilities account for roughly one-quarter of the total and serve 70–80% of the load. The market is highly fragmented, with only a handful of retailers considered national and less than 20% operating in multiple states. The following summarizes some of the firm attributes in non-residential markets as of the Summer of 2005 (Table 15.5): ● ● ● ● ● ● ●

top five comprise ⬃55% of total market; top 20 comprise ⬃85% of total market; concentration increasing slightly; thirteen companies serve over 1000 MW; ⬃75 firms serve business markets; ⬃250 licensed to serve; Thirty-five out of top 50 operate in Texas.

The following summarizes some of the firm attributes in residential markets as of the Summer of 2005 (Table 15.6): ● ● ● ●

thirteen companies serve over 100 k power customers; ⬃45 firms serve residential markets nationwide; ⬃200 licensed to serve nationwide; Forty-five percent of accounts served competitively in Texas.

Although the top retailers in each customer segment have remained relatively static over the past 3 years, competitive retail energy markets have seen significant levels of entry and exit activity. Figure 15.15 summarizes the level of entry, exits, mergers and change in ownership of competitive retail energy companies.

560

Electricity Market Reform Table 15.6. Top five residential retailer customers served, Summer 2005. Top five suppliers

Customers

TXU Energy Reliant Energy Retail Centrica North America Green Mountain Energy Energy Savings Income Fund

2,150,000 1,650,000 950,000 600,000 425,000

Changes in ownership Consolidations Exits Entrants

2002

2003

2004

2005 0

10

20

30

Fig. 15.15. Retailer entrants, exits and transactions, 2002 to current (number of firms, transactions).

Table 15.7. The optimal model for default service. Dimension

Status Quo Service, if available

Last Resort Service

Supply obligation Price frequency

Any qualified entity Unregulated – whatever frequency supplier decides; competitively bid

Any qualified entity Hourly, monthly, quarterly, as appropriate

Price setting mechanism

Market (unregulated)

Hourly real time market or RFP with regulator oversight

Price components

Unregulated – whatever components supplier decides; competitively bid

Full requirements wholesale supply, plus retail components and risk premiums

Utility separation

Fully separate T&D service from retail generation service, including related retail services

CTC mechanism

Fixed non-bypassable charge with true up over set time period; no deferral mechanisms beyond set time period; applies to all customer, not just customers that switch

Supplier Ts and Cs

(1) Reasonable credit requirements, (2) imbalance prices set based on spot energy markets with no penalties, (3) data availability, (4) reasonable billing policies and payment hierarchy rules for consolidated billing

Customer Ts and Cs

No limitation on ability of customer to switch on and off default service

Wholesale design

LMP-based ISO with day-ahead energy markets and no mandatory capacity markets


561

15.7. The Optimal Default Service Model Given the principles set forth and the experience to date previously summarized, what then is the optimal model for default service? As the leading US example of a market with robust customer choice and competition, Texas provides important context for the Optimal Model set forth in this chapter. The optimal model for default service can be defined according to the dimensions previously identified and differentiated between Status Quo and Last Resort Service. Table 15.7 summarizes the optimal model.

Box 15.1 The Texas Model: Closest to Optimal The Texas retail power market enabled in 1999 by Senate Bill 7 and specified in the PUCT Chapter 25 Substantive Rules, represents the closest real world example of the optimal model we advocate. The primary components of the Texas model are: ● ●

●

● ●

●

●

●

●

●

Prices have been decontrolled for Status Quo customers over 1 MW. Prices for Status Quo customers under 1 MW are subject to the PTB, a retail generation price based on historical rates subject to changes through a fuel factor formula based on the price of natural gas. Temporary service is provided through the competitively selected Provider of Last Resort. Although not an hourly price, it was designed as a temporary service and prices and service terms are generally unfavorable compared to those in the competitive market and Status Quo Service. PTB rates are decoupled from the underlying supply costs incurred by retail affiliates. Retail affiliates’ earnings and losses are dictated by market conditions and their ability to perform in those conditions; risks associated with affiliate earnings fall fully on shareholders, not rate payers. Incumbent utilities are legally and structurally separated into a generation company, a retail company and a wires company. PTB and POLR generation service includes a retail service obligation; it is a single retailer model. There are three separately administered tariffs: (1) delivery service tariff (Delivery Service), (2) PTB tariff (Status Quo) and (3) Provider of Last Resort (Last Resort Service). One incumbent – Reliant Energy – fully separated its delivery business (CenterPoint) by spinning it off, while another – AEP – sold its PTB obligation to a third party. Non-price regulations are clear and support competition, including a centralized clearing house for retail transactions and extensive customer protection rules.

References Federal Trade Commission, Staff (2001). Competition and Consumer Protection Perspectives on Electric Power Regulatory Reform: Focus on Retail Competition, September. Flaim, Theresa (2000). The big retail “bust”: what will it take to get true competition. Electricity Journal, 13, 41–54.

562


Gordon, Kenneth and Olson, Wayne (2004). Retail cost recovery and rate design. Edison Electric Institute White Paper, December. Graves, Frank and Joseph Wharton (2003). New Directions for Safety Net Services – Pricing and Service Options. Edison Electric Institute, May. Implementing Principles of Default Service: A Roadmap for Competitive Retail Power Markets. White Paper with Frank Lacey, Strategic Energy, April 2004. Jim Rossi (2000). Universal service in competitive retail electric power markets: whither the duty to serve? Energy Law Journal, 27(1). Joskow, Paul (2003). The Difficult Transition to Competitive Electricity Markets in the U.S. AEI-Brookings Joint Center for Regulatory Studies, July. Reitzes, J., Wood, L., Quinn, A. and Sheran, K. (2002). Designing standard-offer service to facilitate electric retail restructuring. Electricity Journal, November, 34–51. Tschamler, Taff (2000). Designing competitive electric markets: the importance of default service and its pricing. Electricity Journal, March, 75–82. U.S. Energy Information Administration (2003). Status of State Electric Industry Restructuring Activity, February, Available at URL: http://www.eia.doe.gov/cneaf/electricity/page/restructure.html

PART V Other Markets


Chapter 16 The Case of Brazil: Reform by Trial and Error? JOÃO LIZARDO R. HERMES DE ARAÚJO Centro de Pesquisas de Energia Elétrica, Cidade Universitaria, Rio de Janeiro, Brasil

Summary The chapter starts with a brief analysis of the historical evolution of the Brazilian electricity supply industry (ESI) from the 1950s to the 1990s, highlighting its specificities and the context of the Brazilian reform. Next the reform process is discussed, with its strong points and shortcomings. A striking feature of market liberalization, which followed the debt crisis of the 1980s, was that the divestment process and market reform followed two parallel and nearly independent paths. This, plus the mismanagement of reform and transition – particularly in view of the difficulties found in privatizing large generators – led to the 2001 power crisis. The new arrangement instituted by the Lula administration purports to ensure adequate expansion investment with an expanded role for central planning and co-ordination including a mechanism for regulated expansion auctions and contracts. There also is a role for short- and middle-term contracting, which may grow. In principle, the new arrangements could solve the conundrum of thermal power investment in a large hydro-dominated system. There are some positive signs that investment in transmission is working. However, two problems remain to be solved: streamlining of environmental licensing of hydro plants, and how to ensure an adequate supply of gas to thermal plants and build the gas network in a market-oriented context.

16.1. Background of the Brazilian Reform Brazil is a large country (8.5 million km2, larger than the USA without Alaska), with continental distances, a population of 180 million, GDP over US $1400 billion in 2004 (the exact value depends on the exchange rate, see Fig. 16.1) and a sizable power system, both in terms of generating capacity (cf. Table 16.1) and of grid extension (cf. Fig. 16.2). A prominent feature of the Brazilian power system is the weight of hydro power, like Canada and Norway, but with very little international trade, unlike either of those (cf. Table 16.2). The significant hydro potential, both operational and to be developed, has shaped Brazilian power policy for the better part of a century. It has also played a prominent role in both power sector reforms undertaken in the last decade, led respectively by the Cardoso and the Lula administrations. 565

566


Rationing 700,00

4.5 4

600,00

Exchange rate

3.5 3

400,00

2.5 Northeast drought

300,00

2

Drought in the South

200,00

R$/US$

R$/MWh

500,00

1.5

SE/CW South NE North R$/US$

1 100,00

0.5

26 /8/ 20 05

20 05 7/2 /

20 04 22 /7/

20 04 4/1 /

26 /10 /20 01 14 /5/ 20 02 30 /11 /20 02 18 /6/ 20 03

20 01 9/4 /

21 /9/

20 00

0

Date Fig. 16.1. Evolution of MAE spot prices for heavy loads and exchange rate (2000–2005). Source: CCEE, July 29, 2005 for prices, Brazilian Central Bank for exchange rate.

Table 16.1. Generating capacity in 2005 by ownership and use (GW). Self-production Self-production with sale of surplus Traders IPPs Public service of which Federal* State/Municipal Private Total capacity

2.86 2.15 0.07 26.30 60.77 32.07 17.06 11.64 92.15

*This includes the Brazilian half of Itaipu, as well as utilities temporarily under federal intervention. Source: Compiled from ANEEL “Banco de Informações de Geração”, updated on September 3, 2005.

These reforms differ very much between them (not least in political motivation); but their differences should not be construed as a simple market versus government dispute. As we shall see, despite a forceful pro-market rhetoric the Cardoso reform included a spot market that was based on a central optimization algorithm with little role for price bids (effectively no role at all in normal times) and a mechanism that removed all competition among hydro power plants; moreover, in 8 years very few hydro power generators were privatized. On the other hand, despite a strong distrust of market schemes at the roots of the Lula reform, free contracting now answers for 18% of all consumption, much more than in the previous administration. Reality has a way of insinuating itself, and it would appear that each reform

567

The Case of Brazil: Reform by Trial and Error?

Sta.Elena Comitê Coordenadordo Planejamento da Expansão dos Sistemas Elétrico s

Ministério de Minas e Energia CoaracyNunes

Boa Vista

Santana Balbina

S.Maria

Altamira Itaituba Belo Monte Marabá Imperatriz

Manaus

Porto Velho Rio Branco Abunã

Santarém

São Luiz

P. Dutra

Fortaleza Açu Natal

Teresina

S.J.Piaui

Colinas

Samuel Ariquemes

Recife

Sobradinho Miracema

Ji-Paraná Guajará Mirim

V.Conde Tucuruí

P.Bueno Vilhena

Futuro

Sinop Sorriso

Gurupi S.da Mesa

Manso

Cuiabá

Rondonópolis

Irecê Gov.Mang

B.J.Lapa

Maceió Xingo Aracaju Salvador

Funil

Brasília Goiânia T.Marias

Corumbá

C.Grande

Existente Futuro 138 kV 230 kV 345 kV 440 kV 500 kV ⫹ 600 kVDC 750 kV

Belo Horizonte

Ivaiporã

São Paulo

Itaipu Sto.Angelo Garabi

Itá

Vitória

Campos Rio deJaneiro Curitiba Blumenau C.Novos

Uruguaiana Livramento

Porto Alegre Candiota

Fig. 16.2. The Brazilian transmission system (tension above 138 kV). Source: Esmeraldo (2004).

Table 16.2. Generating capacity and generation by source and imports. Operational and Generation in the Net international imports monitored capacitya interconnected systemb not counting Itaipuc Conventional hydro (%) Conventional thermal (%) Nuclear (%) Other (%) Total

74.9 21.4 2.2 1.5 92.2 GW

88.7 8.4 2.9 ND 352 TWh

NA NA NA NA 1.0 TWh

a

Data for September 3, 2005; ANEEL, “Banco de Informações de Geração”. Data for January to December 2004; ONS, “Histórico da Geração de Energia”. ONS data refer to the interconnected system only, and include Itaipu. c Data for January to December 2004; ONS, “Histórico da Operação”. b

started from a bold statement but then proceeded to incorporate hard facts by tâtonnement and negotiation. In the end, whether a reform is successful depends on it being able to attract investment in a sustained way. The Cardoso reform failed that test (we shall look into some possible causes), and the jury is still out on the Lula reform. In Brazil, electricity was initially developed by private capital. Until the 1930s, two groups dominated the growing electricity market: the American–Canadian Group Light,

568

Electricity Market Reform 100 90 80

Percent

70 60 50 40 30 20 10 19 7 19 3 7 19 4 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 80 19 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 97 19 9 19 8 9 20 9 0 20 0 0 20 1 0 20 2 03

0

Nuclear %

Thermal %

Hydro %

Fig. 16.3. Installed capacity shares (1973–2003). Source: National Energy Balance (2004).

which controlled power supply in the largest centers of the Southeast, and the American Foreign Power Company (AMFORP), which supplied electricity to several state capitals and a score of middle-sized towns. Besides these large groups, many public and private enterprises supplied power in a small scale to poorer regions. By 1950, 81% of installed capacity was owned by Light and AMFORP in the Southeast. Things began to change during the Vargas administrations: the State entered power generation, transmission and distribution, first through regulation and later via direct investment. Even so, till the 1970s international capital controlled distribution in São Paulo and Rio de Janeiro, and thus in the country’s industrial hub (Araújo, H., 1979, 27). Since the end of the 19th century till now, hydro power has dominated power generation in Brazil. Figure 16.3 shows installed capacity shares for the last 30 years; the share of hydro in effective generation was and is much higher, having reached 95%; even today, after significant effort in adding thermal plants, hydro power plants account for 88% of generation in the interconnected system (Table 16.2). The significant hydro power potential close to major load centers led the first Vargas administration to edit federal legislation (the Code of Waters – Decree 24.643 of July 10, 1934, which became a central piece of Brazilian electricity regulation until recent reforms), and to create a regulatory agency – National Council for Waters and Power – to guide the exploitation of these resources. This regulation replaced the existing ad hoc concession contracts, and led to numerous conflicts between private investors and the government.1 The net outcome of these conflicts was an unwillingness of 1

Conflicts did not arise because the new law was draconian in any sense whatever. In fact, it was remarkably similar in form and spirit to the US legislation enacted at the same time. However, it replaced comfortable contracts for investors (with gold standard clauses, for instance), and there were a number of unclear issues which led to disputes over interpretation, finally solved only in 1957 when a detailed procedure for updating capital base was set. This happened soon after the creation of the BNDE, which managed a National Electrification Fund financed with resources from a specially created tax, the IUEE (Unique Tax on Electric Energy), and created the conditions for the subsequent expansion of the ESI.


569

Fig. 16.4. Hydropower plants in operation, installed capacity above 30 MW. Source: Daher (2005), updated by LABGIS/CEPEL (thanks to P.C. Menezes).

private firms to invest any further, in face of a growing demand and rising consumer complaints against deterioration in the quality of service, which also led self-generation to reach 16.7% of installed capacity in 1953. As a result, federal and state governments were forced to invest, in order to guarantee the supply of electricity (Araújo and Oliveira, 2005). This nationalization process was aided by the country’s vast hydro potential, which required large projects and – given the size of the territory and diversity of water regimes (Fig. 16.4) – led to significant economies of co-ordination in generation with interconnected operation, over 20% in firm energy2 (Santos, 1996). Finally, mainstream economic theory and practice of the 1950s favored state-led growth for economic development.3 Thus, after a decade of political negotiations, Eletrobrás was created in 1961 together with the Ministry of Mines and Energy (MME). The federal government (and a few state governments) gradually took over the generation and transmission activities, more demanding in investments and with longer maturation lags. Distribution, initially under charge of foreign companies, was progressively transferred to state governments. With the sale of Light to the federal government in 1978, the public sector had effective ownership of the system by the late 1970s (Fig. 16.5). 2

This remains a relevant issue, and motivates some apparently peculiar aspects of the Cardoso reform. cf. Hunt (1989), pp.52ff.

3

570


Total das Empresas Públicas

12000

Privadas

MW

10000

Grupo Light

8000 6000 4000 2000 0 1930

1953

1956

1959

1962

1965

1968

1971

Fig. 16.5. Ownership evolution of power firms in Brazil (1930–1972). Source: Till 1962, Memória da Eletricidade (1988); from 1963 to 1972, Araújo (1979, 1986). Table 16.3. Structure of the Brazilian ESI in 1995. Level

Roles

Companies

Binational Federal

Generation Holding, planning Generation and transmission Generation and distribution Nuclear engineering; research Generation and distribution Distribution Distribution Distribution

Itaipu Eletrobrás Furnas, CHESF, Eletrosul Eletronorte NUCLEN; CEPEL CESP, CEMIG, COPEL, CEEE 23 companies 5 companies 25 companies

State Municipal Private

Source: Rodriguez-Pardina and Estache (1996), updated by Araújo and Oliveira (1996).

Nevertheless, the Brazilian power system was never completely centralized as it happened, for example, in France. Eletrobrás was created with the control of the electricity supply assets that belonged to the federal government, without absorbing them. Both the São Francisco Hydroelectric Company (CHESF), created in 1945 and which started generating in 1954, and Furnas Utility in the Southeast, created in the 1950s but which only began generating in 1964, retained their identities and organizational cultures. These utilities, together with other smaller units including isolated thermoelectric power plants in the North absorbed by Eletrobrás, represented at the time circa 20% of total installed capacity in the country. Over the years, other utilities joined what came to be called the “Eletrobrás System”4 and which answered for 61% of generating capacity in the early 1990s (Table 16.3). 4

This name was given to a group of firms controlled by Eletrobrás. In the early 1990s it comprised four large federal generators – Furnas, CHESF, Eletronorte, Eletrosul – plus the Brazilian half of Itaipu; an R&D Center (Centre for Electric Energy Research, CEPEL); and a few more or less accidental accretions: two distribution companies (Light, after nationalization in 1978, Escelsa as a result of state debt negotiations), and a nuclear engineering firm (Nuclen) in 1989 after the breaking up of Nuclebrás. Over the last 10 years there were important changes in the Eletrobrás Group.


571

Besides the direct control of federal government-owned utilities, Eletrobrás participated in state-owned utilities as a minority shareholder. These utilities represented in the early 1960s another 20% of installed capacity (Dutra and Salles, 1975), and reached 33% in the 1990s (the remaining 6% of installed capacity at this time were self-generation). Eletrobrás also controlled sector planning and the financing of sector expansion via the Federal Electrification Fund, taking over this responsibility from the National Bank for Economic Development (BNDE). This fund came from taxes added to tariffs, notably the Single Tax on Electric Power (IUEE) and a Compulsory Loan (charged to large consumers). To understand some crucial features in the evolution of the Brazilian ESI from the 1960s to the 1990s, it is important to note that the system never became wholly centralized, despite the dominant role of Eletrobrás. Even within the Eletrobrás System, nominal control of its subsidiaries was not synonymous with effective control. Appointment to executive office in these reflected in some measure regional political interests, and utility operation was in fact largely autonomous, leading to distinctive local cultures. This was more noticeable in Furnas and CHESF, which antedated the creation of the “mother house” Eletrobrás (a point often cited by their people, half in jest); but it was also perceptible in Eletronorte and Eletrosul. Conflicts of interest arose sometimes, especially with large state-owned utilities like Companhia Energética de São Paulo S.A. (CESP), Companhia Energética de Minas Gerais S.A. (CEMIG) or Companhia Paranaense de Energia S.A. (COPEL), of which Eletrobrás was a minority partner. However, Eletrobrás did control the Brazilian ESI through two main instruments. It managed the electrification fund, acting as a development bank for the sector; and it headed key Co-ordinating Groups or Committees: one for Interconnected Operation (GCOI), one for the North–Northeastern Operation (CCON), and one for System Planning (GCPS). For the better part of two decades after the creation of Eletrobrás, this co-ordinated and partially centralized system had significant success. Availability of cheap money, from multilateral donors like the World Bank, made it possible to exploit economies of co-ordination and interconnection, and to keep up with rapid growth of energy demand (reaching 11% a year during the 1970s), creating large interconnected markets in the South–Southeast and in the Northeast. It also created significant technical expertise in hydro power and large system construction and operation, since the continental dimensions of Brazil and the hydro power option required the construction of long transmission lines at ultra high voltages. By 1970 there were only 32,000 km of transmission lines; in 1993 the figure was 150,000; Itaipu, which began operation in 1984, was a milestone both in plant size (12 GW) and in transmission lines to the Southeast market (1000 miles at 750-kV AC and 600-kV DC). If one juxtaposes the Brazilian grid on a map of Europe, the interconnected transmission system will extend from Lisbon almost to Moscow. The creation of Eletrobrás was followed by that of a new regulatory body. Department of Waters and Electric Energy (DNAEE), created in 1965, absorbed the National Council of Waters and Electricity (CNAEE) and the National Department of Waters and Energy (DNAE). Besides setting tariffs and controlling utility accounts, it was also responsible for all aspects of water resources. By 1968, rate-making procedures were consolidated, ending more than 30 years of ad hoc methods. In 1974, electricity rates were equalized all over the country; this was aimed at creating incentives for industrial investment in less developed regions. Given the structure of the Brazilian ESI, this was done through an equalization fund managed by Eletrobrás: low-cost utilities transferred their surplus to the fund, and the money was channelled to high-cost ones. This balanced utility accounts, but had flaws that became all too visible when troubles began. Trouble started in the mid-1970s, just when the model seemed most successful. The use of public tariffs by economic authorities as an instrument to fight inflation, in the aftermath of

572


the 1973 oil shock, began to deteriorate the financial situation of utilities. Costly investment decisions, made possible by centralization and an authoritarian regime, also contributed. In the late 1970s, heavy borrowing for large expansion projects, spurred on by the abundance of “petrodollars”, backlashed when interest rates skyrocketed after the Iran–Iraq war and the second oil shock of 1979. The sector debt snowballed and, after the 1981 Mexican default, investment in expansion could not be kept at the previous rates. Economic recession brought in by the debt crisis, which characterized the “lost decade” in Latin America, also meant that suddenly there was surplus capacity, adding to the industry’s troubles. The combination of restricted tariffs and high-debt service strangled the sector. The capacity for self-financing,5 which had been around 60% in the mid-1970s, dropped to extremely negative levels, oscillating between ⫺60% and ⫺80% in the second half of the 1980s (Fig. 16.6). The Acronym for a task force created in the mid-80’s for Institutional Review of the power sector proportion of resources aimed at new investments fell from 71% in 1974 to 54% in 1980, 40% in 1985 and 29% in 1988 (Fig. 16.7). Paying service on sector debt consumed twothirds of resources by the late 1980s. The tariff equalization scheme further aggravated things: low-cost utilities, facing cash problems, ceased to contribute to the equalization fund. The holding company and development bank Eletrobrás thus had to bear the whole burden of debt, also for utilities it did not control – among them CESP, which served the huge São Paulo market and began defaulting on energy payments to the generators. Solutions were attempted to solve the financial impasse along two lines: recovering tariff levels (Programme for Sector Recovery or PRS) and renegotiating a feasible institutional agreement (a sector task force, known as REVISE, was created for this) while keeping the overall framework intact; but all attempts failed. On the one hand, tariff recovery in a context of very high inflation proved a Sisyphus task. On the other, conflicts of interest between federal and state governments impeded agreement on the REVISE proposals. Another issue was created by the 1988 Brazilian Constitution, which abolished the Federal Electrification Fund, together with IUEE and the compulsory loan. Given the twin financial crises of the power sector and of the Brazilian State, the scene was set for reform. By 1992 no less than six bills proposing increased role for private capital, with various degrees of decentralization, were being debated in Congress (Araújo and Besnosik, 1992), but the first steps were taken in 1993.

16.2. The First Reform Process (1993–2002) Electricity reform took place in a vastly changed macroeconomic context. Since the first Vargas administration in the 1930s, economic policy had been centered on import substitution as the way to industrialize. The State took a leading role in the industrialization process after the 1950s, especially for infrastructure industries, following mainstream development theories of the time (Hunt, 1989); this was (rather paradoxically) enhanced under the military governments, despite their rhetoric. During the 1970s and 1980s, another feature characterized Brazilian macroeconomic policy: widespread indexation, first introduced as a “gradualist” instrument to fight inflation, became both a mechanism for sustaining inflation and a way to live with it. But after two oil shocks and the debt crisis, inflation was effectively running out of control, in three-digit figures. The failure of heterodox plans to fight

5

(Own resources less debt service)/new investment.


573

inflation, starting with the Cruzado Plan,6 led to a radical departure with past economic policy. Import substitution policy was abandoned, and protection schemes to Brazilian producers were progressively removed. Privatizations and market liberalization were also started at the time, in several industries. The process was carried out by three distinct administrations: having started under Collor de Mello (1990–1992) with much fanfare but debatable measures, it was continued with less drama under the Franco (1992–1994) and Cardoso (1995–2002) administrations. The first steps for electricity reform were taken in 1993. Law 8631 ended guaranteed remuneration, as well as tariff equalization, and annulled intra-sector debts; the National Treasury absorbed the US $23 billion debt in order to clean the sector slate. In the same year, Decree 915 (superseded in 1996 by Decree 2003) allowed the formation of public and private consortia for power generation, and another one (Decree 1009) created the National System of Electric Transmission (SINTREL), an agreement among utilities to open the interconnected grid for third-party use. Since joining the agreement was mandatory only for federal enterprises, being voluntary for others, this had the same destiny of similar attempts elsewhere – it never really worked because large integrated state-owned utilities felt they had nothing to gain, and much to lose from it. Reform went ahead in 1995; Law 8987 dealt with the overall concession regime, introducing price cap and competitive bidding for all public concessions, and stimulating selfproducers as well as private investment in general. Another act (Law 9074) constrained utilities to give up several of their concessions for hydro power plants that they had not begun building, and also to present financial plans for finalizing those power plants for which construction had been halted. Moreover, this act forced utilities to give open grid access to large energy consumers (above 10 MW immediately and above 3 MW after 2000), who were free to choose suppliers, and also allowed room for independent power producers (IPPs) to supply these consumers. These reforms led state-owned utilities, with power plants under construction, to look for private partners to conclude them. Some private investors formed consortia, often including large industrial consumers, to take advantage of this opportunity to invest in power plants. (cf. Table 16.4). Still in 1995, the federal government launched a project for restructuring the power sector, which came to be known as RE-SEB. The original idea was to privatize all of distribution, transmission and generation, with exception of nuclear plants and the Brazilian half of Itaipu, and to introduce competition. To this purpose it commanded a study to a consortium led by Coopers and Lybrand, which delivered its report in July 1996. Even as the study started, the government decided to go ahead with utility privatization to show their commitment to reform. The divestment program began with two distribution companies which had come into the Eletrobrás Group more or less accidentally – ESCELSA in Espírito Santo and LIGHT in Rio de Janeiro. It was not an easy operation, since in the absence of a clear picture of the structural and regulatory set-up investors were understandably hesitant. In the end, ad hoc clauses – like an X factor set to zero for 8 years, a readjustment index (IGPM, a bulk price index) correlated with commodity prices and exchange rates, and lowered requirements for energy quality – had to be added to the contracts to find takers. 6

This was based on the notion that inflation had become essentially inertial and approximately stable although at high rates. Thus, what was needed was a change of currency, together with price and wages freeze (with some corrections to average out effects). It might have worked if adjustments had been made to allow for violations of its basic assumptions; but they weren’t made and inflation rebounded when the plan was abandoned, less than 1 year later.

574


Table 16.4. Major events in the Brazilian ESI reform. Event

Date

Comments

Law 8631

1993

Decree 915 Decree 1009 Law 8987

1993 1993 1995

Law 9074

1995

RE-SEB started

1995

ESCELSA divested LIGHT divested Law 9427

1995 1996 December 1996 1997 1998 1998 1999 1999

Ended guaranteed remuneration, tariff equalization, annulled intra-sector debts Allowed public and private consortia for power generation Created SINTREL to allow third-party access Introduced price cap and competitive bidding for concessions Forced utilities to let go of projects not started and present financial plans for finishing ongoing projects. Also forced open access to large consumers and gave space to independent producers Project for restructuring and privatizing the ESI. Commanded study to consortium led by Coopers and Lybrand First privatization, on ad hoc bases. First large DISTCO privatization, also on ad hoc bases Created ANEEL; first board appointed 1 year later

Law 9478 Law 9648 Initial contracts MAE agreement CCPE created ANEEL Resolution 178 MAE begins operation Divestment ceases ANEEL Resolution 290 Rationing starts Intervention in MAE Kelman report Rationing ends Law 10433 Laws 10847, 10848 Decree 5184 First auction of “existing energy”

2000 September 2000 2000 2000 May 2001 2001 July 2001 March 1, 2002 April 2002 March 15, 2004 August 2004 December 2004

Created CNPE, but first activities in November 2000 Created ONS (System Operator), MAE (Bulk Market) Assigned by ANEEL to existing generators, to expire in 2005 Provisional agreement for the bulk market Organism charged with transmission planning, plus indicative for generation, replacing GCPS Limits to market shares Definition on rules; “NV”

Required changes in market rules Creation of GCE; Kelman report commanded MAE restructured; COEX replaced by COMAE Diagnosed management and investment problems GCE changed into CGSE. Risk aversion curves introduced into reservoir operation Restructured MAE, changing governance New electricity arrangements, after 10 months of negotiation EPE; CMSE replaced CGSE; CCEE replaced MAE Created EPE Prices below what was expected

Following the study recommendations, in December 1996 Law 9427 created ANEEL (Agência Nacional de Energia Elétrica) as regulating agency for the sector (although its first directing board was nominated in December 1997, when 10 utilities had been privatized for a total sum of US $12 billion). ANEEL is linked to the MME, and was assigned the responsibility for granting concessions and authorizations for all agents, as well as organizing auctions, besides more classical regulatory functions. It has administrative and financial autonomy (being financed by a 0.5% surcharge on electricity tariffs), and a directing board whose five members have fixed 4-year terms, staggered so that each year one member is replaced.

575


60 40 Percent

20 0 ⫺20 ⫺40 ⫺60 ⫺80 1975

1977

1979

1981

1983

1985

1987

Year Fig. 16.6. Self-financing capacity (Sector resources less debt service)/investment. Source: Araújo and Besnosik (1992).

Board members are nominated by the President of the Republic and have to be approved by the Federal Senate. Law 9427 assigned ANEEL a broader mission than that usually associated with the role of regulator: it was also charged with planning, granting concessions, operating auctions and other government functions, under government delegation. Given the breadth of its mission, and the complexity and novelty of the tasks it faced, it has done rather well (despite some criticism of its tariff reviews, understandable since its staff had to deal with a host of nonregulatory functions); its autonomy was only threatened in the 2001 crisis, when pressure was exerted on its Director-General to quit (he did not) and a special committee took extraordinary powers. The Lula administration has recently passed a law redefining its mission; some quarters have seen in it a threat to the autonomy of the agency, but in fact it only takes back government functions that Law 9427 had delegated to ANEEL. Although ANEEL is linked to the MME, it also answers to Congress and has shown a remarkable transparency in its actions. In 1998, Law 9648 created the Operador Nacional do Sistema (ONS) for operating the physical system, and the Mercado Atacadista de Energia Elétrica (MAE) to handle the bulk power market, besides creating conditions for restructuring and divesting enterprises in the Eletrobrás Group. ONS is a private association, with members divided into nine categories: centrally dispatched GENCOS (including the Brazilian representative at Itaipu), TRANSCOS, large DISTCOS, large consumers, power importers and exporters, MME, consumer councils (with representatives indicated by ANEEL) and other producers and distributors. The last three categories have a seat but no right to vote in the General Assembly (however, MME has a vote in the Administrative Council). Operationally, ONS absorbed utility personnel who were active in GCOI and CCON, so that system operation carried on much as usual. MAE was to be instituted through a Market Agreement among interested parties. For many reasons, not least of which was an unsolved conflict of interests between GENCOS and DISTCOS, this proved a lengthy and awkward process, with less than successful resolution. MAE arrived at a preliminary agreement in 1999, creating a governance and administrative structure, and rules were approved in 2000; but operational problems led to delays, then to substantial changes in its governance mechanisms in 2001 and 2002, and finally to its substitution by another scheme in 2004. An administrative act by MME created, in 1999, a Co-ordinating Committee for Electricity Planning (Comitê Coordenador do Planejamento da Expansão dos Sistemas Elétricos, CCPE). This organism, which took over from the earlier GCPS, would prepare indicative plans for generation expansion, but mandatory plans for transmission expansion. It answered

576


Milhões de R$ (2000)

25.000

20.000

15.000

10.000

5.000

0 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Ano Fig. 16.7. Investment in the Brazilian ESI (1978–2000). Source: Pinhel (2000), updated by Pinhel, and Araújo (2002).

to MME, although its activity should be consistent with that of an inter-ministerial body, the National Council for Energy Planning (CNPE). This latter body had been created in 1997 by Law 9478, but only started functioning in November 2000, in the eve of the 2001 power crisis. The Cardoso ESI reform followed in broad lines the original central pool of the British approach: centrally co-ordinated physical dispatch, transportation remaining a natural monopoly but competition being introduced in generation and marketing (Oliveira, 1998). An important element of the reform was the bulk electricity market (MAE), intended to replace the former economic central co-ordinating procedures by market mechanisms, while ONS operated and dispatched the system. The bulk electricity market maintained in part the centrally co-ordinated economic dispatching of power plants, decided by computer optimization. Computer optimization codes determined ex-ante spot prices, based on demand forecasts, to inform the market about expected prices, though actual prices were fixed ex-post based on real dispatch. Hydro power plants provided information on technical parameters, while thermal power plants informed prices and technical parameters as well, to construct a supply curve. Large consumers were free to choose their supplier and might submit demand bids. Thermal power plants might choose the amount of power that would be subject to the optimizing process and the amount of power that should be dispatched, irrespective of cost. The former one was called flexible, the latter inflexible power. A short explanation may help understanding how the spot price is fixed in the MAE setup (see also Fig. 16.8). First, demand is estimated and free consumers inform the maximum price they are willing to pay, so that a demand curve can be constructed. Next, dispatch of inflexible thermal power is taken into account to produce the residual demand curve, to be supplied by flexible thermal (who bid their prices; inflexible ones have to accept system price) and hydro power. The supply curve for this ensemble results from the optimum dispatch generated by computer models, which estimate water value.7 The intersection of 7

This was taken from a chain of stochastic optimization codes, from a 5-year horizon with monthly decisions down to a weekly horizon with half-hourly dispatch, based on 2,000 generated future water flow series.


577

demand and supply curves provides the system spot price and the energy to be generated by hydro and flexible thermal power plants, as a pre-dispatch. Actual price is given ex-post, including system charges. There was to be an additional capacity charge (similar to the British one but with availability declared yearly), which was never implemented after negative reaction at public hearings. The “spot” market actually gave weekly8 prices and quantities bought and sold, while physical dispatch was made hourly. As a hedge against volatility, generators, distributors and free consumers had to make bilateral contracts for at least 85% (later 95%) of their production or consumption, reported to MAE. To start the procedure, existing generators received initial contracts by fiat from ANEEL in 1998, at the going rates. Beginning in 2002 the amount would be progressively decreased at a rate of 25% a year until 2005, when initial contracts would be ended. It was hoped that freely made contracts would replace them, but things did not work out as planned – after the 2001 crisis, spot market prices remained so low that distribution companies would rather buy from the spot market than enter into long-term contracts. As a result, generator revenues fell vertically. In order to avoid the abuse of market power, restrictions were imposed on vertical integration of firms: a utility could not own generation and distribution representing more than 25% of the South, Southeast and Center-West submarket, 35% of the North/Northeast one or 20% of the Brazilian market (ANEEL Resolution 178, 2000). Another mechanism was created to limit cost pass through by distribution companies signing bilateral contracts. “Normative values” (NV) were specified by generation technology, according to expected generation costs. The idea was that utilities could pass through 100% of the costs, if these lay within 5% of NV. If the price was below 95%, the utility could retain part of the extra profit and if above 105% of NV, utility had to incur in extra costs. Unsurprisingly, this matter gave rise to much discussion on what would be correct values for the NV’s and about their necessity within a market-oriented reform. Acknowledging that hydro power plants use what are essentially public resources, all hydro plants with installed capacity above 30 MW had to obtain a regulation concession; smaller hydro plants, and thermal plants as well, required only authorization. Recognition of the peculiar problems faced by the hydro subsystem led to another mechanism: hydro power plants9 participated in the Energy Reassignment Mechanism (MRE), a financial device to reduce hydrological risks. In this, each plant receives a share of the total energy generated by hydro power plants, proportional to its assured energy,10,11 irrespective of its actual generation. The plant can often generate non-assured energy, which may be sold in the spot market but not contracted. By contrast, thermal power plants (outside isolated systems) were able to contract their effective installed capacity, after allowing for down time.

8

Prices were set monthly from September 2000 to June 2001. And thermal plants in isolated systems, which had their fuel subsidized through the CCC (Conta de Consumo de Combustíveis, or fuel consumption account) mechanism. 10 In a hydro-based power system with large reservoirs, energy is more important than capacity. Normally, installed capacity is much larger than firm, or even average capacity due to seasonal and yearly variation in water inflow. This makes it relatively easy to cope with peaks of demand, in contrast with a thermal-based system. 11 Assured energy is defined as what the plant (hydro or thermal) can produce under optimized system operation with 5% risk, simulated for a large number of synthetic series of water inflow. This is a relatively new concept. Traditionally, firm energy is defined as the maximum energy that can be continuously generated under the worst historical water inflow sequence (critical period). Recognition that water inflow is a random process led to the definition of assured energy as a probabilistic version of firm energy. 9

578


This peculiar arrangement has been justified because of the large economies of co-ordination in hydro power generation, both within a hydrological basin and among basins – over 20%. Norway has developed an alternative mechanism to deal with within-basin co-ordination: basin agreements, whereby plants within a basin operate a sort of watercredit system, thus decoupling physical and market operation. However, in Norway there is no co-ordination economy among basins; in the Brazilian case, gains from competition would have to be greater than the loss from poorer co-ordination among basins, and few people in Brazil are willing to gamble on that.12 The dominance of hydro power has also led to the use of stochastic optimization programs to define prices and dispatch. It is worth noting that other hydro-rich countries did not have to resort to similar schemes, either because their systems have a large share of thermal power (see chapter on Colombia and Argentina by Dyner et al in this book) or have a large trade with thermal-based systems, as Norway does (see chapter on Nord Pool by Amundsen et al in this book). The MRE may be interpreted as a feature to introduce co-ordination into a market-oriented system, and thus to reap the non-negligible co-ordination gains in a large hydro-based system as the Brazilian one. Thus, an interesting consequence of this financial hedging mechanism is the elimination of competition among hydro power plants, as energy assigned to each power plant depends only on assured energy, not on actual generation. Thermal power plants, on the other hand, have to compete not only with each other, but also with the co-ordinated supply of hydro power plants. The issues regarding the transmission system led to another scheme. Given the historical centralized system planning and operation, there was no incentive to allow capacity slack in the grids larger than justified by risk management and energy optimization. Indeed, from the viewpoint of such a system, slacks large enough to accommodate effective competition were meaningless and an inefficient use of resources. The Brazilian grid was built to bring electricity from (often distant) hydro power plants to consumption centers, allowing just enough redundancy for reliability and economies of co-ordination in a centrally commanded system. Under a market-oriented restructuring, it effectively divided the territory into several subsystems. This was recognized and MAE was split into four distinct submarkets (Fig. 16.9): Southeast/Center-West (the largest, with most competition potential), South, Northeast and North. Meanwhile, privatization proceeded even though rules were still in the making. Most distribution companies were privatized, as well as a few generators. Eletrosul was split into a federally owned Transco (which kept the original name) and a Genco – Gerasul, acquired by Tractebel. In São Paulo, CESP was split into a state-owned Transco (CTEEP, now Transmissão Paulista after merging with another state-owned Transco) and three GENCOS, two of which were privatized and one could not be sold, retaining the name CESP. Tables 16.3 and 16.5 show the system before restructuring, in 1995, and its present status. Overall, 23 enterprises were privatized, for roughly US $22 billion (Losekann, 2003) – the system adopted was that of onerous concession: bidders had to buy the concession from the government, basing their bids on existing tariffs and starting from a government estimate for the net concession value. The highest bidder won the auction. However, despite the impressive performance of DISTCO privatization, this was not a uniform success. Of the large integrated utilities owned by state governments, only CESP followed the federal line – because the São Paulo and Federal administrations were both

12

Also, the Norse scheme rests on a long history of co-operation and would require extensive tests before introduction in the Brazilian context.


ps

579

S

ps* QI Q Dr

⌬Q

QD

Qmax

Q

Fig. 16.8. Price-making in the MAE spot market. In this diagram Qmax is the existing capacity in the system, QD the forecasted demand, QI inflexible capacity. Demand (maximum price) bids from large consumers are represented in the step-ladder curve, and the residual demand curve starts at QDr. System price ps* is found at the intersection between the residual demand curve and the supply curve construed by an optimization program.

run by incumbents from the same party, PSDB. The remaining three large integrated utilities in the Southeast and South (CEMIG, COPEL and CEEE, Companhia Estadual de Energia Elétrica S.A.), in states run by opposition parties, remained state-owned and integrated. CEEE, the Rio Grande do Sul state utility, went as far as to unbundle its coal-fired generation into a new Genco (Companhia de Geração Térmica de Energia Elétrica S.A., CGTEE), but it remained unsold and, in 2000, was absorbed into the Eletrobrás Group. To this, another hurdle was added. It proved much harder to privatize the large generators than it was initially thought. Several causes may be pointed out for this. Regional political interests were one of them, but by themselves would not explain the difficulty found; after all, those same interests were happy to privatize distribution companies. Other important, perhaps dominant, factors to be considered were the drying up of international capital for direct investment in emerging economies after the crises in the Far East and in Russia (plus the 1999 devaluation of the Real), and the role of large hydro power plants in their regional economies. The first factor dampened the willingness of potential buyers, while the second one strengthened the opposition to selling large generators. The Brazilian hydro power system was built upon big plants, with large reservoirs. Of 144 plants with more than 30 MW of installed capacity, 21 have reservoirs with more than 5 million m3 and 32 have reservoirs with capacity of regulation over several years. Such reservoirs have multiple uses, and in fact some large plants like Furnas (operated by the utility of the same name) and Sobradinho (operated by CHESF) had and still have a profound impact in the economies of Minas Gerais and of the Northeast, to cite only two. Divestment of such a plant would have significant implications for stakeholders in these regions, in ways that are far from clear, and would therefore require extensive negotiation with all interested parties. A parallel with TVA or BPA springs to mind. For all of those reasons, privatizations lost momentum before reform had been completed. One striking feature of the Cardoso ESI reform was that privatization and reform followed nearly independent schedules, instead of the “text-book” sequence restructuring– regulation–privatization followed by the British. The misalignment between reform and divestment processes generated a heavy backlog of ad hoc contracts and measures, and

580


Table 16.5. Structure of the Brazilian ESI in 2005. Type

Activities

Enterprises

Binational Federal ownership

Generation Holding Generation Generation, transmission Generation, transmission, distribution Distribution Nuclear Engineering Research Generation, transmission, distribution Transmission Distribution Distribution Generation Distribution

Itaipu Eletrobrás Eletronuclear, CGTEE Furnas, Chesf, Eletrosula Eletronorte Boa Vista, Manausb NUCLEN CEPEL CESP, CEMIG, COPEL, CEEE Transmissão Paulista 11 companies 5 companies 23 companiesd 40 companiesb

State ownership

Municipal Privatec

Source: Araújo and Losekann (2001), updated by Araújo with ANEEL data. a Between 1998 and 2004 Eletrosul was limited to transmission. b Five DISTCOS have been returned to government hands and are being temporarily managed by Eletrobrás: Ceal (AL), Ceam (AM), Cepisa (PI), Ceron (RO) and Eletroacre (AC). Eletronorte also controls Boa Vista Energia and Manaus Energia. c There are also 423 free consumers, 63 IPPs, 13 self-producers, 1 importer/exporter and 44 energy traders registered with CCEE, of a total 608 market participants (CCEE data from September 7, 2005). d Including special purpose societies and joint ventures.

some unpleasant surprises. One such surprise was finding that substantial modifications to contract clauses would have to be added to ensure the quality of electricity supplied by distribution companies, after a series of brown- and blackouts in the very hot Summer of 1997/1998 obliged the newly appointed board of ANEEL to apply heavy fines on Light/EDF for below-par maintenance. Subsequent concession contracts were much stricter, but ANEEL had to laboriously renegotiate contracts already signed. Other problems concerned consumer tariff readjustments. During the first years of the divestment process everything went well. After the 1999 Real devaluation, however, the choice of index meant that utility tariffs started growing much faster than the consumer price index (see Fig. 16.10), leading to discontent, pressures on the budget of poorer households, and an increase in defaulting and other commercial losses. Investors were not happy either, because, measured in dollars, revenues did fall somewhat against their expectations. And some utilities, most notably Light/EDF in Rio and Eletropaulo/AES in São Paulo, had contracted substantial overseas loans before devaluation, and were consequently in financial difficulties. More serious problems arose with the design of rules for the bulk market MAE. This task was left to the market agents (generators and distribution utilities), and the result submitted to ANEEL and discussed in public hearings. The process started in 1998, but led to many conflicts and judicial disputes. In 2001 ANEEL had to intervene, replacing COEX by an executive board (Conselho do Mercado Atacadista de Energia Elétrica, COMAE) without links to agents. This intervention was followed by others in the aftermath of the 2001 crisis; and in April 2002 Law 10433 restructured MAE. Administrador do Sistema do MAE (ASMAE) and COMAE were replaced by an Administrative Council and Superintendence answering to a MAE General Assembly, and regulated by ANEEL. Initial settlements were made in December 2002. Even so, only in 2004 the settlement backlog was finished.


581

Fig. 16.9. Submarkets of the interconnected system. Percentages refer to shares of the interconnected market in 2004. Source: ONS and CCEE.

Perhaps the most serious problem was lack of investment in generation expansion. Even though there was some increase in investment relative to the mid-1990 trough (Fig. 16.7), it was insufficient to avert the 2001 crisis, covered in the next section. To compensate for it, the transmission arrangements were a success in the sense of attracting investment and avoiding strategic behavior by integrated companies, although unbundling was only partial. This resulted from the conjunction of three factors: independent operation by ONS, mandatory planning by CCPE (upon ONS studies) and competitive bidding for expansion projects with fixed payment for each individual project, upon availability. Competition in grid expansion has been lively since remuneration (under price cap contracts) is guaranteed, while centralized planning and operation avoid potential manipulation (while perhaps inserting central planner biases) and bidding improves efficiency. 16.3. The 2001 Power Crisis and Its Causes In February 2001, ONS reported that prospects for electricity rationing had become likely after 40 rainless days in the wet Summer season, a major reversal from previous forecasts. During March a series of government and industry meetings took place to discuss the situation, despite official denials of the possibility of rationing. By the end of April this had ceased to be a taboo word, and several proposals to cope with the crisis were openly discussed.

582


R$/MWh (IPCA 1995 ⫽ Global average, for comparison)

300 250 200

Residential Industrial Commercial Global average IPCA

150 100 50 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year

Fig. 16.10. Average electricity tariffs and consumer price index (1995–2005). Source: ANEEL, IPEA. (Note: IPCA – Ample consumer price index is averaged from January to December except for 2005, where it is averaged January to April, same as tariffs).

On May 15, the federal government created13 a special body to deal with the emergency, called “Chamber for Management of the Energy Crisis” (Câmara de Gestão da Crise de Energia, or GCE for short). Emergency measures were introduced on May 16 by GCE, suspending supply to new loads (save residential and rural), load increases, nightly sports events, and for advertisement and ornamental purposes; public lighting was reduced by 35%. On May 18 further measures were taken, aiming at a 20% reduction in electricity consumption starting from June 1; consumers were expected to achieve given goals of reduction for each category. There were tariff bonuses for those reaching set goals, sanctions against laggards, and a large information campaign was launched. Price in the spot market was held at the ceiling of R $684 (roughly US $280, see Fig. 16.1), but free auctions for surplus energy (i.e. energy unsold in MAE or energy made available by large consumers) were held, fetching prices below the ceiling for small amounts. The only region free from rationing was the South, thanks to a favorable hydrology, because transmission constraints impeded additional energy exchange with the rationed Southeast. One month into the rationing, government goals had been reached for the Southeast and the North; but contingency plans were announced with tougher measures like power cuts, if necessary. Substantial governance changes were made in MAE, as already mentioned, to try to make it work. Besides this, emergency contracts were signed with merchant plants, especially for the Northeast. Power cuts did not materialize, and with the new rainy season and a sizable reduction in consumption (about 20% relative to forecasts in most regions, reaching 26% in some places), rationing was officially ended on the March 1, 2002. This, in a very sketchy description, was the sequence of events of the 2001 Brazilian electricity crisis.

13

GCE started working at once, although its creation only became official on May 29, when Provisional Measure 2198-3 was formally approved.


583

What caused the crisis, what should be done to avoid a repetition? A task group headed by Jerson Kelman (at the time Chief Regulator for the National Water Agency – Agência Nacional de Águas (ANA), and now heading ANEEL) was set up early on by GCE to investigate those issues, having produced a report that became known as the “Kelman Report” (Kelman, 2001). Its chief conclusions were: 1. Despite unfavorable hydrology (actual inflow was only 75% of historical average), if planned generation projects had been executed on time the crisis would have been averted since demand growth was consistent with existing plans. 2. Construction delays in ongoing projects (22 TWh), and plants not yet implemented although programmed (40 TWh), totalled 62 TWh not generated, which would have been more than enough to maintain safe levels. 3. Assured energies used in the initial contracts had been magnified, giving the government a false sense of security and no incentive for prospective investors. 4. Federal generators, though exposed to risks because of the upward bias of assured energy and willing to invest, were constrained by macroeconomic policy goals (to guarantee fiscal surplus goals, federal expenses were cut) and apparently did not or could not take the issue to National Budget discussions. 5. Several attempts had been made to induce investment in thermal generation (including the Priority Thermal Plant Programme – PPT, involving 49 plants), but without success. 6. The main culprit was ineffective government management and lack of communication between governmental agencies and departments – particularly between ONS, ANEEL, MME and the Presidency of the Republic. 7. Institutional causes were lack of clear responsibility for overseeing energy policy, lack of clear and stable rules, lack of faith in contracts (from the MAE failure in reaching agreement on settlements), incomplete and inadequate legislation. The report offered suggestions to avoid future crises while dealing with the problems listed above. The Kelman Report was a serious and comprehensive document, and justly regarded by all concerned as an accurate technical diagnosis. Nevertheless, from an economic viewpoint it leaves a few ends open; and to brand government inefficiency and misinformation as main culprit appears a little bit like a catch-all: while it is an appropriate description for overblown assured energy, and for constraints to investment by federal generators,14 on the other hand lack of investment in thermal plants, despite several incentive programs, points to more basic design flaws (Araújo, 2001). The point is that, in a large hydro-dominated system like the Brazilian one, under normal conditions economic dispatching will only dispatch a thermal plant for a small fraction of the time (see Fig. 16.1: since the end of rationing, spot prices in the Southeast/Center-West submarket remained around US $6/MWh most of the time, and only occasionally reached US $18/MWh).15 Thus, under competitive conditions of thermal against hydro when the latter dominates, a thermal plant is an excessively risky investment. This is shown by the lack of power purchase agreements for thermal plants, despite incentive programs. In other words, in order to have thermal plant investment in such a system it would be necessary to use extra-market mechanisms (losing economic efficiency or

14

There is anecdotal evidence that the finance minister rejected investment requests from Eletrobrás as unworthy of consideration. 15 I profit to thank Albert Melo and Roberto Caldas, colleagues at CEPEL who helped with the MAE price series.

584


transferring revenues), at least until the generation mix would allow proper working of the market. The 2001 crisis had impacts on the economy and on the ESI and spilled on Brazilian politics – it was a factor in the 2002 Presidential election, being presented by the opposition as the result of mismanagement by the outgoing administration (Californian readers may find this familiar – see chapters by Joskow and Sweeney in this book). One point fiercely criticized (Sauer et al., 2003) was the contracts made with merchant plants, which were hardly if at all dispatched but which required a steady payment for their availability even though it had ceased to be necessary. Leaving this aside, one outstanding consequence was stagnant GDP (⫹1.3%) and a drop in electricity consumption (minus 6.6%) in 2001 relative to 2000 (BEN, 2004), which becomes more significant when taking into account that it occurred from June to December only. Between March and July 2001, consumption dropped 25% (Fig. 16.11). Even more significantly, residential per capita consumption dropped 13% and stayed at the new level in subsequent years, while self-production increased from 7.5% to 10.5% of consumption. This represented a discontinuity in electricity consumption trends, which significantly reduced the revenues of distribution companies and further aggravated their financial situation. To compensate for utility losses due to lower consumption, GCE and the federal government created a special levy on electricity tariffs; this was resented by consumers as adding insult to injury – not only were they required to curtail consumption, but also had to pay for consumption foregone. The crisis also had consequences for sector structure and operation. GCE was ended in February 2002, but Câmara de Gestão do Setor Elétrico (Chamber for Electric Sector Management or CGSE) was created in its stead, to fix what were perceived as significant failings and to monitor events. The governance of MAE was thoroughly reorganized, as

50000 45000

Average (MW)

40000 SE/CW South S/SE/CW Northeast North N/NE Nat.Int. System*

35000 30000 25000 20000 15000 10000 5000

January 2000 April 2000 July 2000 October 2000 January 2001 April 2001 July 2001 October 2001 January 2002 April 2002 July 2002 October 2002 January 2003 April 2003 July 2003 October 2003 January 2004 April 2004 July 2004 October 2004

0

Fig. 16.11. Recent evolution of electricity consumption. Monthly average demand of interconnected system (2000–2004). Source: ONS, Carga Própria de Energia SIN (2000–2004).


585

mentioned in Section 16.2. Besides, economic dispatching was altered to introduce “risk aversion curves”16 in optimization algorithms to signal need for thermal dispatching before the reservoir situation reached critical levels. Another consequence was that divestment plans for federal generators were postponed sine die, in view of the hurdles mentioned above and the existence of more urgent tasks like keeping the lights on.

16.4. The 2002 Picture: Unfinished or Warped Reform? The ESI which emerged from the 2001 rationing was scarred in many points. Distribution companies that were indebted in foreign currency were badly affected, with lasting results. Several distribution utilities in the North and in the Northeast, one that had been divested and others belonging to state governments, were forced onto Eletrobrás (which also had to bear their financial liabilities) upon insolvency. Generators were also financially uncomfortable, since they were losing initial contracts (less 25% each year, to be completely ended in 2005) and were exposed to very low spot prices (Fig. 16.1), since the post-rationing market lacked demand (from a stagnant economy, increased self-generation and reduced per capita residential consumption) and supply was plentiful thanks to a series of wet years – which also discouraged potential buyers from entering into bilateral contracts. Electricity tariffs, on the other hand, grew faster than consumer price index (Fig. 16.10) and frustrated consumers. Investment in thermal plants remained an unresolved issue. Free contracting was not showing up. An official document (the above-mentioned Kelman report) pointed out a chaplet of inefficiencies. It would appear that no one was happy with the reform outcome. Nevertheless, this would not be totally fair. At least in two respects the Cardoso reform presented positive results. First was the setting up of a regulatory agency that worked transparently and endeavoured to deal seriously with the issues set before it. Despite a mission far broader than what one would expect from a regulator, and despite the pitfalls of inexperience, the accomplishment of almost 7 years of ANEEL is mostly positive. This result is all the more impressive as its predecessor DNAEE had withered to little more than a shadow, crushed between the large utilities and the finance ministers. This is not to say that all is well. Far from it, as some recent tariff reviews have shown17 (Araújo and Oliveira, 2005a). Nevertheless, for the first time the Brazilian ESI has a regulatory agency that works with a high degree of autonomy and openness, trying to protect consumers and create competitive pressures for efficiency. The mission breadth has been corrected in the latest reform proposals, and its operational autonomy appears established as witness the nomination of respected names for its board. Second, the institutional set-up for the transmission system proved to be viable and able to stimulate investments while maintaining incentives for efficiency. Contrary to concessions for hydro power generation and for distribution utilities, which were onerous and went to the highest bidder, 30-year concessions for transmission line projects went to the

16

To this effect, curves (called “guide curves” or “risk aversion curves”) were constructed for secure reservoir operation. When reservoir levels fall below the security levels, thermal plants are dispatched even if they are not the cheapest option. 17 ANEEL was criticized for using a “reference company” for tariff reviews, eschewing a thorough analysis of information available. It also had to make provisional reviews only at the appointed dates, since lack of a clear time schedule led to late starts and forced the use of preliminary estimates for important parameters.

586


bidder offering the least service charge, under a reverse auction scheme. The low risk attracted investment, and most auctions to date have been successful, with participation of both national and foreign firms. Since projects are defined by a central planner (CCPE), distortions such as those pointed out by Wu et al. (1996) do not appear, while the possibility of central planning biases is tempered by the fact that the main players, public and private, have a say in the technical committees of CCPE (Portaria 150, MME, May 10, 1999).

16.5. Reform of the Reform (2003–2004): New Institutional Arrangements Under Lula The Workers’ Party had long fought the Cardoso privatizations, opposing many points of the reform carried out by his administration, and studies were prepared for the energy sector during the 2002 campaign.18 Upon Lula’s team taking office, Energy Minister Dilma Rousseff carried out a government study which arrived at proposals more pragmatic in form and content, through lengthy negotiation. Although former criticisms were retained, the new administration had to mould a viable, working electricity industry from what it found. This meant reaching an agreement with all existing players, and thus a compromise between party doctrine and investor interests. After many months of debate in and out of Congress, legislation mustering a fair amount of consensus (and a large number of amendments) was finally passed on March 15, 2004. Some outstanding differences between the new and the old arrangement may be briefly described. First, companies in the Eletrobrás Group were taken off the privatization list (this did not apply to the Distribution utilities which Eletrobrás had to take on for insolvency), and Eletrosul was allowed to invest in generation (it had retained only transmission assets under the previous administration). This could be read as de jure recognition of the de facto difficulty in selling these generators, but it changed their status. Second, the federal government reclaimed functions of planning and policy-making. Law 10847 authorized and Decree 5184 (August 2004) created a public enterprise named Energy Research Enterprise (EPE) charged with planning studies and answering to the Ministry of Energy. A third change was the creation of a Monitoring Committee for the Power Sector (CMSE) instead of CGSE, with a distinct composition (four representatives of MME, plus chief officers of ANEEL, ONS, EPE, CCEE, Câmara de Comercialização de Energia Elétrica, and ANP, Agência Nacional do Petróleo, the oil and gas regulator), and charged with permanent monitoring and evaluation of security and quality of electricity supply. However, the changes most hotly debated refer to the structure and operation of the ESI. MAE was replaced by the Chamber for Electricity Trading (CCEE), although for now MAE operates following the convention for electricity trading (also CCEE) until the new body is formally set up, under ANEEL supervision. CCEE is responsible for contract administration, settlements in the spot market, and conducting energy auctions. Every transaction has to be registered with CCEE, but there are now two contract environments: a regulated environment (Ambiente de Contratação Regulada, ACR) and a free environment (Ambiente de Contratação Livre, ACL). While the free environment refers to the usual bilateral contracts, the regulated environment represents a significant break. Briefly, ACR contracts are made

18

See Rosa et al. (2001), Tolmasquim et al. (2001), Tolmasquim and Campos (2002), Sauer et al. (2003).


587

through auctions conducted by CCEE, in which distribution companies are required to participate (distribution companies must be 100% covered by contracts at all times). The ACR aims at correcting one of the failings of the Cardoso reform: the lack of voluntary long-term contracts, which did not materialize as expected. There are basically three kinds of auctions: 1. Auctions for long-term contracts of available energy from operating plants (“existing energy”). 2. Auctions for adjustment contracts in existing energy 1 year ahead of need. 3. Auctions for long-term contracts of new plants 5 or 3 years ahead of need. One stated purpose of this set-up is to avoid self-dealing between generation and distribution by a single owner. Another concern is that sellers must guarantee their supply; in the case of thermal plants, they must guarantee that fuel will be available when needed. Auctions are conducted in the following way: distribution companies announce to CCEE their requirements 60 days ahead of auction date; generators compete for blocks of energy, in the case of existing energy auctions, and for plant projects in a list made public at the auction (prepared by EPE and MME, but to which prospective generators can add projects of their choosing); bidding follows a reverse auction, that is to say the bidder offering the least energy price wins the block or the project. One singular aspect of the contract arrangement concerns how it is done. Every bid won joins one generator to all distribution companies through a large number of bilateral contracts, proportionally to their demand in the auction. The accounting and registering of the contracts is the responsibility of CCEE. This set-up attempts to avoid strategic behavior against single-buyer schemes, with which it has points in common, by spreading out risks among all buyers (which also reduces risks to sellers). Another effort to reduce investment risk for prospective bidders in new energy auctions is that every project presented at such an auction should have a preliminary environmental licence. In effect, every power plant in Brazil is required to have a clean “bill of health” from the Ministry of the Environment, and previously prospective investors had to incur all the risks of environmental licensing. This had been a deterrent in the previous scheme, and it is hoped that this idea may solve the problem. So far, there have been two auctions of existing energy, but the first auction for “new” energy is yet to be carried out, due to problems with environmental pre-licensing of the projects to be auctioned. 16.6. The Present: Prospects and Uncertainties Have the new arrangements finally addressed the many challenging aspects of the Brazilian ESI reform? If the issue is put out of context like this, the answer should probably be no. One can imagine a future scenario, 20 years from now, where a generation mix with much greater thermal power participation will allow schemes closer to, say, a Pennsylvania–New Jersey–Maryland (PJM) or a Nord Pool operation, described in other chapters of this book. On the other hand, it is misleading and even meaningless to think of an electricity reform in terms of an optimal, “One size fits all” solution. History does matter, and the diversity of reform schemes around the world should make one pause (Cf. e.g. Newbery (2002), chapters 6 and 9). Accordingly, it appears more meaningful to inquire whether the present arrangements are viable for the present system; if it has incentives or disincentives to efficiency and investment, what impacts it will have on prices, what hurdles it will have to face, what interests are present. We shall try to examine these briefly.

588


Two features are apparent in the new arrangements: 1. More centralization of decisions. 2. More weight for government policy than the Cardoso arrangements, but much less than the pre-reform situation. The creation of planning organisms is seen by most observers and market participants as a good thing, especially after the 2001 crisis called attention to this weakness of the Cardoso administration. The centralization of decision-making, however, has both disadvantages and bonuses; the former are well known, but among the latter there is a possible way out of the thermal power conundrum. In fact, one of the purposes of the new contractual set-up is that auctions contemplate thermal power expansion projects, which would have long-term contracts. In this way, physical dispatching might be disconnected from financial contracts. In periods of surplus water, thermal generators could buy cheap energy in the spot market and supply it to the system to honor contracts, even if they are not dispatched. Another possibility, which appears likely to be adopted, is that thermal generators will have long-term capacity contracts from the auction, being paid for their availability (although they would still be required to guarantee operation if dispatched). The requirement that distribution utilities have 100% coverage for their market through contracts19 is reminiscent of the PJM capacity requirement; the difference, of course, is that in a hydro-based system energy is more restrictive than capacity. On the other hand, the requirement that DISTCOS have to forecast their market with a 5-year lead roused many objections; in the end, adjustment auctions with 1-year anticipation for existing energy were deemed satisfactory.20 The new set-up is quite complex and centralized; but the Cardoso reform had already recognized the need for central hydro power operation, which leads to long periods of very cheap energy with few spikes far apart (excepting the 2001 crisis, which was the result of mismanagement as pointed out by the Kelman report) – see Figure 16.1. On the other hand, there is scope for competition along three main routes: free bilateral contracts in the ACL, spot market and competitive bidding for regulated bilateral contracts, whether for expansion projects or for existent free energy. Objectively viewed, there appear to be incentives for efficiency. Whether the downsides of complexity and centralization will prevail upon the good sides of the new arrangements depends very much on attention to details. One thing is certain: with more than 600 players (see footnote (c) of Table 16.5) in the market, both public and private, there is no going back to the old ways. The touchstone of success or failure of the new model will hinge upon the results of auctions for generation expansion, for which the two past auctions for existing energy are a poor estimator. As mentioned before, these have taken place in a context where existing generators had no contracts for a large part of their generation. This had to be sold at very low prices at the spot market, since few buyers were willing to enter into long-term contracts when supply was plentiful. As a result, prices reached in the first auction fell below expectations, both of outside observers and of government officials. Even so, generators who got the

19

Own generation has to follow the CCEE route to avoid self-dealing, which had occurred in a few awkward instances in the previous regime. 20 In the first version, adjustment auctions existed only for 3-year forecasts, to be satisfied by new thermal plants; the 1-year, existing energy auctions were added after negotiation.


589

cheapest contracts felt they were better off than before. The second auction contained less surprises, perhaps because sellers had more experience and were financially less strangled. The coming auction for new projects will contain plants that should start operation in 2009–2010, when they will be badly needed. In principle, contract prices should adequately remunerate prospective investors under these conditions. Nevertheless, two hurdles remain that have little to do with the arrangements themselves, but which are vital for a viable Brazilian ESI. One hurdle concerns environmental licences. After a relatively late awakening of Brazilian politics to environmental issues, there is now a series of environmental requirements that power projects have to satisfy, after examination by the Brazilian Institute for the Environment (IBAMA), of the Ministry of Environment. Both power plants and transmission lines have faced trouble obtaining environmental licenses, but hydro power plants are subjected to especially rigorous scrutiny in view of past sins. However, it is felt in the power sector that procedures are excessively slow and ad hoc, which might jeopardize investment plans for hydro power plants. Given a very significant hydro power economic potential remaining to be exploited, this would mean renouncing efficiency and scarce resources. The problem lies less with rejection versus approval than with the pace of the analysis process. The seriousness of the issue may be seen from the fact that, of 17 projects to be submitted to the first expansion auction, only one has already had its preliminary licence approved by IBAMA. The solution requires inter-ministerial agreement on adequate and timely procedures, but until the moment of writing (September 7, 2005) results have been limited despite the good will of both ministers. The problem seems to be understaffing of IBAMA and lack of clear guidelines. Another issue concerns gas-fired plants. Even if hurdles to hydro power plants are removed, there is a growing need for thermal generation for two reasons. First, although significant, the remaining economic hydro power potential (roughly 50 GW, taking into account environmental constraints) will not be enough if a planning horizon of 15–20 years ahead is considered. Second, for environmental reasons (all the best sites have already been developed), remaining plants will have much smaller reservoirs and thus little capacity for regulation. To enhance their generation potential, thermal plants are necessary. Thermal generation in Brazil means to a significant extent gas-fired plants, at least in the short term. And herein lies the problem: although Brazil has enough potential gas supply from domestic reserves and from neighbors, its gas industry is woefully immature. All over the world, development of the natural gas industry has been based on long-term contracts because of the considerable investment in pipelines and distribution networks. This normally means take-or-pay contracts between buyers and sellers. In a mature gas market, where most network investment has been depreciated and alternate users exist, a seller might be content with lower take-or-pay specifications and a buyer might have a secondary market to sell unused gas. In the Brazilian case, it is necessary both to create a gas infrastructure and to develop a secondary gas market side by side (or alternatively, dispatch thermal plants irrespectively of cost and abundance of water, but this policy cannot claim efficiency). This is far from being simple, and the Ministry of Energy has struggled with the problem for some time now. A “Gas Law” is to be presented this year (2005), aimed at solving the issue. After it comes out, it should be possible to evaluate the prospects for solution of this dilemma. Meanwhile, gas plants are being encouraged to run as dual-fuel, burning gas or diesel according to availability. This is not the best of worlds, but in the short term it would not be excessively costly, since thermal plants are seldom dispatched. Other issues concern market players. Although more than 600 are active, only a handful concentrate a large part of the market. In particular, the Eletrobrás Group concentrates

590

Electricity Market Reform Table 16.6. Distribution market shares in 2004 (%). Federal utilities* State and municipal utilities Investor-owned utilities Total

9.0 24.3 66.7 100.0

*CHESF and Eletronorte-controlled DISTCOS, as well as utilities temporarily under federal intervention. Source: Compiled from ANEEL “Participação dos Agentes Econômicos no Mercado de Energia Elétrica”.

roughly 53% of generation capacity of Brazil in public service, although IPPs now boast 26 GW (cf. Table 16.1); CHESF has 38% of the N/NE market and Eletronorte 32%, while Furnas has 15% of the S/SE/CW, Eletronuclear 3% and CGTEE 0.8% of the same (ANEEL, 2005).21 It is clear that no significant competition can exist in the N/NE markets at the moment; but, given present arrangements, there is not much scope for market power except at auctions. Therefore, the government has strictly watched the auctions to avoid any collusion between group firms, and succeeded to the point that some analysts darkly suggested the results discouraged prospective investors; as a matter of fact, the lowest prices in the existing energy auctions were bid by CHESF. While federal utilities dominate the generation for public service, the converse is true for distribution: investor-owned utilities have two-thirds of the market (Table 16.6). Furthermore, free (large) consumers now represent 18% of total electricity consumption, nearly one fourth of the distribution market (CCEE data). These consumers operate in the free contract environment (ACL), which has thus become significant. A player with large prospects, especially in thermal generation, is Petrobrás, the big federally owned oil company. Petrobrás is large by any standards, with yearly net revenues of 30–40 billion dollars, profits and investments of the order of 6–9 billions. It has recently become an energy enterprise, declaring interest in power generation, and many existing thermal plants are owned wholly or in part by Petrobrás. It is still a bit unclear whether it will find in its interest to develop a gas infrastructure, and actions by other actors have not been very significant as yet, barring the 2001 merchant plants whose contracts are being renegotiated or sold. Other sources do not have significant interest groups, except in a localized sense. However, they should not be written off: in the long term, 20 or 30 years from now, both coal and nuclear will be needed, as hydro and gas approach their limits. A word should be said about IPPs and self-producers, who constitute most of the 600 active agents. Although small, they are connected to a large segment of Brazilian industry and are well organized into several associations; they have thus a non-negligible pressure power, besides representing together more than a fifth of electricity consumption. Finally, in transmission and distribution most problems seem to have been ironed out, and there seems to be no obvious trouble to face in these segments barring environmental

21

The shares of the national market are respectively 12% for CHESF, 10% for Eletronorte, 11% for Furnas, 2% for Eletronuclear, and 0.5% for CGTEE. This refers to the effective generation in 2004, and Itaipu is not included (ANEEL, Dados de Geração, http://www.aneel.gov.br/arquivos/PDF/4_ TRIM_2004_INTERNET_GERACAO.pdf).


591

licensing. The existence of frequent auctions in the free contract environment and the growth of free consumption, on the other hand, signal dynamism and appear to indicate that little by little the Brazilian electricity players are beginning to learn how to deal with the significant hurdles that exist in this system. In conclusion, there are reasons for a cautious optimism. However, two problems must be solved for the Brazilian reform to be successful: how to design guidelines and operational procedures for environmental licensing that do not cripple investments in generation and transmission expansion, and how to solve the natural gas dilemma. Neither seems to be of easy solution, and creativity as well as good will and much hard work are required. Acknowledgments I would like to thank the many colleagues and friends who collaborated with discussions, criticism and information. Special thanks go to Ricardo Raineri, Perry Sioshansi, Nelson Martins, Helder Pinto, Adilson Oliveira, Jerson Kelman, Albert Melo, and Roberto Caldas. I would also like to thank Paulo Cesar Menezes, from CEPEL’S Geographical Information Systems Lab, for updating and converting Figures 16.4 & 16.9 to high definition images. Last but far from least, I am indebted to former Eletrobrás President Altino Ventura Filho, who not only read the manuscript carefully but took time to discuss it and straighten a number of imprecise points. Of course, I alone am responsible for the remaining errors and omissions of the text. References ANEEL, Agência Nacional de Energia Elétrica, http://www.aneel.gov.br Araujo, H.P.dE M. (1979). O Setor de Energia Elétrica e a Evolução Recente do Capitalismo no Brasil, Rio de Janeiro, COPPE/UFRJ, Tese de Mestrado, 1979. Araújo, J.L. (2001). Investment in the Brazilian ESI: What went wrong? What should be done?, Workshop on Competition and Regulation: The Energy sector in Brazil and UK/EU. Oxford, St. Anne’s College, 4–5 June. Araújo, J.L. (2002). A Questão do Investimento no Setor Elétrico Brasileiro: Reforma e Crise. Revista Nova Economia, 11(1), pp. 77–95. Araújo, J.L. (2005). Brasil em desenvolvimento: expansão da infra-estrutura. In A.C. Castro, Antonio Licha, Helder Queiroz Pinto Jr and João Sabóia (orgs). Brasil em desenvolvimento: economia, tecnologia e competitividade, Civilização Brasileira, Rio de Janeiro, ISBN 8520006809. Araújo, J.L. and Besnosik, R. (1992). Regulation, Institutional Structure and the Performance of the Brazilian Electricity Sector. COPED Report, Rio de Janeiro. Araújo, J.L. and Losekann, L.D. (2001). Atualização da Indústria Elétrica Brasileira. Report to PREVI Pension Fund, Rio de Janeiro, August. Araújo, J.L. and Oliveira, A. (1996). De-Regulation (and Re-Regulation) of the Brazilian Energy Industries: Recent Trends and Long Term Prospects. USAEE/IAEE Annual Meeting, Boston, October. Araújo, J.L. and Oliveira, A. (2005). Diálogos da Energia: Reflexões sobre a última década 1994–2004, Editora 7 Letras, 2005, Rio de Janeiro, ISBN 85757718. Araújo, J.L. and Oliveira, A. (2005a). Questões teóricas e práticas da regulação por preço teto, in Araújo and Oliveira (2005). Araújo, J.L., Oliveira, A. and Zendron, P. (2000). Power market issues in the Brazilian ESI reform. 23rd Annual IAEE International Conference, June 7–10, 2000, Sydney. Araújo, J.L. and Pinto Jr., H. (1998). Changes in the Structure, Regulation and Financing of the Brazilian Electricity Supply Industry. IAEE Annual Meeting, Quebec, May. ASMAE (2000–2001). “Regras do Mercado”, “Procedimentos do Mercado” and other documents. BEN 2004 – Balanço Energético Nacional 2004, MME, Brasília 2005. CANAL ENERGIA, http://www.canalenergia.com.br

592


CCEE, Câmara de Comercialização do Setor Elétrico, http://www.ccee.org.br (also http://www. mae.org.br) Centro Da Memória Da Eletricidade No Brasil (1988). Panorama do Setor de Energia Elétrica no Brasil. Memória da Eletricidade, Rio de Janeiro, 1988. Daher, M. (2005). Operador Nacional do Sistema Elétrico: Papel, Funções e Características. Presented at IE/UFRJ on 25 May 2005. Dutra, W. and Salles, V.C. Padrão de Financiamento em Empresas Estatais. M.Sc. thesis, COPPE/UFRJ, Rio de Janeiro, 1975. Esmeraldo, P.C. (2004). O Planejamento da Transmissão: Uma Visão do CCPE, Workshop CEPEL 30 Anos, Rio de Janeiro, November 2004. Hunt, D. (1989). Economic Theories of Development. Harvester Wheatsheaf, New York. IPEA, Instituto de Pesquisas em Economia Aplicada, http://www.ipea.gov.br Kelman, J. (Co-ordinator) (2001). Relatório sobre o desequilíbrio entre oferta e demanda de energia elétrica. Comissão de Análise do Sistema Hidrotérmico de Energia Elétrica, Brasília 31/07/2001. Losekann, L.D. (2003). Reestruturação do Setor Elétrico Brasileiro: Coordenação e Concorrência. D.Economics Thesis, Instituto de Economia da Universidade Federal do Rio de Janeiro, December. MME, Ministério de Minas e Energia, http://www.mme.gov.br Newbery, D. (2000). Privatization, Restructuring and Regulation of Network Industries. MIT Press, Cambridge (Massachusetts). Oliveira, A. (1996). Reforming the Brazilian Energy System: Challenges and Opportunities. Seminar of the Institute of the Americas, Rio de Janeiro, June. Oliveira, A. (1997). Reforma do setor elétrico: Que podemos aprender com a experiência alheia?, IE/UFRJ, mimeo, February. Oliveira, A. (1998). As Experiências Internacionais de Reestruturação. In de Oliveira and Pinto JR (1998) (Orgs.). Financiamento do Setor Elétrico Brasileiro: Inovações e Novo Modo de Organização Industrial. Garamond, Rio de Janeiro. Oliveira, A. and Losekann, L. (1999). O Novo Mercado Elétrico: Perspectivas para o Gás Natural. Relatório Gaspetro.{not referred} ONS, Operador Nacional do Sistema, http://www.ons.org.br Pinhel, A.C. (2000). Simulação de uma usina térmica a gás no novo contexto do setor elétrico – Análise Risco X Retorno. M.Sc. thesis, UFRJ, COPPE, Rio de Janeiro, December. Rosa, L.P., Tolmasquim, M., Oliveira, L.B., Sampaio, M.R., Cecchi, J.C. and Schaeffer, R. (2001). Um País em Leilão – Das Privatizações à Crise de Energia. Vol. 2, 1 ed. COPPE/UFRJ, Rio de Janeiro, 191 pp. Santos, M.F. de M. (1996). Os novos desafios do setor elétrico: Coordenação, integração e uso eficiente de energia. II Encontro de Estudos Estratégicos, October 14, 1996, Rio de Janeiro. Sauer, I.L., Rosa, L.P., D’Araujo, R.P., Carvalho, J.F., Terry, L.A., Prado, L.T.S. and Lopes, J.E.G. A Reconstrução do Setor Elétrico Brasileiro, Editora Paz e Terra/Editora UFMS, São Paulo – SP/Campo Grande – MS, 2003. Tolmasquim, M.T. and Campos, A.F. (2002). A Reforma do Setor Elétrico em Perspectiva. In IX Congresso Brasileiro de Energia, 2002, Rio de Janeiro. Anais do IX Congresso Brasileiro de Energia, Vol. I. COPPE/UFRJ, SBPE, Clube de Engenharia, Rio de Janeiro, pp. 454–459. Tolmasquim, M.T., Pires, J.C.L. and Rosa, L.P. (2001). New Strategies for Power Companies in Brazil. In Atle Midttun (Org.). European Energy Industry Business Strategies. Oxford, pp. 337–374. Wu, F., Varaiya, P., Spiller, P. and Oren, S. (1996). Folk theorems on transmission access: proofs and counterexamples. Journal of Regulatory Economics, 10(1), 5–25.


593

Appendix: Most Used Acronyms ACL (Ambiente de Contratação Livre) – Free contracting environment, accessible to generators, independent producers, traders and large consumers. Created by the Lula reform. ACR (Ambiente de Contratação Regulada) – Regulated contracting environment, mandatory for distribution companies. Created by the Lula reform. AEF (Alocação do Excedente Financeiro) – Mechanism for compensation of windfall losses and gains in trading between regions with different spot prices, due to initial contracts (assigned by the government). ANA (Agência Nacional de Águas) – Federal Regulatory Agency for Water Issues. ANEEL (Agência Nacional de Energia Elétrica) – Federal Electricity Regulatory Agency, replaced DNAEE in 1997. ANP (Agência Nacional do Petróleo) – Federal Regulatory Agency for Hydrocarbons, created in 1996. ASMAE (Administrador do Sistema do MAE) – Technical body of MAE, replaced in 2002 by a Superintendence answering to the General Assembly of MAE. CCEE (Câmara de Comercialização de Energia Elétrica) – Chamber created by the Lula reform and charged with overseeing and registering power trading, in lieu of MAE. CCON (Comitê Coordenador de Operações do Norte-Nordeste) – Former Committee for co-ordinated power system operation in Northern and Northeastern Brazil, headed by Eletrobrás. Its functions were absorbed by ONS. CCPE (Comitê Coordenador do Planejamento da Expansão Elétrica) – Co-ordinating committee for electric power expansion planning, created in 1999. The planning is indicative for generation, and mandatory for transmission. CEEE (Companhia Estadual de Energia Elétrica S.A.) – Integrated utility controlled by the State of Rio Grande do Sul, in Southern Brazil. CEMIG (Companhia Energética de Minas Gerais S.A.) – Integrated utility controlled by the State of Minas Gerais, in the Southeast. CESP (Companhia Energética de São Paulo S.A.) – Generator controlled by the State of São Paulo; part of its assets was divested prior to 2000, but the company remains in the state government’s hands. CGSE (Câmara de Gestão do Setor Elétrico) – Permanent body charged with overseeing the power sector, created in 2002. CMSE replaced it in 2004. CGTEE (Companhia de Geração Térmica de Energia Elétrica S.A.) – Thermal generator (mostly coal) based in Rio Grande do Sul, controlled by Eletrobrás. CHESF (Companhia Hidrelétrica do São Francisco) – Generator in the Northeast, controlled by Eletrobrás. CMSE (Comitê de Monitoramento do Sistema Elétrico) – Committee charged with monitoring and assessing continuity and security of the power supply system, created by the Lula reform. CNPE (Conselho Nacional de Planejamento Energético) – High-level inter-ministerial body for co-ordinating energy planning, created in 2000. COEX (Comitê Executivo do MAE) – Executive Committee of MAE, replaced in 2001 by COMAE, due to governance problems. COMAE (Conselho do Mercado Atacadista de Energia Elétrica) – Executive board of MAE, replacing COEX, and replaced by its turn in 2002 by an Administrative Council. COPEL (Companhia Paranaense de Energia S.A.) – Integrated utility controlled by the State of Paraná, in the South. CTEEP (Companhia de Transmissão de Energia Elétrica Paulista) – Transmission company originating from the unbundling of old CESP. Now named Transmissão Paulista, it is being privatized by the state government to pay for debts of the new CESP.

594


DISTCO – Distribution Company. DNAEE (Departamento Nacional de Águas e Energia Elétrica) – Electricity regulatory agency from 1965 to 1997, when ANEEL replaced it. ELETROBRÁS (Centrais Elétricas Brasileiras S.A.) – State-controlled holding company for the Eletrobrás Group. ELETRONORTE (Centrais Elétricas do Norte do Brasil S.A.) – Generator in Northern Brazil, controlled by Eletrobrás. ELETRONUCLEAR (Eletrobrás Termonuclear S.A.) – Nuclear generator in the Southeast, controlled by Eletrobrás. ELETROSUL (Eletrosul Centrais Elétricas S.A.) – Company in Southern Brazil controlled by Eletrobrás, most active in transmission (its generating assets were bought by Tractebel), now returning to generation. EPE (Empresa de Pesquisa Energética) – Despite the name (Energy Research Enterprise), it is a public organism charged with the execution of technical planning studies to assist CNPE and CCPE decisions. ESCELSA (Espírito Santo Centrais Elétricas S.A.) – Distribution utility in the State of Espírito Santo, in the Southeast, privatized in 1995. ESI – Electricity Supply System. FURNAS (Furnas Centrais Elétricas S.A.) – Generator in the Southeast, controlled by Eletrobrás. GCE (Câmara de Gestão da Crise de Energia Elétrica) – Emergency task force created in 2001 to deal with the power crisis. CGSE replaced it in 2002. GCOI (Grupo de Coordenação da Operação Interligada) – Former co-ordinating group for interconnected power system operation, headed by Eletrobrás; its functions were taken over by ONS. GCPS (Grupo Coordenador do Planejamento do Sistema) – Former co-ordinating group for system expansion planning, headed by Eletrobrás, has been replaced by CCPE since 2000. GENCO – Generation Company. IBAMA (Instituto Brasileiro do Meio Ambiente) – Chief technical organ of the Ministry of Environment, responsible for environmental impact assessments of energy projects. IGPM (Índice Geral de Preços do Mercado) – Bulk price index used to readjust tariffs in most electricity concession contracts in Brazil. Its original function was to readjust banking contracts. IPCA (Índice de Preços ao Consumidor Amplo) – Consumer price index covering households earning from 1 to 40 minimum wages in nine metropolitan areas plus the Federal Capital. LIGHT (Light Serviços de Eletricidade S.A.) – Distribution utility serving the metropolitan area of Rio de Janeiro, since 1996 controlled by Electricité de France. MAE (Mercado Atacadista de Eletricidade) – Organism for managing the bulk power market, replaced in 2004 by CCEE. MME (Ministério de Minas e Energia) – Ministry for Mines and Energy. MRE (Mecanismo de Realocação de Energia) – Mechanism for apportioning generation revenues among hydro generators. ONS (Operador Nacional do Sistema) – National System Operator, charged with the physical operation of the interconnected system. PETROBRÁS (Petróleos Brasileiros S.A.) – Federally controlled oil company. PSDB – Brazilian Social – Democratic Party. REVISE – Acronym for a task force created in the mid-80’s for Institutional Review of the power sector. TRANSCO – Transmission Company. VN (Valor Normativo) – ANEEL-defined prices for different generation technologies, to serve as criteria for cost pass through by distribution utilities to consumers.

Chapter 17 Understanding The Argentinean and Colombian Electricity Markets ISAAC DYNER,1 SANTIAGO ARANGO,2 ERIK R. LARSEN3 1

Energy Institute, Universidad Nacional de Colombia, Medellín, Colombia; 2Energy Institute, Universidad Nacional de Colombia, Medellín, Colombia and System Dynamics Group, University of Bergen, Bergen, Norway; 3Faculty of Economics, University of Italian Switzerland (USI), Lugano, Switzerland

Chapter Summary In this chapter, we will make comparisons between Argentina and Colombia to develop a better understanding of how the different models of deregulation have shaped the evolution of their electricity systems. Representing the different approaches to deregulation in South America, these two countries are both developing nations with large hydroelectricity generation bases, with sufficient similarities to make a useful comparison.

17.1. Introduction Argentina and Colombia were not the first countries to deregulate in South America. At present they can be seen to be among the most successful, although this might change in the near future. Both countries deregulated over 10 years ago and as such provide a period where analysis and comparison can begin to yield insight into the consequences of both the initial and later decisions on the evolution of the electricity system. While deregulation has already taken place in a number of countries and regions around the world it is interesting to compare countries in South America. In some dimensions they are very similar (e.g. technology, level of economic development, and culturally), while other aspects are quite different (e.g. institutional frameworks, market arrangements, and time of deregulation). As deregulation has taken place at various times during the last 20 years, it is fair to ask if the latecomers to deregulation have learned from the earlier experiences of neighboring countries. Other chapters of this book cover Chile and Brazil, two other important markets in Latin America. The changes in the electricity sector in South America have been actively promoted by international agencies such as the World Bank (1993) and the Inter-American Development Bank (IDB, 2000; Bacon and Besant-Jones, 2001), based on the need for private capital, among other drivers. The first country to deregulate in Latin America was Chile, followed by Argentina, which introduced a number of adjustments to the Chilean model; 595

596


Table 17.1. Current situation of the ESI in Latin-American and the Caribbean countries. Private property Vertical segmentation with incompatibilities

Argentina*** (1992) Bolivia** (1994) Guatemala** (1997) Panama** (1997) Chile** (1986) El Salvador** (1996) Peru** ()

Vertical integration allowed Private and State Property

Cuba Ecuador Jamaica Mexico Trinidad and Tobago Honduras

Costa Rica Granada Guyana Uruguay Venezuela

Exclusively State property

Surinam

Haiti Paraguay Autonomy for companies with regulation

Partial competition in generation

* Colombia** (1995) Brazil (1998) Nicaragua

Open market

Note: (year of liberalization) if applicable. *With or without vertical segmentation; **with weak horizontal participation; and ***with strong horizontal participation. Source: Adapted from OLADE et al. (2004).

other countries such as Ecuador, Peru, and Bolivia used a similar deregulation model. The adjustments to the Chilean model made in Argentina prevented some of the problems observed in Chile during the late 1990s, but not others that have emerged since the beginning of 2004. Despite some technological similarities with the Chilean electricity system, Colombia adopted the British model, with some adjustments to take account of the differences in the energy systems, in the mid-1990s. Table 17.1 show the date when the energy supply industry (ESI) was liberalized in the different Latin-American countries, and the current ownership status. In this chapter, we will make comparisons between Argentina and Colombia. They are both developing nations with a large hydroelectricity generation base, making a comparison meaningful. These two countries, broadly speaking, represent the two main approaches to deregulation undertaken in South America. This chapter is organized as follows. First, we present introductions to each of the countries and the evolution of their individual electricity system after deregulation. The second section compares and contrasts the two countries. We finish by drawing some general points in the conclusions. Figures 17.1 and 17.2 show the map and main cities of Argentina and Colombia, while the key economic, demographic, ownership, concentration, and technological indicators for the ESIs in the two countries are provided in Table 17.2. For each country we provide the main information describing the background for the reform, the chosen reforms, market structure, major changes that took place, and other

Understanding The Argentinean and Colombian Electricity Markets

597

Fig. 17.1. Map of Argentina.

significant events. To compare the evolution of the two countries, standard economic indicators including efficiency and quality are used (Newbery, 2001; Hunt, 2002; Stoft, 2002). For efficiency measures, we present not only price, supply, and demand evolution but also a measure of market concentration in terms of the Herfindahl Hirschman index

598


Fig. 17.2. Map of Colombia.

(HHI).1 Efficiency is reflected in relative market prices and depends largely on concentration; reliability is affected by volatility; and quality is measured by the frequency and intensity of interruptions of service (grid losses) and by service perception when data are available. In addition, this chapter explains specific events that have had a significant impact on the ESI in each country.

17.2. Argentina Argentina deregulated its electricity industry in 1993. This was prompted by a number of problems that the sector confronted during the late 1980s (Ente Nacional Regulador de la Electricidad, ENRE, 1999; ENDESUR, 2002; Pistonesi, 2002). As tariffs were used for

1

The HHI Index is usually used by regulators to measure market concentration (Herfindahl, 1950; Hirschman, 1964; Kwoka, 1985). It is calculated by squaring the market share of each firm competing in the market and then summing the resulting numbers. The US Department of Justice normally considers a market with an HHI less than 1000 as a competitive market, HHI between 1000 and 1800 as a moderately concentrated market, and an HHI of 1800 or greater as a highly concentrated market. Bacon and Besant-Jones (2001, p. 15) show how the HHI index can be understood “as an indicator of how much the price could be raised above the marginal cost of production where there is no regulation to control price”.


599

Table 17.2. Comparison of economic indicators and the electricity industries in Argentina and Colombia. Argentina

Colombia

Population (million) GDP 2002 (billion US$) Size (*1000 km2) GDP per capita (*1000 US$/person)

37.928 102.19 3.761 2.69

43.745 82.19 1.039 1.88

Installed capacity (MW) Hydro Thermal Others

8857 13,010 1018

8810 4366 0

Total

22,884

13,176

Technological composition (%) Hydro Thermal Others

39 57 4

67 33 0

Private sector participation (%) Generation Transmission Distribution

60 100 70

70 10 50

30 80 50

50 100 60

Market share of the three largest firms (%) Generation Transmission Distribution

World Bank (2003). Only the MEM, major subsystem with more than the 90% of the total electricity system in Argentina (Energía, 2002). ISA (2003) and Espinasa (2001).

anti-inflationary and income distribution purposes, state-owned utilities ran large operational losses, provided poor services to customers, maintained high non-technical losses, and lacked sufficient financial resources for capacity investments. These led to a serious crisis of electricity supply, with blackouts that amounted to losses of 1 million MWh in a single year. The deregulation of electricity was, furthermore, part of a wider process in which the country opened up to capital markets, privatization, and deregulation of public services in the late 1980s and early 1990s. Law 24065 of Argentina’s Congress, passed in 1992, established the foundation for the liberalization of the Argentinean ESI and created ENRE, the National Regulator of Electricity. ENRE supervises and regulates the ESI, which includes the control concessions, the prevention of anticompetitive behavior or entry barriers, the promotion of incentives for infrastructure investment, the resolution of conflicts, and the protection of the environment. Argentina’s electricity sector is, in reality, composed of two separate markets: the Main Electricity Market (the one that we are focusing on in this chapter) and the Wholesale Electricity Market of the Patagonia System. It is the responsibility of the regulator to implement clear market rules, thereby providing confidence to potential and actual investors, as well as making the system adaptable to

600


accommodate the necessary changes. The distribution and transmission of electricity are considered regulated monopolies that have to guarantee “free access” to producers and consumers. Generation is operated under market conditions by means of a cost-based pool. The wholesale market is organized in to two parts: a spot market and a bilateral contract market. The spot price is determined hourly as a result of the optimal dispatch based on short-term marginal costs; generators submit their cost and the system operator performs an optimal dispatch based on the short-term marginal cost (ENRE, 1999, 2002, 2003; CAMMESA, 2003a). Bilateral contracts are negotiated freely between the participants and the price is that agreed between the parties. Some regulated activities have been licensed: the high-voltage net from the central government to TRANSENER S.A., and electricity distribution from the provincial governments to regional utilities. The dispatch of the system is cost based, and price and generation are established through an optimization process where the price is set on a nodal basis. In addition to the dispatch price, generators are paid for making capacity available to the system, which has been fixed at 10 US$/MWh. This is paid to generators when they make energy available in the 90 hours defined by the regulator as the weekly peak-demand periods (Rudnick and Montero, 2002). Currently, Argentina’s electricity network is interconnected with Uruguay, Paraguay, and Brazil; and Argentina is planning to develop a bilateral market with Chile to take advantage of their potential electricity complementarities. On the supply side, the system was expanding up until 2002 at an average rate of 6% per year, not only in thermal capacity, but also in hydro capacity, reaching 22,884 MW of installed capacity for the main electricity market. Demand was also increasing until 2002, when there was a fall in demand of 2% as shown in Figure 17.3, resulting in a growing reserve margin and consequently a drop in prices. The reserve margin (calculated as the percentage difference between the installed capacity and the maximum demand, divided by maximum demand) has always been relatively high in Argentina; however, it increased further to around 60% in 2002 due to the combined effect of increases in the installed capacity and decreases in peak demand. The fall in peak demand can be explained by the major economic crisis in Argentina, when the Argentinean peso was devalued by about 300% in 2002. As we can observe in Figure 17.3, there

25,000

MW

20,000

Installed capacity

15,000 Peak demand 10,000

5000 1992

1996

2000 Time

2004

Fig. 17.3. Evolution of the Argentinean electricity market: installed capacity and peak demand.

601


has been no substantial increase in capacity since 2001, while the economic crisis has been overcome and demand is growing at almost the same rate as before the crisis, that is around 5% p.a. The rather large system margin of more than 60% has been reduced to around 45% in 2005. This is still a relatively large margin. However, electricity shortages started to occur in 2004 as a consequence of reductions of gas supplied to gas-fired plants and low water-inflows into hydroelectricity reservoirs. Taking into account the current growth in demand, and the construction times of new power plants, there might well be problems within the medium term. This indicates the need for revisions to the government-decided energy tariffs and price structures, as well as for reviews of regulation regarding investment incentives, including capacity charges. The Argentinean gas supply problem is not only affecting electricity supply to Argentina but also the supply of natural gas to Chile, as described in Chapter 3. Hydrology has an effect on price, although it is rather limited during normal years, while temperature largely explains the yearly cyclical behavior. Figure 17.4 shows this situation, where spot prices peak with falling temperatures, largely due to household heating. An increase in temperature implies a decrease in the demand, which leads to a decrease in prices, and vice versa. This dynamic relationship between temperature and spot prices in Argentina is reflected in a correlation coefficient of ⫺0.39 for the available data, which might be considered relatively high given that the demand is affected by many factors. As we can observe in Figure 17.4, the spot price has more than doubled in the last 3 years, a phenomenon that might indicate problems in the structure of the market, as well as the world market energy prices (imports are in US$).2 It should also be noted that the prices for electricity and gas paid by the domestic sector are fixed by the government. The prices have been fixed since the beginning of the economic crisis and have only seen minor adjustments since. The prices are relatively low compared with the spot market prices paid by the non-regulated sector (Haselip et al., 2005).

30

80

25 60 20 15 40 10 Monomial spot price 20 1998

2000

2002

2004

Time Fig. 17.4. Spot price in Argentina, compared with average temperature.

2

Note the price reported is in local currency: 1US$ ≈ 3 Argentinean pesos.

5 2006

Temperature (°C)

Argentinean peso/MWh

Average temperature

602


Argentinean peso/MWh

80

60

40

20 1992

1996

2000 Time

2004

Fig. 17.5. Monomial average annual price in Argentina.

Consumers have to pay for other services in addition to the price for electricity. These services are included as part of the tariff, such as ancillary services and transport (CAMMESA, 2002). A significant part of the demand, 37%, is contracted, and the rest is traded in the spot market; even so, this is a low percentage compared to most other deregulated electricity markets. However, the number of contracts for electricity has increased rapidly since the market started operations, from nine contracts signed in 1993 to more than two thousand in 2002 (CAMMESA, 2003a & b). Price reductions in the spot price (see Fig. 17.5) can be partly explained by increasing competition. The generators, which increased in number from 13 in 1992 to 44 in 2002, are competing in a market that trades around 2335 million Argentine pesos (in 2001) and where the largest five companies take about 43% of total sales (ENRE, 2002). According to 2001 data, 74% of the installed capacity is privately owned. There are 44 companies in the generation market, indicating significant competition (ENRE, 2002). The HHI was around 1500 for generation, 1250 for installed capacity, and 1400 for distribution, in 2001 (ENRE, 2002), a relatively low HHI index for all segments of the market. This indicates the right condition exists for competition in the market and that market power should be a minor problem. As we can observe in Figure 17.5, the general price fall that followed the economic crisis only lasted very briefly, and the price has now returned to the pre-crisis level. However, there has been a significant increase in competition in generation and supply to large customers. The Argentinean system defines two kinds of large customer: very large and large, where very large customers consume at a rate in excess of 2 MW/year, and the large customers at a rate between 0.1 and 2 MW/year. The number of customers in the very large group has been stable at about 350, and the number in the large group has increased from 207 in 1995 to almost two thousand in the year 2002, as the threshold has been reduced, thereby opening up a larger part of the market to competition. Argentina’s market has also experienced a reduction in total grid losses as can be seen in Figure 17.6. While there was a small increase immediately after deregulation, grid losses started dropping from the late 1990s, from around 11% to below 6% in 2002, largely due to competition and increased focus on revenue recovery. The number of yearly interruptions was among the lowest in South America, 12 interruptions for a total of 12 hours (Larsen et al., 2004).


603

Grid losses (%)

12

8

4

0 1992

1994

1996

1998

2000

2002

Time Fig. 17.6. Grid losses in the Argentinean electricity market. Source: Adapted from CAMMESA (2002).

Table 17.3 provides an overview and summary of the main milestones in the Argentinean ESI since deregulation, up to 2005, as discussed above. Bacon and Besant-Jones (2001) present the improvement in performance of the ESI before and after privatization for two Argentinean distribution companies, which is summarized in Table 17.4. Both internal indicators (number of employees, customers/employee) and external indicators (number of customers, sales, bad debt, etc.) measure the improvements of those distribution companies as a result of privatization and competition in the ESI. This is similar to what has been experienced elsewhere, such as in the UK, where employment in generation decreased by 60% (Bunn, 1994). There can be little doubt that these companies are much better run following deregulation and privatization, due to the introduction of competition. The latest Argentinean economic recession has made living conditions difficult and electricity tariffs have become a very sensitive issue. The devaluation of the Argentinean peso relative to the US$3 led to increases in gas prices, and consequently increasing electricity costs for industry. Furthermore, after realizing the potential problem for the poorest part of society, the government reduced tariffs for domestic users of electricity and gas, tariffs that have been increased only marginally since 2002, creating problems for the profitability of electricity and gas companies and consequently stopping most of the investment in the sector. Looking ahead, one of the main issues is that even though the capacity margin is large, demand is quickly recovering to historic growth levels and new investments will be required. However, questions have been raised in connection with price structure and lack of incentives for new investment. Rudnick and Montero (2002) argue that the centralized dispatch and regulated prices to final consumers are under attack, and that retail prices are unsustainable as they do not reflect real market conditions, which may lead to a slowing down in new investments and limit the entry of new competitors. This is particularly worrying in the context where demand has started to increase, as discussed above. Problems in 2004 showed what the

3

The Argentinean peso devalued by over 300% in December 2001.

604


Table 17.3. An overview of the development in the Argentinean electricity sector after deregulation. Event

Date

Comments

Energy crisis

1988–1989

Energy crisis with frequent shortages all around the country. Argentine confronted conditions of inefficiencies, bureaucracy, and inoperability that strengthened the energy crisis.

Electricity Act

1989

In August 1989 the Law 23696 appeared which introduced the transformation of the electricity sector. It promoted market principles, such as competition, changes in the role of the state, private participation, etc.

Creation of ENRE and CAMMESA; start of the electricity market

1992

The electricity market began in 1992. The Law 24065/92 provided the regulatory framework and the market rules. The Law created ENRE, whose role is the control of public services, prevention of monopoly behavior, estimation of tariffs, etc. CAMMESA, also created with this Law, was assigned to manage the market based on a national dispatch. The role of the state changed from a firm to a regulator.

Major privatizations

1992

A number of privatizations took place. Among the generators are Central Puerto S.A. (April 1, 1992), Central Costanera S.A. (May 29, 1992) and Central Alto Valle S.A. (August 26, 1992). The distribution, originally called SEGBA (a state-own monopoly), was divided and privatized into three areas: EDENOR S.A., ENDESUR S.A., and EDELAP S.A. From the liberalization to 2004, 2370 MW were installed.

Continuation of privatizations

1993–1994

The private operators had less than 1% in 1991, which was increased to 62% in 1994 after all the privatizations. In 1993, privatizations took place for transport and for liquid combustibles, which strongly affected the electricity costs.

Entrance of minor deregulated customers

1997

In 1995 the number of contracts increased considerably with the entrance of the first minor deregulated customers. It passed from 91 contracts in December 1994 to 448 in December 1995. In 1997 the number of contracts reached 1190.

Azopardo episode

1999

A failure in a substation that caused a number of problems for users and created doubts about the way ENRE managed the crisis.

Continuous price reduction

1992–2001

The prices were reduced by about 70% from 1992 to 2001, motivated largely by the increment of low-cost natural gas supply and the transformation of that sector.

Macroeconomic crisis

2002

Strong alterations in the macroeconomic environment in 2002. The Argentinean peso lost parity with the US$ and suffered a drastic devaluation, which affect the country at all levels. The electricity consumption was reduced around 2% for the first time in 10 years.

Expansion in Brazil connection

2002

The connection with Brazil was increased from 1050 to 2100 MW in 2002.

Low-investments report

2003

Reduction of the investments in generation and transport, partially due to the distortions of the relative prices of the sector.


605

Table 17.4. Improvement in performance of two Argentinean distribution companies before and after privatization, up to 1998. Company Year privatized Energy sales* (GWh/year) (%) Energy losses** (%) Number of employees*** (%) Customers/employee* (%) Net receivables** (days) (%) Provisions for bad debts** (% sales)

ENDESUR 1992 ⫹79 ⫺68 ⫺60 ⫹180 ⫺38 ⫺35

EDENOR 1992 ⫹82 ⫺63 ⫺63 ⫹215 n.a. n.a.

*Positive if amount has increased; **positive if level has declined; and ***positive if amount has declined. Source: Bacon and Besant-Jones (2001).

future can bring in terms of possible blackouts and pressure to increase the regulated domestic price to levels that reflect the cost associated with the generation of electricity. The pressure could further increase if Argentina was to face a period of adverse weather conditions, limiting generation from the hydroelectric sector. We will draw further lessons from the Argentinean experience after the section about Colombia.

17.3. Colombia The deregulation of the ESI in Colombia started in 1994, and the spot market initiated operations in July 1995, supported by Laws 142 and 143 (Congreso, 1994a, b). This was prompted by three main causes (García and Dyner, 2000; Larsen et al., 2004). Firstly, two major blackout periods: the first one in 1983 and the second during 1992–1993 (these blackouts were politically unacceptable and the government needed to find a way of avoiding further blackouts); secondly, the government was unable to finance capacity expansion; thirdly, the system was increasingly inefficient and ran large non-technical losses, adding further to the financial problems. Despite many economic and technological differences, Colombia adopted a version of the UK central-pool model. The Colombian electricity market is the only one in the region where pool prices are settled in a bidding process, as opposed to the cost-based scheme that operates in Argentina, Chile, and Brazil (Millán, 2001), that is the system is price based, rather than cost based as in the rest of the subcontinent. Companies submit daily bids of both energy and prices (from hourly bids originally it has now changed to block bids) to the Centro Nacional de Despacho (CND, the system operator). The CND decides on dispatch according to merit order, taking into account system restrictions. There is a capacity payment mechanism in place, intended to provide investment incentives to generators. These incentives are allocated according to an “optimization–simulation model” operated by the CND and regulated by the Comision de Regulacion de Energia y Gas (CREG, Regulation Commission for Energy and Gas). The capacity payment is fixed at 5.25 US$/kW/month. The liberalization also followed the UK model initially, as competition was only created in the wholesale market, but with no explicit commitment to move to retail competition. One major change from the implementation of the Colombian deregulation was that companies were allowed to be both generators and distributors; however, limits were set on how much electricity a distributor could buy from its own generation company (maximum 60%). Most transmission is part of an independent company, ISA, which has recently been floated on the

606


Installed capacity

MW

12,000

9000

Maximum monthly demand

6000 1996

2000 Time

2004

Fig. 17.7. Installed capacity and monthly maximum power demand in Colombia.

Colombian stock exchange; while some generation companies remained publicly owned others were sold to private investors. The Colombian electricity industry is characterized by a large hydroelectric component, close to 70%, and is considered to be one of the most competitive markets in the developing world (Larsen et al., 2004). The installed capacity and the maximum demand for ≈ 1995–2005 are shown in Figure 17.7. In 1998–1999, however, the country faced the worst recession in a century (growth of about ⫺5% in GDP), reflected in a totally unexpected fall in demand (UPME, 1996, 1999), of about 750 MW, from which it has recovered gradually in recent years, as shown in Figure 17.7. At the end of 1997 and beginning of 1998, El Niño South Oscillation occurred, which led to a reduction in the water supply to the hydro-based electricity system, resulting in sharp spotprice increases but little effect on the average contract price (Fig. 17.8). It is important to notice, however, that blackouts did not happen during this period as the system was capable of producing sufficient electricity to satisfy demand. Compared with 1992, when Colombia faced the same macro-climatic phenomenon with serious consequences in terms of shortage and blackouts, what happened during 1998 “proved”, to many, that deregulation had important benefits, as the system could successfully confront a Niño of such intensity. However, it is not clear whether the system will deliver the necessary increase in capacity now that the economy is back in growth mode and the system is showing weaknesses, especially regarding the capacity payment mechanism (as discussed below) (Larsen et al., 2004). While there have been ongoing concerns in the UK system about market power and possible price manipulation, this has not been true to nearly the same extent in Colombia. Monthly estimates of HHI since 1995 provide the broad picture of the situation. As expected, HHI shows more variability for generation than for capacity or power availability. For capacity, according to the HHI, concentration has declined from about 1400 in 1994, to below 1200 in 2003. Note that at this level, the HHI indicates that there is moderate concentration, which might seem unproblematic; but when examined seasonally and locally, the HHI for generation shows values closer to 1800 because of grid restrictions or hydro-power unavailability.


607

120

Price (US$/MWh)

Spot price 80

40

Contract price

0 1996

2000 Time

2004

Fig. 17.8. The evolution of pool and contract electricity prices in Colombia from 1995 to 2002.

Losses in the Colombian electricity market are considered moderately high compared with other countries in the developing world (Larsen et al., 2004). The Grid Company (ISA) reported 21% losses in 1994. After deregulation, losses fell to around 15% in 2000, with important differences between regions, mainly due to competition. Another, uniquely Colombian, problem is the number of terrorist attacks aimed at the electricity infrastructure: 483 pylons were taken down in 2002, 329 in 2003, and 127 in 2004. However, the system operator has learnt to cope with these emergencies and kept the system on line. One side effect is that volatility increases as the grid gets disabled in places. Table 17.5 provides an overview of the major ESI milestones after deregulation in Colombia. The events are discussed in the subsequent text. There are limited network interconnections between Colombia and its neighbors. Some electricity transactions are taking place through International Energy Transactions (TIEs), with Ecuador, Peru, and Venezuela. These countries conform to what is called “Mercado Eléctrico Andino”, MEA. During 2004, Colombia exported 1634 GWh to Ecuador, which represents 13% of the Ecuadorian demand (MEA, 2005). This is a significant step toward integrating the region, which may take advantage from the complementarities that exist, including: hydrological differences between Ecuador and Colombia, and technological and time differences between Venezuela and Colombia. Market evolution has been satisfactory in terms of investment, competition, efficiency, and reduction in electricity losses. Market (pool) prices have remained low but regulated (domestic) tariffs and subsidies are still a major issue. This has created problems for a number of mainly smaller distribution companies that seem non-viable; not only are customers incapable of paying for electricity, but subsidies are insufficient, and in many of the affected areas the distribution companies experience high losses. A detailed account of the Colombian electricity markets is presented in Larsen et al. (2004). Possible reforms are now under review. CREG (the regulator) has taken under consideration a variety of studies (COMILLAS, 2000; UN-COLCIENCIAS-ISA, 2000; TERA, 2001) to adjust the capacity charge mechanism. However, these studies were conducted in 2000 and there has been

608


Table 17.5. An overview of the development in the Colombian electricity sector after deregulation. Event

Date

Comments

Energy crisis

1992–1993

Major blackouts in 1992–1993 revealed the problems with the current system and led to the initial proposal for deregulation.

Electricity Act

1994

Immediately after the Electricity Law was made public in 1994, investments in gas-fired plant took place in the North of Colombia, taking advantage of unsatisfied demand and system restrictions.

Start of the electricity pool

1995

Colombian regulator created the one-day-ahead pool, similar to the one that had been operating in the UK since 1990.

Learning period

1995

The initial 8 months of the pool went fairly smoothly. Prices were slightly higher than expected and there was overreaction from both regulators and companies. The minimum operational levels for hydroelectricity plants, set by regulators, caused higher prices than indicated by above-average hydrology conditions for December 1995.

Announcement of El Niño

1995

December 1995: the announcement of El Niño encouraged further investment in thermoelectricity plants.

Occurrence of El Niño

1997

The build up of the worst Niño registered this century revealed a number of facts related to companies’ (a) inefficiencies, (b) management problems, (c) lack of trading capabilities, (d) capacity payment inconsistencies. Nevertheless the system passed its most important test. No blackouts!

Inefficiencies revealed

1997

More than 10 small distribution companies plus an important regionally integrated utility proved non-viable. Losses were higher than book value. These companies were taken over by the government and sold (in the process). Corelca (an important regional utility) was separated into three different activities: generation, transmission and distribution.

Loss of the most important multiutility company

1997

The most important multi-utility company in the country revealed weaknesses. As it owns the only multi-annual reservoir in Colombia it had over-contracted power, not fully taking into account the reservoir intervention regulations (and its possible revision), before El Niño took place. As a result of a strong Niño and not being allowed to dispatch its hydro plant, Empresas Públicas de Medellín E.S.P. had to buy electricity at prices sometimes seven times as high as the price contracted for delivery to other companies. The losses amounted to about 100 US$ million in just a few months.

Occurrence of La Niña

1998–1999

La Niña period 1998–1999 has revealed how difficult it is for a purely thermoelectric company to survive within a system with a large hydroelectric component, strong weather variations, and heavily regulated prices. Prices have been much lower than during the period prior to deregulation.

Adjustments of capacity payments

1999

There is great satisfaction among generators and politicians in 1999 with adjustments in regulatory issues related to the capacity payment and also with respect to the minimum operation level of hydroelectricity plants. There is a threat that thermoelectric plants will disappear from the central region of Colombia because of the large number of hydroelectric units in this region


609


Date

Comments and the slowdown of the economy. This has created temporary over-capacity, making it difficult for the gas units to compete with water turbines, as the former operate at zero fuel costs.

Announcement of privatization

2000

It has been announced that the transportation (grid) utility company will go private. About 12% of it has already been sold to a number of small investors. The company manages to attract over 60,000 new owners in this way.

Major attacks on infrastructure

2001

Major attacks on the transmission infrastructure have isolated some important generators’ cheap generation capacity. As a consequence of this some companies in the electricity system have to run with more expensive plants in operation and some companies have apparently managed to exercise market power.

Intervention for apparent market power

2001

Major political pressure has built up as prices increase dramatically during the time of attacks on the infrastructure. A new interventionist rule has been put in place by CREG (Resolution 034, 2001) intending to control the apparent market power exhibited by some generators.

2002

A second tranche of ISA shares (in excess of 10%) was sold.

2005

Creation of the new standardized electricity-contracts market.

Electricity resolution

little or no progress since then. This may create major problems for the system in the midterm if generators do not build new capacity. We will discuss further the insights from the Colombian case, below.

17.4. Comparison of the Evolution of the Two Electricity Markets We have briefly discussed two case studies to provide an overview of two different electricity-market structures in South America. These countries have chosen two different models on which to base their market and regulation. The deregulation process is generally not well understood in the sense that it is difficult a priori to predict which markets will be successful and which will fail. We need to examine the factors and conditions that determine the circumstances under which the alternative “deregulation models” are likely to succeed. One way of building a better understanding is to compare and contrast countries that are similar in many dimensions (i.e. to hold one set of variables constant) and see how other sets of variables have affected the outcome. However, to obtain a broader insight we need to make more than a pure economic analysis. Argentina and Colombia have some similarities (e.g. culture, language, etc.) but also have many differences in relation to geographic location, technology, economy, and also to the way they chose to deregulate (e.g. market framework, regulatory intervention and control, timing, etc.). In this section we present a comparison between these countries, based on a number of factors from the previous section. We focus on the common structural elements found in these markets. Such lessons can inform not only countries in South America but also other countries that have recently deregulated or are about to deregulate. Table 17.6 summarizes the two cases.

610


Table 17.6. Summary of development in Argentina and Colombia.

Pre-restructuring political environment and ownership

Selected model Pool design Number of firms*

Argentina

Colombia

Weak economic growth and unstable political institutions since the 1940s; however, they have strong provincial institutions. State-owned industry until the recent privatization

Until recent privatization, government owned almost the whole industry, with a continued political instability, up to the current date.

Adapted with improvements from Chile

Adapted from UK

Cost-based bidding 38

Bid based 26

763

1681

Cost Price cap Price cap

Price Price cap Price cap

10.04 15.25 7.30

6.42 7.78 4.19

12

60

12

58

38

70

Proceeds from sale of electricity distribution entities* (million US$) Price-setting mechanisms Generation Transmission Distribution Average electricity prices in June 2001 (US cents/kWh) Residential Commercial Industrial Quality of the service Average number of interruptions per year Hours of interruptions per year Electricity trade in contracts (%) *Source: Bacon and Besant-Jones (2001).

A qualitative summary of what we have observed is presented in Table 17.7. It shows a cross comparison of the two cases and the current state of their performance in a number of areas. Argentina and Colombia have had high reserve margins, which led to very low spot prices at the beginning of the new century. The underlying reason for the high reserve margins is, largely, the deep recessions suffered in their respective economies. Insufficient electricity generation capacity could constrain future economic growth. The natural question is then, do the current electricity markets, given their structure, provide the incentives for expansion to meet consumers’ demand? We will discuss a possible answer to these questions, in two parts, below. Figure 17.9 illustrates this problem in Colombia, as it shows the rate of return in the sector in 1999 and 2000, returns which have not improved much since. It is clear that investors


611

Table 17.7. Qualitative cross comparison of the electricity markets’ performance in Argentina and Colombia.

Reserve margin Volatility Market power Losses Interruptions

Argentina

Colombia

High Low Moderate Low with improvements Improvements

High High Moderate and local High with improvements Little improvements

5 4 3

Generation Transmission Distribution

2 1 0

1999

2000

⫺1 ⫺2 Fig. 17.9. Estimated percentage rates of return on investment in the different electricity sectors in Colombia.

will not be lining up to invest if this is the level of return. Furthermore, is there potentially country risk as well, which would lead most private investors to require even higher returns than if it had been in Europe? Deregulation does not mean that a queue of investors will instantly form to begin investing in additional capacity. Investment decisions take significant amounts of time. Decisionmakers will be looking for the appropriate signals to start new generation projects. These often include considerable time to establish the government’s commitment to reforms and the regulator’s determination to carry through deregulation. Further, when a decision to build capacity has been made, it takes a significant period to bring this capacity on line (in the best of cases, 2–3 years). When changes occur, there is high uncertainty and the potential investors and current owners of capacity must learn about the operation of the new rules, a learning process that takes time. A similar situation occurs when there has been a long period of excess capacity in a market (i.e. high reserve margin and low prices). As economic recovery takes

612


place in Colombia and Argentina, it will most likely require a significant amount of time before investors will be committed. Investors need to be confident about the regulatory framework and also be sure of the demand-growth patterns before they commit to investment, a process likely to take a few years. Both the regulatory framework and government policy can help or hinder the process. If investors feel that the regulatory framework is, for example, not favorable to thermal generation – as might be the case in Colombia given the nature of the capacity payment – they will be even more reluctant to return to add new required (thermal) capacity. The same is true for government policy; for example, the government of Argentina put a price cap on electricity and gas prices during the economic crisis and has not yet removed them completely, which has caused the investment in new electricity capacity as well as gas exploration to dry up. Deregulated electricity markets tend to induce cyclical behavior in reserve margin (Bunn and Larsen, 1992, 1999; IEA, 1999; Ford, 2001, 2002; Dyner et al., 2003). Even though there is not yet enough data to analyze this statistically, we might observe two phases of the cycles. While Argentina and Colombia were increasing their reserve margin during the late 1990s these have been rapidly decreasing since the turn of the century. It should be possible to take advantage of this situation as discussed below. The spot electricity prices in Colombia and to a lesser extent Argentina are driven strongly by rainfall, a situation that makes them vulnerable to extreme El Niño/La Niña events. These create excessive rainfall in some countries, while at the same time, droughts in other countries. Lack of rain leads to water scarcity, which increases prices and volatility, while excess rain creates very low prices (making it difficult for non-hydro-based generators to run). Colombia has faced two events of this type, both water scarcity, since deregulation: the first in 1997–1998, which significantly increased prices but posed no serious threat to the electricity supply; and the second in 2002, which had no significant implications as it was much less intense and Colombia had a larger reserve margin compared with the previous event. Another important aspect that can be observed is the partial decoupling of the wholesale and retail markets, especially with respect to prices, as these are not passed on directly from producers to consumers (except for large users). The way that changes in the wholesale market are transmitted is through a smoothing process that incorporates the price variation into the consumers’ tariff (or price), with long time lags (Stoft, 2002). For other energy forms, such as oil, the end-consumer is subjected to price variations causing rapid changes even in the relatively short term (e.g. the price of petrol). Part of the explanation for this is the long-term contract that exists between (large) customers and their suppliers, including distribution companies; Table 17.6 shows the fraction of the electricity traded on contracts, around 70% in Colombia. In other cases where the market is not fully deregulated, the domestic sector is still a monopoly and prices are only allowed to change very slowly (largely influenced by politics). In systems that depend on hydroelectric power, this might create even worse problems, especially during the long dry periods (causing less inflow to the reservoirs) as has occurred in the countries discussed in this chapter. If consumer prices are not providing signals of real electricity cost in the wholesale market, the national system will be caught in a potentially unsustainable situation where the price increases in the wholesale market will have no effect on consumption, as tariffs to consumers remain constant. Colombia and Argentina have shown that both price-bidding and cost-based models can produce acceptable results in deregulated electricity systems. However, in the late 1990s and early this century, both countries have had major recessions in their respective economies, which stopped or reversed any growth in demand. The test will be when economic growth returns, as has been the case in Argentina, whether the system will provide the right signals to increase generation capacity. Although it is still early, recent experiences in Argentina


613

might indicate that there will be major problems and the system will not be able to respond with new investments fast enough to avoid major disruptions. A similar situation could occur in Colombia, where there is real concern about the current return on investments and the effect this might have on future investment (Larsen et al., 2004). The real test of these regulatory systems will be in the years to come, when there will almost certainly be a strong need for investment in new capacity. However, as discussed above, the effect of lags might prevent enough new capacity being in place when it is most needed (Fischer and Galetovic, 2000). To make the situation even more complex, the political instability in Latin America triggers debates on ownership of the electricity sector (public versus private) as well as prices, tariffs, cross-subsidies, and the robustness and integrity of the regulatory institutions. All of these factors affect the perceived risk and uncertainty to foreign investors and might, in the end, have detrimental effects on the willingness of private investors to provide the necessary capital (Spiller et al., 1996). These problems will be further amplified as long as there is seen to be major political involvement in the markets, as investors are likely to delay investments while waiting for political stability, that is for less political involvement in the markets. How long this period is likely to last is difficult to foresee (Larsen et al., 2004).

17.5. Discussion and Conclusion The first impression is that Argentina and Colombia did well overall during the years after deregulation; however, when one takes a closer look the complete picture is more ambiguous. The adjustment processes of the frameworks have not been “maintained”, creating potential major problems in the future. Leaving aside the initial determination to solve old problems, we have seen a reluctance to face the imperfections that emerged within the newly deregulated industries – which has led to a stalemate in the future development of the electricity industry. This said, one should not underestimate the successes that these systems have had over the last decade. Here we have presented and compared two cases of national deregulation of electricity markets in Latin America. These markets are far from completion but provide a number of warnings to other countries. We have pointed out both some existing and potential problems. Argentina had electricity shortages in 2004 and Colombia is now facing a dangerous decline in reserve margin. There are, however, a large number of positive experiences. These include increases in private investments, better managed systems (lower losses), and potential (in some cases realized) lower prices. There are also a number of cases in which the systems managed to provide enough electricity to cover demand, where many people would have thought it not feasible. Another important question to ask is whether the systems would have done better without deregulation. Of course, it is not possible to answer this question, but it is worth remembering that in many cases it was blackouts and bad management that led to the deregulation decision. Colombia and Argentina are countries with significant hydroelectric generation capacity (70% and 37%, respectively). Colombia has faced major price volatility because of extreme hydrological conditions. It is obvious that a high percentage of hydroelectricity implies a high volatility in prices, as has previously been seen in Norway. Argentina has experienced a relatively low volatility, which can be explained by its lower hydroelectric capacity, and had not faced a significant reduction in hydrological conditions until 2004. After the initial transformation the two countries almost “froze” the reforms and no substantial adjustments have been made to cope with the challenges that have emerged subsequently. Argentina and Colombia are in unstable political environments that have delayed new rounds of reforms,

614


being termed “Second Generation Reforms” (Millán, 2001). It is should be clear that deregulation is not a “one-off” event, but an ongoing process that will have to continue for at least as long as the systems’ transition period, which can last decades (Dyner and Larsen, 2001). Generally, productivity and efficiency have increased after the reforms, as a result of the introduction of competition. However, not enough of the productivity and efficiency gains have been passed on to consumers (Haselip et al., 2005). Argentina has faced the problem, because the regulator has avoided creating uncertainty for investors about revenues in the future by allowing a relatively high tariff (Bacon and Besant-Jones, 2001). A similar situation occurs in Colombia (Larsen et al., 2004). This might make the situation better for the investors, but create another set of problems as consumers do not see sufficient benefits from the reforms. If there is not a fair sharing of the benefits from deregulation it is likely that there will be increasing political pressure to stop, delay or even reverse the creation of markets, as it is unlikely that consumers will appreciate the fact that the extra investment will prevent shortages in the future. The new round of reforms should take into account different aspects of the problem. Firstly, they should consider an economic recovery of the region. Secondly, they may rely more on market institutions (e.g. financial elements), which would create the appropriate incentives and tools for resource allocation and risk management. This would improve the industry performance and reliability. Thirdly, the reforms should consider the State of the Art of electricity markets; particularly, they should include learning from areas with a similar technology mix, such as NordPool. Fourthly, the future design should take into account the possibility of creating regional integration, again possibly looking at NordPool. Expanding on the last point, we believe the next step for the South American electricity markets has to be the development of regional integration among neighboring countries. Although regional integration has slowly started, this needs to be one of the main areas of focus for Congress, regulators, and transmission companies in the region. It will help markets to improve their performance, not so much because of the competition, but mainly because of the existing complementarities among the neighboring countries. The MEA (2005) is already in operation, and a bilateral market between Chile and Argentina is under study. Regional integration is feasible under the right political climate, but it requires both commitment as well as taking into account other experiences with similar topologies, such as the Scandinavian model. More research and analysis is needed in this direction.

Acknowledgments The authors thank ISA-MEM, the Electricity Market Manager in Colombia, for providing valuable information and Paul G. Ellis for his comments.

References Bacon, R.W. and Besant-Jones, J. (2001). Global electric power reform: privatization and liberalization of the electric power industry in developing countries. Annual Review of Energy and Environment, 26, 331–359. Bunn, D.W. (1994). Evaluating the effects of privatizing electricity. Journal of the Operational Research Society, 45(4), 367–375. Bunn, D.W. and Larsen, E.R. (1992). Sensitivity of reserve margin to factors influencing investment behavior in the electricity market of England and Wales. Energy Policy, 20, 420–429. Bunn, D.W. and Larsen, E.R. (1999). Deregulation in electricity: understanding strategic and regulatory risk. Journal of the Operational Research Society, 50, 337–344. CAMMESA (2002). Argentine wholesale electricity market annual report. Compañia Administradora del Mercado Eléctrico Argentino, Buenos Aires, Argentina, February 2002.


615

CAMMESA (2003). Argentine Wholesale Electricity Market Annual Report. Compañia Administradora del Mercado Eléctrico Argentino, Buenos Aires, Argentina, March 2003. CAMMESA (2003a). Argentine wholesale electricity market annual report. Compañía Administradora del Mercado Eléctrico Argentino, Marzo. COMILLAS (2000). Estudio cargo por capacidad en Colombia. Final Report for Asociacion Colombiana de Generadores de Energía Eléctrica – ACOLGEN, Universidad Pontificia COMILLAS, Madrid, May 2000. Congreso de la Republica de Colombia (1994a). Ley 142 de 1994. Ley de Servicios Publicos, Bogota, Colombia. Congreso de la Republica de Colombia (1994b). Ley 143 de 1994. Ley Electrica, Bogota, Colombia. Dyner, I. and Larsen, E.R. (2001). From planning to strategy in the electricity industry. Energy Policy, 29, 1145–1154. Dyner, I., Lopez, B. and Arango, S. (2003). Modeling uncertainty in electricity markets for learning, policy and strategy. International System Dynamics Conference, New York. ENDESUR (2002). Luces Argentinas: Una Historia de la Electricidad en Nuestro Pais. Empresa Distribuidora Sur S.A., Argentina, ISBN: 987-20209-0-6. Energía, S.D. (2002). Informe del sector electrico año 2001. Secretaría de Energía, Ministerio de Economía, Buenos Aires, Argentina. ENRE (1999). El Informe Electrico – Cinco años de Regulación y Control: 1993–Abril 1998. Ente Nacional Regulador de la Electricidad, Buenos Aires, Argentina, p. 106. ENRE (2002). Informe Anual 2001. Ente Nacional Regulador de la Electricidad, Buenos Aires, Argentina. ENRE (2003). Informe Anual 2002. Ente Nacional Regulador de la Electricidad, Buenos Aires, Argentina, p. 276. Espinasa, R. (2001). Marco Institutional de los Sectores Electricidad y Telecomunicaciones en America Latina. Research Report, Research Department, Inter-American Development Bank, Washington, DC. Fischer, R. and Galetovic, A. (2000). Regulatory Governance and Chile’s 1998–1999 Electricity Shortage. Serie Economia, 84, Universidad de Chile, Chile. Ford, A. (2001). Waiting for the boom. A simulation study of power plant construction in California. Energy Policy, 29, 847–869. Ford, A. (2002). Boom and bust in the power plant construction. Journal of Industry, Competition and Trade, 2, 59–74. García, M. and Dyner, I. (2000). Reform and regulation of electricity in Colombia – UK experiences taken across the Atlantic. International Journal of Public Administration, 23, 603–623. Haselip, J., Dyner, I. and Cherni, J. (2005). Electricity market reform in Argentina: assessing the impact for the poor in Buenos Aires. Utilities Policy, 13, 1–14. Herfindahl, O.C. (1950). Concentration in the steel industry. Unpublished Ph.D. dissertation, Columbia University, New York. Hirschman, A.O. (1964). The paternity of an index. American Economic Review, 54(5): 761 . Hunt, S. (2002). Making Competition Work in Electricity. Wiley, New York. IDB (2000). Energy Sector Strategy. Inter-American Development Bank, Washington, DC. IEA (1999). Electricity Reform Power Generation Costs and Investments. International Energy Agency. OECD/IEA, Paris, France. ISA (2003). Interconexión Electrica S.A.E.S.P. Market Report, Medellín, Colombia. Kwoka Jr., J.E. (1985). The Herfindahl index in theory and practice. Antitrust Bulletin, 30, 915–947. Larsen, E., Dyner, I., Bedoya, L. and Franco, C. (2004). Lessons from deregulation in Colombia: successes, failures and the way ahead. Energy Policy: 1767–1780. MEA (2005). Mercado Eléctrico Andino. Available: www.mercadoelectricoandino.com. Accessed: 2005. Millán, J. (2001). The Second Generation of Power Exchanges: Lessons for Latin America. InterAmerican Development Bank. Washington DC. Newbery, D.M. (2001). Privatization, Restructuring, and Regulation of Network Utilities. The MIT Press, Massachusetts. OLADE, ACDI-CIDA, UNIVERSITY OF CALGARY, (2004). Competencia en mercados energeticos: una evaluacion de la reestructuracion de los mercados energeticos en América Latina y el Caribe. Advance Project Report, Mentor Poveda, MSEE, December 2004. Pistonesi, H. (2002). El Sistema Eléctrico Argentino: Inversión y Competencia. Energética, 28: 9–25.

616


Rudnick, H. and Montero, J.P. (2002). Second generation electricity reforms in Latin America and the California paradigm. Working paper series No. 216, Universidad Católica de Chile, ISSN (electronic edition) 0717-7593. Spiller, Pablo T. and Viana L.M. (1996). How should it be done? Electricity regulation in Argentina, Brazil, Uruguay, and Chile. In R.J. Gilbert and E.P. Kahn (eds.), International Comparisons of Electricity Regulations. Cambridge University Press. Stoft, S. (2002). Power System Economics, Designing Markets for Electricity. Wiley Inter-Science, New York. TERA (2001). Escisión del SIC – Bolsa de Energía Eléctrica de las Actividades Comerciales de Interconexión Eléctrica S.A. E.S.P. y Definición del Esquema Institucional para las demás funciones del Actual Centro Nacional de Despacho. Programa de las Naciones Unidas para el Desarrollo (PNUD), Banco Mundial, presented by Teknecon Energy Risk Advisors, LLC, February, 2001, Bogota, Colombia, 4 volumes. UN-COLCIENCIAS-ISA (2000). Opciones de Manejo del Recurso Hídrico en el Mercado Eléctrico Colombiano. Research Project Report, Energy Institute, Universidad Nacional de Colombia, Medellin, Colombia. UPME (1996). Plan de expansion de referencia. Ministerio de Minas y Energía, Bogotá, Colombia. UPME (1999). Plan de expansion de referencia. Ministerio de Minas y Energía, Bogotá, Colombia. World Bank (1993). The World Bank’s Role in the Electric Power Sector. Policies for Effective Institutional, Regulatory and Financial Reform. World Bank, Washington, DC. World Bank (2003). World Development Indicators database. World Bank, Washington, DC.

Chapter 18 A New Stage of Electricity Liberalization in Japan: Issues and Expectations MIKA GOTO AND MASAYUKI YAJIMA Socio-economic Research Center, Central Research Institute of Electric Power Industry, Tokyo, Japan

Summary Electricity liberalization in Japan entered a new stage in April 2005 with the expansion of eligibility to include all high-voltage customers and the commencement of operations of the Japan Electric Power Exchange. This was preceded by the introduction of a market mechanism for the wholesale sector in 1995, which involved competitive bidding for additions to the generation capacity. In 2000, retail competition was initiated for customers whose maximum demand is 2000 kW or more as well those who are supplied electricity at extra-high voltages (⭓20,000 V). In this context, the Japanese Government will begin a major policy discussion regarding whether the eligibility should be further expanded to include household customers. The implications of the evolving market on the country’s electricity supply industry (ESI), including issues such as the fate of nuclear power in Japan, will also be discussed. This chapter provides an overview of the fledgling market reform initiatives and their long-term consequences for the Japanese ESI. 18.1. Introduction Electricity liberalization in Japan entered a new stage in April 2005 when contestability was expanded to include all high-voltage customers. New market reform initiatives also include the establishment of two new organizations, the Electric Power System Council of Japan (ESCJ) and the Japan Electric Power Exchange (JEPX). Since Japan adopted market reforms rather late, it does not have much to offer in the way of empirical results or experience thus far. However, its unique characteristics, including the paramount energy security and environmental concerns, make it an interesting market to study. Moreover, the Japanese work culture involves a substantial amount of deliberation, which encourages consensual decision-making and offers interesting insights for other countries that are contemplating market reform. Japan has adopted a cautious and gradual approach toward electricity deregulation by examining the experiences of other countries. This chapter is organized as follows. Section 18.2 briefly describes the historical background of the Japanese ESI and its market structure. Section 18.3 describes the deregulation 617

618


Fig. 18.1. The 10 electric power companies by service area. Source: The Federation of Electric Power Companies of Japan (2005).

process and new market institutions. Section 18.4 discusses the observed changes in the ESI resulting from the electricity liberalization, including changes in price levels, customer switching rates, wholesale trading, and trends in investment and productivity. Section 18.5 discusses a number of remaining issues, which are yet unresolved, including the wholesale power market, the governance of the ESCJ, full retail liberalization, and the future of nuclear generation under deregulation. Section 18.6 summarizes the main issues and challenges facing the liberalization process in the years ahead.

18.2. The Electric Power Industry in Japan There are 10 vertically integrated, investor-owned electric utilities (IOUs) in Japan, each with an exclusive service area (Fig. 18.1). They have been performing all the functions that are required to supply electricity, from generation to retail sales. With the exception of the Okinawa Electric Power Company, the service areas of all these IOUs are interconnected by extra-high-voltage transmission lines. Before the recent introduction of deregulation in Japan, these utilities were regulated by the Ministry of Economy, Trade, and Industry (METI). The current structure of the ESI dates back to 1951. At that time, there was a heated discussion regarding the organization and ownership of the industry, and whether it should be state owned or privately owned. Japan opted for the latter, believing that it would lead to a more efficient outcome despite the fact that at that time many other countries were operated by state-owned companies. Table 18.1 provides the relevant statistics for the industry. In addition to the IOUs, there are two large-scale wholesale utilities, Dengen Kaihatsu and Nihon Genshiryoku Hastuden, whose capacities were 16,375 and 2617 MW, respectively. Other than these two companies, there also exist more than 50 joint thermal power plants defined as deemed wholesale utilities that sold power to IOUs before the deregulation. The peak load in Japan was approximately 167 GW in 2003. The peak load for the largest company, Tokyo Electric Power Company (TEPCo.), was approximately 57 GW in 2003, and the second, and third, largest companies, Kansai and Chubu, experienced peak loads of 31 and 26 GW in 2003, respectively.

619

A New Stage of Electricity Liberalization in Japan Table 18.1. Installed capacity, number of customers, and the total revenue for the 10 electric power companies in 2004. IOUs from North to South Hokkaido Tohoku Tokyo Chubu Hokuriku Kansai Chugoku Shikoku Kyushu Okinawa Total

Installed capacity (MW)

Number of customers

Total revenue (million yen)

6,584 15,515 62,825 32,585 6,754 35,761 12,205 6,861 19,422 1,916

3,873,983 7,672,967 27,719,529 10,337,578 2,004,238 13,155,600 5,206,272 2,872,963 8,291,184 784,250

512,872 1,464,137 4,851,769 2,071,121 460,958 2,464,645 965,359 529,599 1,338,444 137,646

200,428

81,918,564

14,796,550

Data source: Handbook of Electric Power Industry (2005).

Each company is responsible for continuously balancing electricity supply with demand, maintaining the quality and ensuring the security of the grid in its service area, principally by using its own generation, transmission, and distribution facilities. Bilateral trading between IOUs and other generators were also used as supplements for the fulfilment of their responsibilities. The interconnection between adjoining companies is in the form of a one-point connection with very limited capacity. This is due to the concept of network configuration, a legacy dating back to the foundation of the industry. In other words, it is based on the concept that a blocked transmission network system in each company is desirable from the perspective of reliable power supply; this is because if network trouble occurs in any one supply area, it does not affect the other areas due to the capacity limitation among areas (Nambu, 2003).1 The electricity trading among IOUs has historically been rather constrained due to these reasons.2 Another major limiting factor is that the Japanese ESI currently operates as two distinct systems with different frequencies. In the Eastern region, electricity is supplied at 50 Hz, while in the Western region it is supplied at 60 Hz.3 A frequency converter is used to provide an interconnection between the two regions; however, it has serious physical restrictions 1

Such a basic concept of limited interconnection is also related to the constraints that originate from Japan’s mountainous geography. In addition, nowadays, a construction of the new interconnection is further difficult, which generally requires more than 10 years due to various institutional constraints such as regulations for environmental protection. 2 Even under the physical constraints of the limited interconnection capacity; however, there existed a significant need to trade electricity through bilateral contracts between each IOU as well as between IOUs and wholesale utilities. For example, in 2003, the ratio of such net trading volume of electricity to the final demand was approximately 25% for TEPCo. and 30% for the Kansai Electric Power Company. Moreover, the average ratio of the net trading volume was approximately 20% for the 10 IOUs in 2003. Most of them are fixed by using long-term bilateral contracts, for example, 15-year purchase contract is concluded at the beginning of the operation of the specified generation plant. 3 This is a historical consequence of the fact that when the industry was in its nascent stage, the Eastern region imported electric generators from Germany, while the Western region imported generators from the USA. Hence, both transmission systems were independently developed.

620


Fig. 18.2. National trunk transmission interconnections. Source: The Federation of Electric Power Companies of Japan (2005).

(Fig. 18.2).4 Such limited linkage capacities present a technical issue that needs to be resolved because it presently restricts opportunities for power trading.5 An efficient use and operation of the interconnection will be discussed in the future, which includes issues on current restrictions on minimum volume and step size for the use of frequency changer (F.C.) and a reevaluation of a required margin of the interconnection. Figure 18.3 illustrates the output and fuel mix over time. During the period from 1970 to 2004, when consumption grew from 252 to 785 TWh, Japan experienced two “oil shocks,” which resulted in a major shift in its power sources from oil to liquefied natural gas (LNG) and nuclear energy. Since Japan has very limited natural resources, energy security and balanced energy supplies are of paramount importance.6 The share of fossil fuels decreased from 77.1% to 57.5% during this period. Conversely, the share of nuclear power increased significantly from 0.5% in 1970 to 33.4% in 2004. Moreover, the share of hydro has also shrunk over time.

18.3. Electricity Liberalization 18.3.1. The deregulation process Discussions regarding the reform of the Japanese ESI date back to 1994 in the Electricity Utility Industry Council (EUIC), which is an advisory committee to the METI. The primary 4

Specifically, the capacity of the Shin-Shinano F.C. is 600MW, and that of the Sakuma F.C. is 300MW, together totaling 900MW. 5 Currently, there is a plan to acquire new F.C. equipment with a capacity of 200 MW for the purpose of beginning operations after 2007. 6 For example, according to the 2002 statistics, the share of imported oil in Japan’s domestic oil consumption was 99.8%; the corresponding figure for the USA was 59.5% (OECD/IEA, 2002).

621

A New Stage of Electricity Liberalization in Japan 1970

0.5%

1980

252 TWh

17.4% 22.4%

15.5% Hydro Fossil fuel Nuclear

Hydro Fossil fuel Nuclear

77.1%

67.1%

1990

2004 637 TWh

785 TWh 9.0%

10.3% 28.4%

413 TWh

Hydro Fossil fuel Nuclear

Hydro Fossil fuel Nuclear Other

33.4%

57.5%

61.3%

Fig. 18.3. Changes in the total volume of generation and shares by fuel mix of nine electric power companies. Data source: Handbook of Electric Power Industry (1971–2005). Note: “Other” in 2004 represents renewables. Table 18.2. End-user electricity prices in five OECD countries in 1995 and 2003. Total price (US$/unit)

Total price (US$/unit) (using PPP)

1995 2003 1995 2003 Industry Household Industry Household Industry Household Industry Household Japan

0.19

0.27

0.12

0.19

0.10

0.14

0.10

0.15

USA UK Germany France

0.05 0.07 0.10 0.06

0.08 0.13 0.20 0.17

0.05 0.05 NA 0.04

0.09 0.12 NA 0.13

0.05 0.07 0.07 0.05

0.08 0.13 0.14 0.13

0.05 0.05 NA 0.04

0.09 0.11 NA 0.12

Data source: OECD/IEA (2004).7 Note: Statistics are in US$/unit and that using purchasing power parity (PPP).

motivation for reform was to address the issue of high electricity prices in Japan as compared with those in other developed countries. The background of the discussions – the disparities in the price levels between Japan and foreign countries in general – was apparent in those days due to the yen’s sharp appreciation against the dollar. However, the comparatively higher electricity prices were not attributed only to the exchange rate (Table 18.2). 7

OECD/IEA stands for Organization for Economic Co-operation and Development/International Energy Agency.

622


100

99.8

98.0

96.2

Oil Total

82.2 80 51.1

40 Percent

60.8

59.5

60

27.5

20 0 ⫺20

⫺13.3

⫺40

⫺35.2

⫺60 Japan

US

UK

Germany

France

Fig. 18.4. Degree of dependence on imports in oil consumption and the total energy consumption. Data source: OECD/IEA (2005).

90

82

80

73

Minutes

70 60 46

50 40 30 20

13

10 0 Japan

US

UK

France

Fig. 18.5. Duration of annual electric power failure caused by incidents. Data source: Japan Electric Association (2004). Note: The statistics for Japan pertain to the year 2002, and those for the USA, UK, and France pertain to the year 2000.

It is true that higher electricity prices in Japan are partially attributable to a higher dependence on imported energy (Fig. 18.4) and higher reliability standards (Fig. 18.5); however, there is a consensus that structural and operational issues were also responsible. The growing interest in market reform was further supported due to the stagnation of the Japanese economy, which commenced in the beginning of the 1990s following the bursting of the “bubble economy” as well as the movement toward deregulation in Chile, the UK, and in the Nordic countries.

A New Stage of Electricity Liberalization in Japan

623

As a result of the discussion, the Electric Utility Industry Law (EUIL) was amended in 1995 for the first time in 31 years. The EUIL amendment allowed the entry of new independent power producers (IPPs) to supply electricity to the incumbent IOUs under long-term contracts. This is similar to the enactment of the Public Utility Regulatory Policies Act (PURPA) of 1978 in the USA. Between 1996 and 1999, following invitations for tenders for a total capacity of 6660 MW, tenders were proposed for 28,341 MW; the bid was eventually finalized for a capacity of 7380 MW. The principal bidders, who were large-scale steel and oil companies, revealed the potential for competition within the power sector. Further, the EUIL established “special electric utilities” that supply electricity in newly developed specified areas using their own facilities of generation and electric lines. At the same time, a “yardstick assessment” was applied to the incumbent IOUs on the occasion of their application for an increase in the rates. This new regulatory process provided a certain degree of incentives to improve managerial and operational efficiencies. In 1997, the EUIC initiated a debate on the introduction of retail competition to achieve internationally comparable cost levels in the ESI; this culminated in another amendment to the EUIL in May 1999, which resulted in a partial liberalization of the retail market. The reason for the government’s decision to implement partial liberalization was the careful thought that retail competition should be implemented considering the need for a balance between the nature of a public utility and competition in order to ensure universal service for customers, which would otherwise be jeopardized under fierce competition. Subsequently, the retail market was liberalized for extra-high-voltage customers with a total demand of 2000 kW or more as well those who were supplied electricity at extra-high-voltage levels (⭓20,000 V). New entrants that could supply electricity to these newly contestable customers are referred to as “power producers and suppliers (PPSs).” The PPSs could supply electricity to the eligible customers using the wheeling service that is offered by IOUs. The wheeling price is calculated based on costs using the activity-based costing (ABC) methodology. The partial liberalization commenced on March 21, 2000. A review of the achievements of this partial liberalization policy was scheduled to be conducted approximately 3 years subsequent to its implementation. The review process concluded that fairness/fair access and transparency in the transmission sector must be ensured, a nationwide wholesale market should be established, and a schedule for further retail liberalization should be drawn up. In response to this, the EUIL was further amended in June 2003 and the ESCJ was established as a neutral agency to ensure fairness and transparency in the transmission sector. At the same time, the JEPX was established as a wholesale power exchange. In addition, a schedule for expanding customer contestability was specified for 2004 and 2005, lowering the eligibility threshold to 50 kW. Table 18.3 summarizes the major events in the deregulation of the Japanese ESI. 18.3.2. The ESCJ When liberalizing a power market, new entrants must be assured a fair and nondiscriminatory access to the transmission and distribution infrastructures, which are monopolies by nature. To ensure that these policy requirements are met, the transmission and distribution sectors must by some means be kept separate from sectors that are open to competition.8 8

See, for example, Chapters 2 and 14 in this book for discussion on issues of transmission separation in the context of electricity deregulation.

624


Table 18.3. Major milestones in the Japanese ESI. Event

Date

Comments

The first amendment to the EUIL in 31 years

1995 – Amendment and Implementation

Correction of the disparity between domestic and foreign prices. Creation of competitive bidding system for new generating facilities to introduce the market mechanism into the wholesale power sector, which implies the introduction of IPPs.

The second amendment to the EUIL

1999 – Amendment 2000 – Implementation

Initiation of retail competition for the benefit of extra-high-voltage customers (⭓2000 kW and at ⭓20,000 V) to achieve the internationally comparable cost levels in the Japanese ESI.

The third amendment to the EUIL

2003 – Amendment

Further expansion of retail competition and the establishment of a wholesale electricity market (JEPX) and a neutral agency (ESCJ). Expansion of retail competition to include high-voltage customers (⭓500 kW). Commencement of operations of the JEPX and the ESCJ; expansion of retail competition to include all high-voltage customers (⭓50 kW). Initiation of a debate on the expansion of retail competition to include all customers, including households.

2004 – Implementation 2005 – Implementation

2007

ESCJ

Application METI Appointment

• Establishment of rules • Monitoring of rules and dispute settlement • Disclosure of grid information and central communication function of dispatching

Supervision and directive

Government

JEPX

Evaluation of transmission based on grid information Fig. 18.6. Mechanism of the ESCJ.

This separation can be achieved through unbundling of the accounting for each sector, legal unbundling through the formation of separate companies, and unbundling of ownership through the divestiture of specific functions to third parties. The EUIC debated possible methods of unbundling the transmission sectors in Japan and decided to limit this separation of the generation and transmission /distribution sectors to the unbundling of accounting with codes of conduct applied among the sectors. To compensate for this limitation, the ESCJ was established as a limited liability intermediate corporation. Figures 18.6 and 18.7 describe the mechanism and organization of the ESCJ, respectively, which reports to METI. The ESCJ comprises four groups: (a) incumbent IOUs, (b) new entrants


625

Members IOUs

Generators

PPSs

Scholars

General assembly meeting of members Selection of members

Controller

Audit

Selection of members

Administrative board

Commission of experts

Transmission/distribution function of IOUs

Proposal

Council

Central dispatching communication function

JEPX

Fig. 18.7. Organization of the ESCJ. Source: ESCJ (2005).

(PPSs), (c) wholesale utilities, IPPs, and private power generators, and (d) scholars. In order to secure fairness in operations, the IOUs, PPSs, the generators as stakeholders, have 3 voting rights for each and the scholars as neutral parties have 5 voting rights. A scholar is appointed as chairperson of the board. There are three principal roles for the ESCJ. These are as follows: First, establishing rules with respect to investment in the electric power system, network access, network operation, and disclosure of information; second, performing ex-post monitoring and establishing a dispute resolution system between IOUs as network owners and network users; and third, functioning as a central communication system for dispatching and disclosing grid information with respect to wide-area electricity trading, operation of interconnections, and arrangements for congestion management. The organization also sets rules that limit information exchange between the transmission sector and competitive sectors, restricts cross subsidization between these sectors, and restricts discriminatory activities against competitors. Although the rules established by the ESCJ are not enforceable, the discipline of function is ultimately ensured by the government because METI supervises the activities of the ESCJ. On this basis, the fundamental concept of the ESCJ is rather to govern and resolve problems among members independently. Therefore, if members violate the rules in order to cause some inconvenience to other members, penal provisions are applied to the firm that violates the rules and its name is disclosed in public. Such penalty has a strong incentive to make the members comply with the rules because once they are accused of an unfair act by the other members and the facts are disclosed in public, it causes extreme damage to the firm’s reputation, which results in creating a significant disadvantage for the future business of the firm. However, if an organization with such self-governance does not function appropriately in accordance with expectations, the government can intervene in the ex-post dispute settlement. At the current (early) stage of deregulation in Japan, it is difficult to comment on the effectiveness of the unbundling of accounting and the performance of the ESCJ. To discuss these issues further, we have to carefully keep track of the progress of competition and the achievement of deregulation. If the rules do not function as well as expected and allow the IOUs to use abusive behavior against new entrants, a revision to the current system, including a debate on

626


Spot market

Forward market

Standardized Market

48 half-hourly contracts One day ahead

12 months ahead 24-hour month contracts 14-hour daytime month contracts

Billboard Market

Fig. 18.8. Market structure at the JEPX.

the pros and cons of unbundling of accounting, might be needed. In this sense, the establishment of the ESCJ might be the first step, and we will need to carefully observe for a certain period of time whether or not the ESCJ with its limited authority will be effective in carrying out its functions. 18.3.3. The JEPX PPSs who provide electricity to end customers are required to maintain a balance between supply and demand within a predetermined range at an interval of 30 minutes. An imbalance charge is imposed on PPSs if they are unable to meet the balance requirement.9 Therefore, there arose a need for a system that balances supply and demand as well as manages the risk of imbalance. Consequently, a nationwide wholesale power exchange was developed in order to meet such requirements. The JEPX was established in November 2003 as a limited liability intermediate corporation like ESCJ with 21 fund-raising member companies that engage in the company operation, including IOUs and PPSs. This was because the EUIC proposed that a private institution is preferable to a public institution with respect to economic efficiency and from the viewpoint of being able to respond to the participants’ needs. Trading at the JEPX comprises spot and forward markets, as depicted in Figure 18.8. There are no financial contracts at the JEPX. In the spot market, electricity to be delivered on the following day is traded using 48 half-hourly contracts. Prices are determined by the intersections between the demand and supply functions that consist of biddings from sellers and buyers using the single-price auction. The participants can begin submitting bids 5 business days before the delivery date, up to 9:30 a.m. on the day prior to the delivery. Further, the forward market comprises a standardized market and a “billboard market.” In the standardized forward market, two types of monthly contracts are traded. One is a “base-type” contract for constant power generation for 24 hours, and the other is a “daytime-type” contract for constant power generation for 14 hours from 8:00 a.m. to 10:00 p.m. on weekdays 9

The balance charge is calculated by IOUs based on costs incurred for the balancing activity. The balance charge is required to receive authorization from METI. As observed in British Electricity Trading and Transmission Arrangements (BETTA) in the UK, the balance charge is usually considerably higher than the price of electricity in a normal energy transaction; thus, it consequently assumes a penalty character.


627

and on Saturday. The trading of new monthly products begins exactly 12 months prior to the delivery month and concludes 10 days prior to the beginning of the delivery month. Thus, a total of 24 products are constantly traded at the standardized forward market. As a basic principle of trading at the JEPX, the standardized forward market is a bilateral contract, which implies that the delivery and the settlement have to be concluded between the participants in line with each other’s agreements. The JEPX only contributes to match individual bids from participants using the trading mechanism of continuous sessions and evaluate availability of the transmission capacity for the trading, based on information from the ESCJ. The participants are required to register their generation capacity and load served for customers in advance to secure physical contracts, and they are allowed to trade within the registered volume (JEPX, 2005a). Meanwhile, the JEPX does not commit to the billboard market because its role is limited to only list messages from participants indicating their intention to buy or sell products with several conditions; for example, duration of delivery, shape of contracts, and ranges of volume and price. The JEPX charges participants a commission for the use of the billboard function. The billboard market enables participants to conduct tailor-made trading; however, there is little activity at this stage due to a lack of messages from participants. 18.3.4. Expansion of eligibility In April 2004, direct access was initiated for customers consuming 500 kW or more. In April 2005, the contestability was further expanded to include all high-voltage customers whose consumption was 50 kW or more. Policy discussions regarding the possibility of direct access for household customers will begin around April 2007. Figure 18.9 represents the dates of implementation, scope of customers, and the share of customers by volume of demand since March 2000. As of April 2005, 63% of the customers were eligible to choose their electricity suppliers as a result of the gradual introduction of direct access. In Section 4.1, we review the results of the retail competition on changes in retail electricity prices and the customer switching rate. 18.3.5. Characteristics of electricity liberalization in Japan The most notable characteristic of electricity liberalization in Japan is that the wholesale market was liberalized first, followed by the retail market. A multi-stage approach was adopted and a careful monitoring process was used to liberalize the retail market. Figure 18.10 depicts the current structure of the Japanese ESI. Separation of generation and transmission has been achieved in most of Europe and the USA through legal unbundling of transmission function with an effective combination of establishing independent system operators (ISOs) or transmission system operators (TSOs) (see other chapters in this book for the status and function of ISOs/TSOs in the liberalized electricity markets in other nations/regions). Meanwhile, in the Japanese liberalization model, no ISOs/TSOs were established. Instead, each IOU owns and operates its own transmission system and handles dispatch. On that basis, the ESCJ ensures a fair and nondiscriminatory access to the transmission network for all users of the grid. When evaluating whether generation and transmission should be further unbundled, it is necessary to consider not only anticipated benefit of fair access to the system but also to forecast the extent of the economic impact caused by the unbundling. For such policy debates including the establishment of independent regional transmission organizations,

628


March 2000

April 2004

April 2005

Large plant, department store, large buildings 26%

2,000 kW 20,000 V

26%

Medium scale plant 9%

40%

500 kW Grocery store, medium/small buildings 5%

63%

Small plant 9% 50 kW 6,000 V

Grocery store, medium/small buildings 14% Very small plant, small grocery store 5%

100 V–200 V

Household 14% Fig. 18.9. Eligible customers and expansion. Note: Percentage by electricity consumption.

Power exchange JEPX

IOUs

Gen

GEN

Wholesale utilities

IPP

Special utilities

PPS

GEN

Wholesale Transmission T&D Distribution Wheeling service

Rules and authority of “neutral agency” ESCJ

Supply obligation

Retail

Small customers

High voltage customers (50 kW)

Extra-high voltage customers (2000 kW)

Still regulated Liberalized Fig. 18.10. Structure of the Japanese electric power industry as of 2005.

Customers


629

a cost–benefit analysis may be useful.10 For example, there are several benefits and costs of unbundling. Such benefits include enhancement of independent network management, improvement of effectiveness and efficiency of regulation, enhancement of retail competition, and promotion of wholesale competition. Meanwhile, the costs of unbundling include rising transaction costs, loss of economies of scope, and temporal effects that lead to a decrease in investment.11 With regard to the global movement toward deregulation, a series of occurrences influenced the deregulation trend. These occurrences include the electricity crisis in California from the Summer of 2000 to the beginning of 2001, the Enron bankruptcy in 2001, and the large-scale power outage in New York in 2003. We cannot conclude that they were caused by deregulation itself; rather, they may be closely related to the institutional design of the deregulation. However, these occurrences affected the attitudes of many states in the USA toward deregulation because no state has passed a restructuring legislation since June 2000, when the California and Western power crisis were just emerging (Rose and Meeusen, 2005). Besides, Japan learned from the experience of foreign countries that the wholesale power market was highly volatile and that there were frequent price spikes. In particular, as a result of the California electricity crisis, it has been recently considered in Japan that a complete unbundling of vertical integration should be undertaken with extreme caution. This is probably because the strategic behavior of generation companies to exercise market power to inflate wholesale prices was considered to be problematic.12 After observing a series of turmoil such as the collapse of the California market, policy makers in Japan have become more concerned about the potential unintended consequences of introducing radical market reform initiatives and have adopted a slower, even more cautious approach.

18.4. What Has Deregulation Changed? In this section, we document several factors that resulted from the deregulation by observing data pertaining to IOUs and the entire ESI. Since details of the Japanese experience have not been fully accumulated, it is too early to derive conclusive results based on the limited data and experience till date; however, in the section below, we will assess the influence of market reform on retail prices, the customer switching rate, the development of the wholesale power market, as well as the management practices in the IOUs till date. 18.4.1. Retail electricity prices and supplier switching rate Table 18.4 indicates the levels of system average prices for household customers and commercial and industrial customers in 1995, 2000, and 2003. All price levels are classified into nominal and real terms for each customer group. In addition, the table separately indicates 10

Michaels (2004) refers to the importance of a balanced discussion about the deregulation. Particularly, he indicates that we have forgotten advantages of vertical integration in the policy discussion about unbundling, a possible loss of economies of vertical integration. Chao et al. (2005) propose a “Third Way” between the extremes of permanent vertical integration and instantaneous liberalization of wholesale and retail markets. 11 Mulder et al. (2005) investigate these benefits and costs. 12 See, for example, Borenstein et al. (2002) and Joskow and Kahn (2001) with regard to the issues of market power in California.

630


Table 18.4. Price changes after 1995. Nominal prices

1995 (yen/kWh) 2000 (yen/kWh) 2003 (yen/kWh) Rate of decrease from 1995 to 2003 (%) Rate of decrease from 2000 to 2003 (%)

Real prices

Household

Commercial and industrial

Household

Commercial and industrial

24.6 23.1 21.5 12.7 6.9

16.9 15.4 14.1 17.0 8.9

25.0 23.1 21.9 12.3 5.1

16.3 15.4 14.8 9.1 4.0

Data source: Handbook of Electric Power Industry (1996, 2001, 2004).

the rate of decrease in the price levels from 1995 to 2003 and that from 2000 to 2003. It should be noted here that the average prices in Table 18.4 are calculated by using the revenue and demand pertaining to both regulated and eligible customers who are supplied electricity exclusively by IOUs. In other words, these prices do not reflect the effects of PPSs and mask the individual changes in rates for eligible customers. In nominal terms, the prices for household customers decreased by 12.7%, from 24.6 yen/ kWh (approximately 22 cents/kWh, converted at an exchange rate of 110 yen/dollar) in 1995 to 21.5 yen/kWh (19.5 cents/kWh) in 2003, while those for commercial and industrial customers decreased by 17.0%, from 16.9 yen/kWh (15 cents/kWh) in 1995 to 14.1 yen/ kWh (13 cents/kWh) in 2003. On the other hand, in real terms, the prices for household customers decreased by 12.3%, from 25.0 yen/kWh (23 cents/kWh) in 1995 to 21.9 yen/kWh (19.9 cents/kWh) in 2003, while those for commercial and industrial customers decreased by 9.1%, from 16.03 yen/kWh (15 cents/kWh) in 1995 to 14.8 yen/kWh (13.5 cents/kWh) in 2003. With regard to the rate of decrease from 2000 to 2003, the prices decreased by 6.9% and 8.9% in nominal terms and by 5.1% and 4.0% in real terms for household customers and commercial and industrial customers, respectively. The percentage rates of decrease for household customers over this period are almost comparable to those for commercial and industrial customers, although retail competition was not introduced among household customers. The reasons for this can be explained as follows: The IOUs made efforts to comply with the government policy to achieve internationally comparable cost levels, which was discussed in the EUIC from 1997, before the second amendment of the EUIL. In addition to this, it can be assumed that the IOUs took preemptive action against further development of deregulation on account of experiencing increasing pressures of competition. Next, we investigate the price developments with regard to only the extra-high-voltage contestable customers who are supplied electricity by both IOUs and PPSs, as shown in Figure 18.11; thus, the effects of competition on electricity rates are more evident in this figure as compared with Table 18.4. Figure 18.11 separately describes the average prices for all extra-highvoltage customers, extra-high-voltage industrial customers, and extra-high-voltage commercial customers from 2000 to 2004.13 It is evident that the prices for commercial customers decreased by 26%, from 17.36 to 12.87 yen/kWh, between 2000 and 2004. Although the rate of decrease for industrial customers is moderate as compared with that for commercial customers, the price in 13

Statistics are compiled from results of complete enumeration with respect to the payment and volume of electricity in the first half of each year.

631


18

Industrial

Commercial

15.76

14.3

Total

17.36 16 13.67

Yen/kWh

14 12 10

12.87 11.69 11.04

11.55 11.02

11.1

10.68

10.37

10.67

10.29

9.98

2002 Year

2003

2004

8 6

2000

2001

Fig. 18.11. Price developments only for extra-high-voltage customers. Data source: METI, News Release (2001–2005).

2004 is approximately 10% lower than that in 2000. The reason for a larger decrease for commercial customers than that for industrial customers is that the rates for commercial customers were relatively high in the past and there were consecutive requirements from some representative commercial customers to reduce their rates. This implies that there was significant room for competition in commercial customers. Meanwhile, with regard to rates for industrial customers, there were some special contracts to provide lower prices before the deregulation, in which IOUs and customers agree with each other to reduce loads when it is required, such as in case of an emergency. Therefore, the rates were already low and there was not much room for competition as compared with the situation of commercial customers. These decreasing trends are almost consistent with the price developments for commercial and industrial customers, which are observed in Table 18.4, indicating that the price levels significantly decreased after deregulation for all customers, particularly for contestable commercial customers. As of December 2004, the supplier switching rate by eligible customers was 2.3% for the nation as a whole and 4.5% for the area serviced by TEPCo. The supplier switching rate in Europe ranges from 0% to over 20% in 2003 for industrial customers, and commercial and household customers, respectively, depending on each country’s conditions (EU Commission, 2005). The switching rate in Japan is low as compared with those of several other countries; however, this rate is not directly comparable because of Japan’s specific conditions. In other words, the gas pipeline facility is not fully developed, except in a few metropolitan areas; thus, new entrants with gas turbine technology find it difficult to obtain price competitiveness as compared with the IOUs. On the other hand, in Europe and the USA, the main technology of new market entrants is gas turbines and pipeline networks are well developed; thus, they provide easy access to natural gas. Consequently, as compared with Japan, natural gas prices are relatively low in these two regions, which led to the fact that the power supplied by new entrants remained competitive in the 1990s under relatively stable gas prices.14 14

However, their competitive position weakened with the recent price hikes in natural gas.

632


3.5 PPS supply for high-voltage customers: 500 kW–2000 kW PPS supply for extra-high-voltage customers: 2000 kW Ratio of PPS supply for all eligible customers Ratio of PPS supply for extra-high-voltage customers Total PPS supply

3.0

600

500 2.0

400 1.5

300

Ratio of PPS supply (%)

2.5

1.0

200

2005/02

2004/12

2004/10

2004/08

2004/06

2004/04

2004/02

2003/12

2003/10

2003/08

2003/06

2003/04

2003/02

2002/12

2002/10

2002/08

2002/06

2002/04

2002/02

2001/12

2001/10

2001/08

2001/06

2001/04

2001/02

2000/12

0.0 2000/10

0 2000/08

0.5

2000/06

100

2000/04

Volume supplied by PPS (GWh)

700

Fig. 18.12. Volume and share of PPS supply. Data source: METI, News Release (2000–2005).

Figure 18.12 describes the changes in volume supplied by PPSs for extra-high-voltage customers (⭓2000 kW in demand) from April 2000 and those for high-voltage customers (⭓500 and ⬍2000 kW) from April 2004. In addition, the figure indicates the percentage ratios of electricity supplied by PPSs for the eligible customers according to the category they belong to in terms of total demand. The ratios are measured for extra-high-voltage customers and all eligible customers. The figure reveals an increasing trend in supplier switching since April 2000. The volume supplied by PPSs to extra-high-voltage and high-voltage customers in March 2005 collectively reached 683 GWh/month. However, the total share of PPSs continued to be significantly low, 2.4%, during the same period. If measured only in terms of the extra-high-voltage customers, whose switching rate is higher than that of the high-voltage customers, the share of PPSs supply becomes 3.3%; however, it still remains at a very low level. It can be pointed out that the switching rates in Japan to date are too low compared with those of more mature markets; however, their growth rate is highly encouraging and clearly suggests that there is potential for growth in competition in the future. Meanwhile, a number of PPSs entered the electricity market and commenced operations. As of July 2005, 23 companies have notified METI of their PPS business. Table 18.5 lists the major PPSs with their major investor/category of business and total capacity, based on data from METI (METI, 2005a). Various types of companies, such as a general trading company, an oil company, and a paper manufacturing company, entered into the PPS business. The increasing number of new competitors entering the market represents another promising sign of increasing competitiveness. It is important to note that the situation of the customer switching varies depending on each electric power company. For example, in the supply area of TEPCo., there are seven PPSs in

633

A New Stage of Electricity Liberalization in Japan Table 18.5. Major PPSs as of July 2005. Company name

Major investor/category of business

Diamond Power Marubeni Corp. eRex Nippon Steel Enet Summit Energy Daio Paper Nippon Oil Oji Paper

Mitsubishi Corp. General Trading Company Nittan Capital, Ueda Yagi Tanshi, Hitachi, Toshiba Steel Company NTT Facilities, Tokyo Gas, Osaka Gas Sumitomo Corp., Sumitomo Joint Electric Paper-manufacturing Company Oil Company Paper-manufacturing Company

Total capacity (MW) 618 905 446 358 1351 1445 572 278 268

Source: METI (2005a).

operation and they obtained 660 customers with a total demand of 1900 MW as of 2004 (METI/ANRE, 2005a).15 The second, and third, highest PPSs shares are observed in the service areas of the Kansai Electric Power Company and the Chubu Electric Power Company, where PPSs obtained 168 customers with a total demand of 440 MW and 36 customers with a total demand of 100 MW, respectively. In particular, more than 25% of the electricity is supplied by PPSs if we consider the extra-high-voltage commercial customers in TEPCo.’s area from April to December 2004 (METI, 2005b). For Kansai and Chubu, the figures are approximately 15% over the same period. In summary, although the nationwide customer switching rate is still low and even with the limited competition introduced so far, the IOUs are feeling the pressure and are taking steps to reduce/maintain costs and offer better prices to retain their customers.16 18.4.2. Overview of the JEPX So far trading volume at the JEPX has been marginal, not exceeding 0.1% of the total demand (Fig. 18.13). The primary reason for the low liquidity is that the wholesale electricity market began operations only in April 2005 and is still in the initial stages of its market operations. Thus, most trading members are cautiously studying the behavior of the other members and the market developments with regard to prices and volumes. Another reason for the low liquidity is related to the price ranges submitted by sellers and buyers. In terms of the volume submitted to the market, the selling order is continuously higher than the buying order. Specifically, the volume of selling orders accounts for 30,000–40,000 MWh/day, which is several times larger than that of buying orders depending on the demand–supply relationship. This reveals that there is sufficient supply capacity; however, most bidding prices of buyers and sellers deviate from each other, being higher 15

ANRE: Agency for Natural Resources and Energy. However, we should continue to carefully watch the trend in electricity rates because it partially depends on the trend of fuel prices, which have recently increased. Further, since the managerial effort to decrease investment is reaching a critical limit, the investment will increase in the future to maintain the current high level of reliability. In particular, investment issues have significant implications for the future deregulation policy, because reliability is more important for Japanese customers from a perspective of customer satisfaction as compared with customers in other countries (Ariu, 2005). 16

634


0.10 0.08 0.06 0.04

2005/7/31

2005/7/24

2005/7/17

2005/7/10

2005/7/3

2005/6/26

2005/6/19

2005/6/12

2005/6/5

2005/5/29

2005/5/22

2005/5/15

2005/5/8

2005/5/1

2005/4/24

2005/4/3

0.00

2005/4/17

0.02

2005/4/10

Percentage ratio of JEPX trading volume to total demand

0.12

Year Fig. 18.13. Developments in the daily share of the JEPX trading volume to the total demand. Data source: JEPX data (2005).

among sellers and lower among buyers. Therefore, the demand and supply curves do not necessarily intersect and the resulting contracted volume at the JEPX spot market becomes extremely low. However, such a low liquidity at a formative stage is not an intrinsic feature of only the Japanese market; in other nations as well, markets usually have a low level of liquidity in the initial stages of their operations, and the volume gradually increases over the course of time. As of the end of June 2005, nine contracts have been signed, totaling 11,774 MWh in the standardized forward market (JEPX, 2005b). According to the results of the questionnaire survey conducted on members of the JEPX, the predicted volume of daily spot trading is expected to reach 1000 MWh (360 GWh/year) in the first year of operation, and the total volume is expected to become 4000 GWh 6 years later, when income and expenditure will be balanced (JEPX, 2005b). However, even if this expectancy is achieved, the total volume still remains at a low level that accounts for approximately 0.5% of the total demand for the nine IOUs in 2004. As a measure to increase the trading volume, the JEPX is considering an enlargement of the number of trading members and increasing the variety of forward contracts based on members’ needs.

18.4.3. Changes in the corporate management of electric power companies After the deregulation, several changes occurred in the corporate management of IOUs. These changes should be carefully considered in order to evaluate the outcome of deregulation because it directly affects the progress of industry-wide efficiency and the enlargement of consumers’ surplus through changes in electricity rates, which are the ultimate goals of deregulation.

635


Number of employees Total demand per employee

0

2003

120,000

2002

1000

2001

125,000

2000

2000

1999

130,000

1998

3000

1997

135,000

1996

4000

1995

140,000

1994

5000

1993

145,000

1992

6000

1991

150,000

Total demand (MWh) per employee

7000

After deregulation

1990

Number of employees

155,000

Year Fig. 18.14. Developments in the number of employees and the total demand per employee. Data source: Handbook of Electric Power Industry (1991–2004).

Figure 18.14 indicates the developments in the total number of employees and the total demand (MWh) per employee from 1990 to 2003 for the nine IOUs.17 The total demand per employee is often referred to as a benchmark for the productivity (labor productivity) of companies and industries. Based on the statistics, the labor productivity has become approximately 1.3 times since 1990, from 4638 (MWh per employee) to 6214 (MWh per employee). The growth of productivity was maintained even under the conditions of economic stagnation and a sluggish increase in electricity demand after 1990. This is attributed to the fact that the number of employees decreased rather than increased over the period. In fact, the number of employees increased to 1995 at one point, but decreased thereafter. This decrease can be attributed to the managerial efforts aimed at improving the efficiency of the firm since the beginning of deregulation in 1995. Figure 18.15 illustrates the total expenditure on the extension work of nine IOUs from 1990 to 2003 for each function of generation, transmission/distribution network, and others. From the figure it is clear that the amount of expenditure decreased sharply after 1995, from 2828 billion yen in 1995 to 847 billion yen in 2003. The cutback in the capital investment cost is drastic; it is less than one-third of that in the peak period. The shares of the expenditure for generation and network are almost equal. This decline is correspond to the slowdown in electricity demand in recent years as well as being further evidence of managerial efforts for improved efficiency of the Japanese electric power companies that they reduced their investment cost in preparation for increased competition under deregulation. 17

These statistics cover all IOUs except the Okinawa Electric Power Company. Okinawa is not included because it underwent a different deregulation process as compared with the other nine companies due to its status as a state-owned utility until 1988. This section does not include Okinawa in the statistics provided.

636


Others Network Generation

1000 million yen

2500

After deregulation

2000 1500 1000 500 0

1995

1990

2000

2003

Fig. 18.15. Expenditure on extension work by the nine IOUs. Data source: Handbook of Electric Power Industry (1991–2004).

41.5

Thermal efficiency

41.0 40.5 Percent

40.0 39.5 39.0

After deregulation

38.5

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1990

37.5

1991

38.0

Fig. 18.16. Developments in thermal efficiency. Data source: Handbook of Electric Power Industry (1991–2005).

In addition, a strategy of reducing debt is clearly indicated after deregulation, which is similar to the decreasing trends observed in the number of employees and the amount of capital investment. The total amount of interest-bearing debt for the nine IOUs began to decrease significantly from 1999, along with a 3-year stagnation period just before the year. The total amount of debt was reduced by approximately 20% from 1995 to 2003. Such decreasing trend indicates that the IOUs began to improve the financial position of them in preparation for the progress of competition, because their high degree of indebtedness was often pointed out as a problem by investors and market analysts. Lastly, Figure 18.16 indicates the change in the thermal efficiency over the observation period. After a gradual increase from 1990 to 1995, the thermal efficiency rapidly improved from 39.01% in 1995 to 40.94% in 2004. This improvement is partly explained by the technological progress achieved by the introduction of highly efficient combined-cycle gas turbine


637

generation plants in Japan. In addition, the improvement is attributed to the management efforts to abolish the degraded, low-efficiency, high-cost fossil fuel plants under the pressure of competition. From these figures, it is evident that the Japanese electric power companies modified their management practices aiming to improve business efficiency after deregulation in 1995. These changes are consistent with the decreases in electricity rates over the same period. Thus, it is easily estimated that the benefits of these reduced costs were finally passed on to the consumers through the reduction of electricity rates, as indicated in Section 18.4.1. 18.5. Continuing Issues As the preceding discussion indicates, the Japanese market reform initiatives are relatively new and many of their provisions have not been in place long enough to produce sufficiently clear trends or results till date. Additionally, Japanese policy makers and the industry are still debating on the future direction and scope of reform where a number of important issues remain unresolved. This section provides an overview of four of the main issues. 18.5.1. The wholesale power market The policy change has several implications for electric utility restructuring. The first concern is a trading ability that enhances liquidity in the spot trading volume at the JEPX, which currently accounts for less than 0.1% of the total electricity demand in Japan. This issue can be easily identified by comparing the trading volumes with those of other nations. For example, the trading volume is approximately 40% of total demand at Nordpool, and approximately 10% at Germany’s European Energy Exchange (EEX).18 Based on the experience of other countries, we know that unless the liquidity of the spot market is increased, it will be difficult to develop reliable price indexes to deliver appropriate price signals to the market. Thus, the liquidity of the spot market should be increased because the establishment of a robust wholesale power market is a key factor for the success of electricity liberalization. This is an important challenge that requires the efforts of market players. Relating to this problem, it should be considered that, to begin with, we have to explore the degree of liquidity that is required in the spot market for the purpose of obtaining the appropriate price signals. The examples of England and Wales reveal that a market perception of prices will be formed through active bilateral forward transactions. This appears to be a more practical measure in Japan because under its current system consisting of long-term bilateral transactions and voluntary power exchange, it is generally believed that most of the market participants prefer to trade energy by using bilateral forward contracts.19 Consequently, under the current system centered on bilateral trading, it might not be possible to increase the liquidity of the power exchange by a large extent and the spot market may be unable to deliver a sufficiently reliable price index. This problem can be noted from a technical perspective that physical limitations of interconnection capacities need to be resolved to change the current market structure and increase the liquidity at the power exchange. However, slight indication of an increasing trading volume is observed at the JEPX. Key players with regard to increasing the liquidity include wholesale utilities like Dengen Kaihatsu. 18

See the Nordpool website, http://www.nordpool.no/nordpool/spot/index.html, and Maibaum (2003). As noted in Footnote 2, a significant amount of electricity is traded by long-term bilateral contracts between electric power companies and wholesale generators. 19

638


Currently, they are expected to become active players in the spot markets. Dengen Kaihatsu has recently reviewed its long-term power purchase contracts with electric power companies and is selling the power formerly contracted in the spot market. Therefore, further developments in the spot market should be closely monitored. Another important issue for activating the JEPX is to increase the variety of products. It is important to ensure that not only the spot market but also the forward market is designed to be active in terms of trading amounts. Currently, the standardized forward market offers trading of monthly forward contracts for 24- and 14-hour constant power. In the near future, we need to develop a variety of products related to timing (year, season, month, week, daily contracts, etc.) and type (base, peak load contracts, etc.) in order to satisfy the needs of the market participants. It is evident that forward trading occurs both inside and outside the power exchange, and both are important to foster wholesale power trading. For the purpose of increasing the variety of products, we may consider the introduction of futures (financial instruments) in the years ahead. Futures are traded at Nordpool and EEX, which have relatively high liquidity. The UK Power Exchange (UKPX) in England and Wales also offers futures, although their liquidity is relatively low. In Japan, futures are not yet offered at the JEPX. Moreover, it is also likely that power trading is handled by a variety of brokers. It can be expected that the competition between the power exchange and brokers will increase the market activity. Lastly, the question of how to increase the number of market participants is pointed out as an important issue for the JEPX. There are 28 participants in Japan as of July 2005, while Nordpool has more than 350 companies, EEX has more than 100, and UKPX has more than 40. However, as the transmission network in Japan has been isolated from those of foreign countries, we cannot expect to increase liquidity through international trading, as is done in Europe. Private power generators and PPSs that have not enrolled as JEPX members are prospective participants in the future. 18.5.2. Governance of the ESCJ The board of the neutral agency principally consists of stakeholders. Based on the previous decisional experience of ISOs in the USA, it appears that their decisions are largely driven by politics and consensus among diverse interest groups. In that case, their decisions might be biased and they do not necessarily provide an outcome that delivers maximum social welfare. To avoid this, in Japan, scholars are appointed as board members in addition to the stakeholders; however, the stakeholders form the majority. Scholars are expected to express neutral opinions in order to arrive at reasonable decisions. They are also expected to analyze and objectively evaluate valuable lessons that can be learned from electricity liberalization throughout the world. Their efforts will contribute to an appropriate electricity restructuring plan for the Japanese power industry, particularly if they investigate various issues related to the unique features of the electric power industry. It is too early to evaluate the effectiveness of the ESCJ at this stage. We limit ourselves to the statement that we can expect the unique role of scholars in the board, which is to be important for the successful functioning of the ESCJ. 18.5.3. Full retail liberalization In Japan, there is an ongoing debate on the desirability of eligibility expansion. A discussion will commence around 2007 on whether and how to expand eligibility. We would like to briefly comment on the issue of full retail liberalization.


639

A decision regarding whether the household customer should be allowed eligibility depends on the degree of competition in the wholesale power market, marketing, meter installation and billing costs, the costs of supplier switching, and the degree of competition between suppliers at the retail level. Besides, the circumstances in other countries serve as a reference.20 The experiences of other countries reveal that the economic impacts of full retail liberalization might be very small if we implement it in Japan. Particularly, the most important factor among several conditions to make full retail liberalization truly effective is the meter installation cost. For household customers, the cost to install a real-time meter should be less than the expected benefit from energy savings. Some economists discuss that greater economic benefit and convenience should be attained through the further development of real-time meters in order to expand eligibility to household customers.21 Currently, for household customers, the cost to install a real-time meter is considered higher than the expected benefit from energy savings. In addition, since profiling reduces the incentive to develop real-time meters, caution should be exercised regarding the introduction of profiling. However, the meter costs will be reduced owing to technological development as well as scale merit, which would result from the worldwide expansion of the liberalized retail market in the near future. The intensification of retail competition will foster the vitalization of the wholesale market. Such interaction between wholesale and retail competition is essential to further the progress of the liberalization in Japan. 18.5.4. Future of nuclear power22 The long-term plan of the electric utility companies in Japan is to build 14-GW nuclear power plants by 2014. This plan is essentially consistent with the government’s energy policy; METI recently emphasized the importance of nuclear generation from the perspective of energy security and the prevention of global warming (METI, 2005c). There are several reasons behind this: concerns about a future rise in the prices of fossil fuels, and economic efficiency and reliable power supply using renewable energies. The recent indication of a revival of nuclear generation in Europe and the USA affects the plan as well. Further, the Japanese Government expressed that the share of nuclear energy in the generation mix should be in the range of 30–40% in the future, which is higher than the current share of 30% (METI/ANRE, 2005b). There are serious debates in Japan regarding the impact of further competition in the electricity market on the development of nuclear power plants and the kind of governmental intervention that will be required in order to ensure steady nuclear power development. Nuclear power has certain unique characteristics, namely: 1. High up-front investment costs. 2. High uncertainty regarding the costs of high-level waste disposal and decommissioning of the power plants. 20

In the USA, for example, only two states have a percentage of residential load switching greater than 10% in 2005. One such state is Ohio, where most of the residential switching has been through the state’s aggregation programs. The other is Texas, which is now the most active state in the country in terms of residential customers choosing a supplier. In most states the figure is well below 5%. Nine states are at or near 0% (Rose and Meeusen, 2005). In Texas, the prices of IOU-related companies in the Electric Reliability Council of Texas (ERCOT) area are regulated as “prices to beat” that are set to be relatively high as compared to those of competitors (Zarnikau and Whitworth, 2005). 21 For example, see Joskow (2000). 22 This section is based on Yajima (2004).

640


3. The necessity of having long-term investment guarantees normally provided under rateof-return regulations to vertically integrated IOUs. In this context, what kind of impact can liberalization have on the future of nuclear power? The main issues are the following: 1. The manner in which existing nuclear power plants in a liberalized market should be treated. 2. The kind of incentives that may be needed to sustain the future of the nuclear industry in Japan. In a more competitive environment, stranded investment may occur when revenues are unable to cover the sum of the variable and fixed costs (including investment costs for fuel procurement and processing, and waste disposal). However, the existing nuclear plants will survive if the government will allow the recovery of the stranded costs from the viewpoints of energy conservation and solving the global climate problems. Meanwhile, there appears to be a rather different perspective with regard to new investment in nuclear energy in a competitive environment. According to the modern investment theory, which incorporates the concept of option value, new nuclear power plants will not be built because the option value of each power plant is regarded as a premium for the flexibility of the plant’s operation. In such a case, gas-fired power plants, which usually have smaller capacities, are preferred to large nuclear power plants. Assuming that a certain total capacity is required and that the capacity of three gas-fired plants equals that of one nuclear plant, we choose to construct one natural gas plant as a first step. Later, we have time to be able to decide whether to proceed further with the additional two gas-fired plants. This procedure is useful to avoid taking risks of lump-sum investment. However, in the Japanese context, nuclear generation may need to be evaluated by considering both the environmental aspect and the aspect of the security of supply. Such a perspective of evaluation of the nuclear generation in the context of the security of supply varies among different countries/regions, principally due to restrictions of the lack of uniformity in the endowment of primary fuel. In other words, a country with rich primary fuel endowment has a wider range of choices to maintain an assured level of supply and is relatively less bound to the nuclear option. Meanwhile, a country with less (no) primary fuel endowment, like Japan, may have no option but to select nuclear energy for the purpose of ensuring a stable supply of a given level of electricity to end-users. In addition, the evaluation of nuclear generation in an environmental context depends on societal preferences and various risk assumptions: the risk of climate change as a result of CO2 emissions and the risk of nuclear hazards. Some countries/societies may therefore support nuclear generation under environmental considerations, while others may oppose it because of its potential hazards. In Japan, it has been pointed out that nuclear power development is the only large-scale economic solution to reduce CO2 emissions and to increase the security of supply. Recognizing a role of the nuclear energy in solving problems related to environmental protection and security, it is argued that the presence of market failures requires a higher share of nuclear investment than there is at present. Market failure is principally derived from the inability of markets to take into account both the external effects of fossil fuel combustion and the instability of fuel supply. This implies that large investments in nuclear power plants in more competitive markets should be induced by means of government subsidies, tax preference, etc. In other words, if such a market failure exists, the government can intervene in the market to promote nuclear power generation.


641

One of the significant problems of nuclear power generation is cost uncertainties. The main cost uncertainties arise from the nuclear fuel cycle, especially from the disposal of nuclear waste and decommissioning. The subcommittee for studying costs and other issues, which was set up under the EUIC, estimated the costs of the nuclear fuel cycle related to the operation of the Rokkasho Reprocessing Plant. The estimated cost amounted to 18.8 trillion Japanese yen (METI/ANRE, 2004). Moreover, the subcommittee for the study of the system and measures confirmed the necessity of developing an economically viable system to cover the expenses of the nuclear fuel cycle at the time of generation under the “polluter-pays” principle, based on the unique characteristics of the business. The EUIC proposed a revamp of the system in August 2004. The main feature of this proposal is that the past generation costs will be borne by all customers, including the customers of new entrants, and the future generation costs will be borne by all customers of the incumbent electric power companies. The costs arising from the operation of the Rokkasho Reprocessing Plant are recovered through non-bypassable charges added to the transmission charges (METI, 2004). However, the costs of the nuclear fuel cycle are not limited to these costs. For example, another reprocessing plant may have to be built in the future, and operating the plant might further increase the overall costs. In the future, it might be necessary for the government committee to further investigate how to share responsibility between the government and the market participants in terms of recovering such additional costs. For example, to solve this problem without weakening the polluter-pays principle, the government could provide security to potential investors by setting an upper limit on their financial burden with respect to the costs of the nuclear fuel cycle and decommissioning. If the costs exceed the upper limit, the government could bear the additional costs. In summary, although there are pending issues in terms of cost uncertainty and cost allocation, there is a general consensus among policy makers, electric power firms, and others that the fundamental energy policy in Japan is based on further promotion of nuclear power generation, which is consistently developed with the electricity deregulation. 18.6. Conclusions Electricity liberalization in Japan entered a new stage in April 2005 when the contestability was expanded to include all high-voltage customers. Since competition was first introduced into Japan’s electricity markets in 1995, liberalization of the industry advanced step by step from one level to the next. The electricity rates for eligible customers decreased considerably after the introduction of retail competition; the prices for contestable commercial customers decreased by approximately 26% from 2000 to 2004, and those for industrial customers decreased by approximately 10% over the same period. Although we adopt a very cautious approach toward deregulation and are only at the initial stage, the results of deregulation have thus far been evaluated as successful from the perspective of consumer welfare. However, several issues remain to be resolved to improve the liberalization scheme in Japan. In particular, liquidity in spot trading volumes on the JEPX should be enhanced to a large extent because it currently accounts for less than 0.1% of the total electricity demand in Japan. An efficient and robust wholesale power market is a key factor for the success of electricity liberalization. Therefore, further developments in the spot market should be closely monitored. In case liquidity cannot be increased as expected, the next step could be the introduction of virtual power plants (VPPs), which are adopted in France and in Texas in the USA. VPPs are considered to be an effective way to activate the wholesale market in countries like Japan, where it is considered legally difficult to unbundle the transmission function to an independent organization.

642


Finally, we should always take into account that our ultimate aim for electricity restructuring is to improve efficiency of the ESI that can provide us with reliable energy at efficient prices, which eventually leads to an increase in consumer welfare. Competition is merely a tool to achieve the aim, not our supreme goal. In order to enable this tool to function effectively, we have to maintain flexibility in our policies so that we constantly improve the institutional design based on participant discussion and consumers’ perspectives.

Acknowledgements The authors thank Dr. Sioshansi who provides us with helpful suggestions to improve the quality of this chapter. The authors also thank Masahiro Maruyama and Nobuyuki Yamaguchi for their help in obtaining data. The Research was conducted as part of a project on the deregulation of electric utilities, at the Central Research Institute of Electric Power Industry (CRIEPI) in Japan. The views expressed here should not be identified with those of CRIEPI.

References Ariu, T. (2005). Brand strategy and customer equity of electric power suppliers in the US. Report No. Y04008, Socioeconomic Research Center (in Japanese). Borenstein, S., Bushnell, J.B. and Wolak, F.A. (2002). Measuring market inefficiencies in California’s restructured wholesale electricity market. American Economic Review, 92, 1376–1405. Chao, H., Oren, S. and Wilson, R. (2005). Restructured electricity markets: reevaluation of vertical integration and unbundling. Harvard Electricity Policy Group Research Papers, available at http:// www.ksg.harvard.edu/hepg/Papers/Chao_Oren_Wilson_Reevaluation_Vertical_Integration_ Unbundling.pdf ESCJ (2005). Organization of the ESCJ, available at http://www.escj.or.jp/about_escj/organization/ about_escj_organization.pdf (in Japanese). EU Commission (2005). Annual Report on the Implementation of the Gas and Electric Internal Market. Brussels, 5.1.2005, COM (2004) 863. Federation of Electric Power Companies of Japan (2005). Electricity Review Japan 2004–2005, available at http://www.fepc.or.jp/english/index.html Federation of Electric Power Companies of Japan (1981–2005). Handbook of Electric Power Industry, published under the editorship by METI/ANRE (in Japanese). Japan Electric Association (2004). Current Situation of the Electric Power Industry, published under the editorship of METI/ANRE (in Japanese). JEPX (2005a). Guidelines for Trading at the JEPX, available at http://www.jepx.org/guide.htm (in Japanese). JEPX (2005b). Status of the Japan Electric Power Exchange, available at http://www.enecho.meti.go.jp/ denkihp/bunkakai/22nd/jepx_shiryo.pdf (in Japanese). JEPX data (2005). Available at http://www.jepx.org/transaction information. Joskow, P.L. (2000). Why Do We Need Electricity Retailers? Or Can You Get It Cheaper Wholesale? Revised Discussion Draft. MIT, Cambridge, MA, February 13. Joskow, P.L. and Kahn, E. (2001). A quantitative analysis of pricing behavior in California’s wholesale electricity market during summer 2000. NBER Working Paper Series No. 8157. Maibaum, O. (2003). European power exchange: its development in Germany and continental Europe. 8th Annual AEPX Conference, October 2003, available at http://www.theapex.org/documents/APEX03/ APEX_Meeting_2003.pdf METI (2004). On a Whole Concept of Institution and Measures for Reprocessing Business of Nuclear Generation, Interim Report of the Electricity Utility Industry Council, available at http://www.meti.go.jp/ feedback/downloadfiles/i40714bj.pdf (in Japanese).


643

METI (2005a). List of Power Producers and Suppliers, available at http://www.enecho.meti.go.jp/ denkihp/genjo/pps/pps_list.html (in Japanese). METI (2005b). On the Market Share of Electricity Retail Market, supporting document, available at http:// www.meti.go.jp/committee/downloadfiles/g50815a56j.pdf (in Japanese). METI (2005c). Perspective of Energy Demand and Supply in 2003, Committee Report of the Demand and Supply Council, available at http://www.meti.go.jp/report/data/g50328bj.html (in Japanese). METI/ANRE (2004). On the Investigation by the Subcommittee on Cost and Other Issues, Electricity Utility Industrial Council, supporting document, available at http://aec.jst.go.jp/jicst/NC/tyoki/gijyutu2004/ gijyutu03/siryo3.pdf (in Japanese). METI/ANRE (2005a). On the Status of New Electricity Supply Industry, supporting document, available at http://www.enecho.meti.go.jp/denkihp/bunkakai/22nd/shiryo6.pdf (in Japanese). METI/ANRE (2005b). Nuclear Power in Energy Policy, supporting document, available at http:// www.enecho.meti.go.jp/info/image/050318.pdf (in Japanese). METI, News Release (2000–2005). Bulletin of the Total Quantity of Electricity Demand. Available at http://www.meti.go.jp/press (in Japanese). METI, News Release (2001–2005). Summary of Survey Results on Electricity Demand and Prices. Available at http://www.meti.go.jp/press (in Japanese). Michaels, R.J. (2004). Vertical integration: the economics that electricity forgot. Electricity Journal, 17(10), 11–23. Mulder, M., Shestalova, V. and Lijesen, M. (2005). Vertical Separation of the Energy-distribution Industry: An Assessment of Several Options for Unbundling, CPB Netherlands Bureau for Economic Policy Analysis, CPB Document, No. 84. Nambu, eds. (2003). Institutional Design of Electricity Deregulation. University of Tokyo Press (in Japanese). OECD/IEA (2004). Energy Prices and Taxes. OECD/IEA (2005). Energy Balances of OECD Countries, 2002–2003. Rose, K. and Meeusen, K. (2005). 2005 Performance Review of Electric Power Markets: Update and Perspective. Review Conducted for the Virginia State Corporation Commission. Yajima, M. (2004). Liberalization of Japan’s electric power industry and nuclear power development. Nuclear News, May 2004, 10. Zarnikau, J. and Whitworth, D. (2005). Has electric utility restructuring led to lower electricity prices for residential consumers in Texas? Energy Policy (forthcoming).


Index

AB 1890, 332–335, 338, 339, 340, 366n Access holidays, 68–69 Access to market, 14, 24, 52, 56, 57, 65, 70, 184, 300, 302, 452, 467 Accounting theory, 221–222, 242, 587, 624 Acquisition, 46, 239, 252, 287, 288, 319, 337, 344, 348 ACR, 586–587 Active network competition, 63–64 Additional toll, 94–95 Affiliated Retail Electric Providers (AREP), 392, 395–396, 407, 408, 546 Alberta, 419, 442, 443, 444, 446 Alberta Electric System Operator (AESO), 446–447 Alberta experience, 447 Alberta Market Surveillance Administrator (AMSA), 445–446 Alberta restructuring experience, 442–447 Alberta’s Energy Utility Board (AEUB), 443 Allocative efficiency, 467, 487–488, 498 American Electric Power Company, 385, 408 Ancillary services (A/S), 88, 96, 101, 354, 393, 394, 397, 484n, 485, 494–495 Ancillary services markets, 5, 15, 19–20, 100, 101, 394, 396, 397, 403, 405, 452, 472–474 ANEEL, 574, 575, 580, 585 Arbitrage, 70, 95n, 184, 218, 267 Argentina, xxviii–xxix, 9, 92, 97, 102–105, 595, 598–605, 609, 610, 612, 613, 614 ARR market structure, 467 As-bid system, 336, 368 Assured energy, 577n, 578, 583 Auction model, 51, 131, 301, 444–445, 516, 517, 534–535, 536, 541, 587, 588, 589 Auction Revenue Rights (ARR), 454, 466–467, 516 Australia, xxviii, 40, 42, 173

Australian Competition and Consumer Commission (ACCC), 180, 185 Australian electricity generation, 174–176 Australian electricity market arrangements, 199–202 Australian Energy Market Commission (AEMC), 174, 180, 181 Australian Energy Regulator (AER), 174, 180 Average return on assets, 93 Balancing Energy Service (BES), 393, 394, 397, 401, 403, 406 Balancing market, 116, 133, 137, 163, 239 Balancing Mechanism (BM), 127, 130, 133n, 135, 136, 137, 141 Bargaining power, 131 Basic generation service (BGS) auction, 27 Basic toll, 94, 94n Behavioral rules in markets, 13, 458, 459, 462, 470, 510, 512 Belgium, 268, 269, 277, 284, 288 Bertrand–Nash behavior, 52, 58 BETTA, 44, 122, 109, 626n Bidding model, see Auction model Bidding strategy, 341–345, 347n Bilateral contracts, 9, 127, 186, 286, 331, 393, 458, 600, 619n, 627 Bilateral schedule, see Self-schedule Billing data, 397, 407, 412 Birka, see Fortum Blackouts, 20, 100, 353, 518–519, 605 Blue Book, 11, 330, 413 BNA, 247, 250, 253, 261 Bonneville Power Administration (BPA), 323 Brazil, 15, 29, 47, 565, 567, 568, 570, 571, 574, 578, 579, 585, 587, 589–590, 605 Brazilian ESI reform (2002), 585–586 Britain, 37, 109, 113, 122, 132n, 138, 139, 140 British coal, 21, 37, 116 British Energy, 112, 132

645

646

Index

British ESI, 109, 113, 115 British Gas, 136, 139 Brokers, 361, 638 Buffering hypothesis, 253 Bulk supply tariff (BST), 207–208, 111, 220 Bundesnetzagentur (BNA), 247, 250–251, 253 Business models, 52, 517, 558, 559 CAISO ancillary service market, 337–338, 342, 364, 367 CAISO real-time market, 337, 342, 343–345, 364 California, xix–xx, 7, 8, 10, 11–14, 16, 37, 40, 46, 145, 169, 257, 319, 369–379, 383, 393, 397, 401, 413, 431, 438, 483, 485, 486–487, 501–502, 503, 510, 554 California electricity crisis, 14, 29, 319, 629, 46, 348–350, 353, 358, 366, 369, 376, 430, 431, 438 California electricity restructuring, 46, 319, 326–328 California electricity system before restructuring, 320–324 California financial crisis, 348, 349, 361, 365, 367, 369, 379, 380 California Independent System Operator (CAISO), 330, 336–338, 361, 375, 487, 514, 515, 518, 520, 521 California market design, 46, 319, 554 California Public Utilities Commission (CPUC), 11, 13–14, 320, 324, 327, 328, 329, 330, 330–331, 332, 334, 335, 338, 339, 348, 365, 367, 372, 373, 374, 376, 377, 505 California PX, 335–336, 492 Canada, 358, 419 Canada electricity generation, 420 Capacity by fuel type, 113, 454–455 Capacity Credit Market (CCM), 468, 470–472 Capacity investment, 38, 162, 163–165, 390, 599 Capacity margin, 132, 160, 236, 261, 270, 289, 292–297 Capacity market, 467–472, 507 Capacity payment, 9, 22, 80n, 89n, 114, 118, 119, 127, 132, 136, 141, 184, 186, 505, 605 Capacity payment mechanisms, 9, 505 Capacity price, 78, 90, 92 Capacity resources, 452, 468, 469–470, 505n Capital markets, 177, 599 Captive customers, 36, 116, 117, 118 Cardoso ESI reform, 576, 579, 585 Cartel Office, 239, 242, 243, 246, 252 CDEC, 77, 78, 85, 87, 88, 89, 91, 92, 95n, 97, 98, 101, 103 CE electricity markets, 274, 302, 308 CE restructuring issues, 274–280

CE transmission issues, 272–274 Central Electricity Generating Board (CEGB), 21, 37, 110, 111–112, 120, 121, 122, 127 Central registration agent function, 384, 396, 411, 412, 413 Chamber for Electricity Trading (CCEE), 586–587 Chamber for Management of the Energy Crisis, see GCE Chile, 10, 11, 44, 77, 111, 480, 600, 605 Chilean ESI, 44, 78, 96, 105–106, 106, 596 Climate Change Levy, 129 CNE, 84, 88, 90, 92 CNE pricing model, 90 CO2 emission, 22, 165, 167–168, 640 CO2 emission trading, 255–256, 261 CO2 ETS, 167, 259 CO2 permits, 235, 236, 255, 256, 261 Coal, 37, 80, 111, 113, 116, 118, 174, 192, 196, 232, 236, 237, 259–260, 321, 455, 456 Coal-fired generation, 303, 422, 431, 435, 439, 579 Colombia, 595, 596, 598, 605–609, 610, 611–612, 613 Combined-cycle gas turbine (CCGT), 8, 113, 238, 636–637 Combined heat and power (CHP), 138, 236, 238, 243, 253–255, 261 Commercial customers, 26–27, 210, 333, 341, 410, 630, 631 Commercially Significant Constraints (CSCs), 395, 400 Commission, 388, 393, 398, 399, 401, 402, 412, 414, 534, 552 Competition, 36n, 41–43, 44, 45, 53n, 52, 55, 56, 57, 61, 109, 112, 117, 125, 458 Competition Act, 242, 239 Competition authorities, 53n, 58, 140, 305 Competition law, 49, 53 Competition transition charge (CTC), 330, 331, 332, 333, 340, 348 Competitive advantage, 160, 217–218, 333, 340 Competitive generation/production, 61 Competitive markets, 4–6, 13, 14–15, 19, 22, 23, 29, 49, 53, 138, 145, 265, 325, 328, 334, 336, 383, 389, 403, 431, 452, 458, 480, 499, 530, 533, 543, 554, 555, 556, 557 Competitive pricing, 89–90, 111, 117, 266–267, 347, 494 Competitive retail market, 219, 328, 334, 339, 349, 408, 531, 542, 545 Competitive retail power markets, 22, 529, 555 Competitive wholesale market, 22, 392–393, 454 Computer systems, 407

Index Concentration, 12, 80, 109, 140, 169, 188, 239, 246, 247, 286–288, 458–459, 470, 512, 547, 598, 606 Concentration ratios, 458–459 Conduct-impact test, 510 Congestion, 8, 9, 10, 19, 46, 302, 349, 364, 394, 395, 398, 413, 452, 464–466, 489, 496, 497, 513, 518 Conservation, 214, 364–365, 380, 435, 436, 437, 440 Consumer expectation, 432 Consumer information, 432–433 Consumer price outcomes, 193–194 Consumer response, 435–439 Contestability theory, 57, 221 Contestable markets, 49–50, 51–53, 55–56, 56–59, 221 Continental Europe (CE), 265, 267, 268, 269, 272, 273, 274, 277, 280, 286, 287, 288, 290, 292, 302, 307, 308, 309 Contract for Differences (CfDs), 116 Contract market, 11, 186–187 Contracts, 26, 27, 45, 47, 69, 90, 98, 99, 116, 117, 118, 127, 138, 141, 153, 154, 155, 157, 162, 164, 186–187, 189, 203, 219, 226, 319, 324, 326, 338–339, 371, 378, 438, 444, 458, 468, 497–498, 505, 513, 577, 584, 586, 587, 602, 612, 638 Corporate management, 634–637 Corporatization, 41, 211, 231 Cost-based approaches, 253, 247, 249, 250 Cost-based regulation, 248–249, 253, 473, 482 Cost-efficient operation, 164 Cost-of-service regulation, 17 Cournot–Nash behavior, 58 CPUC regulatory proceedings, 328–332 CPUC Restructuring Order, 328, 331, 332, 334 Cross-border trade, 246, 299n, 309 Cross-border transmission, 270, 297–299 Cross-subsidization, 4, 37, 160, 227, 239, 625 CTC mechanism, 531, 537, 540 Customer choice, 3, 12, 530, 554, 557 Customer education, 412–413 Czech Republic, 282 Dash for gas, 117, 118 Day-Ahead Energy Market, 20, 401, 454, 456, 457, 466, 469 Death spiral, 348 Debt–equity ratio, 225 Dec game, 497 Decentralized decision-making, 40, 440 Decentralized market-driven system, 111 Decision-making process, xxv, 66 Decommissioning, 116, 641

647

Decoupling, 340, 540–541, 578, 612 Decree 915, 573 Decree 1009, 573 Decree 5184, 586 Default contract, 157, 160 Default service, 23–27, 529, 530–533, 534, 541–542, 561 Default service dimensions, 531–532 Default service organizational models, 543 Default services pricing models, 533 Deliverability, 469–470 Demand adjustment limitations, 361 Demand and supply balancing, 162–163 Demand curves, 498–499, 506–507 Demand elasticity, 345 Demand response, 27, 373, 403–406, 414 Demand/supply development, 268–270 Denmark, 153, 154–155, 165, 166, 167 Department of Water Resources (DWR), 362–364, 366, 372 Deregulation, 7, 8, 41, 46, 78, 102, 105, 168, 326, 327, 328, 367, 379, 380, 551, 595, 596, 599, 605, 606, 609, 611, 613, 614, 618, 625, 629–637, 641 Deregulation process, 105, 168, 609, 620–623 Designing pricing models, 539–540 Developments comparison, 280–285 Disaggregated regulation, 43–44, 50, 51, 53, 54, 56, 67 Distribution, 3, 6, 44, 63, 77, 111, 117–118, 181, 192–193, 209, 274, 392, 569, 600 Distribution business, 178, 181, 185, 192 Distribution network operator (DNO), 250, 251, 254, 255 Distribution networks, 16–18, 62, 64, 226–230, 255, 272 Distribution restructuring, 216–219 Distribution value added (VAD), 92–93 Distrigas, 288 Divestiture, 346, 365 DNAEE, 571, 585 Domestic suppliers, 117–120 Dual retailer model, 545 Duopoly generators, 128–129 Dynamic efficiency, 488 Dysfunctional energy market structure, 361 ␧ sunk cost argument, 59 ECNZ (see also NZED), 207, 209, 211, 224, 225, 230 Economic efficiency, 208, 220, 327, 487, 488, 588, 626 Economic forces, 353–364 Economic policy-making, 221–222 Economic withholding, 508

648

Index

Economies of scale and scope, 51n, 53, 489 Edison Mission, 129 EDP, 284 Effective competition, 50, 53, 57, 58, 117, 122, 265, 287, 307 Efficient pricing, 219–221 El Paso pipeline, 358 Electrabel, 268, 282, 284, 287, 288 Electric crises, 96–105 Electric Power Boards (EPBs), 216 Electric power companies, 634–637 Electric power industry, 618–620, 638 Electric power market, 235 Electric reliability council of Texas (ERCOT), 44, 46, 383, 384–387, 391–392, 392–393, 394, 395, 396–397, 397–398, 399, 402, 403, 405, 406, 406–411, 412 Electric Utilities Act, 443, 444, 446 Electric Utility Industry Law (EUIL), 623 Electricity Act 1989, 112 Electricity Commission, 230 Electricity Corporation of New Zealand (ECNZ) (see also New Zealand Electricity Division (NZED)), 207, 209, 211, 224, 225, 230 Electricity crisis, 349, 353, 366, 376 Electricity demand measures, 373–374 Electricity directive, 122, 140, 240, 267, 277 Electricity distribution pricing, 92–93 Electricity generation capacity, 102, 370, 375–376 Electricity generators, 341–343 Electricity import reduction, 357 Electricity industry, 173 Electricity Industry Reform Act 1998, 217 Electricity industry structure 1990, 210 Electricity industry structure 2004, 231 Electricity liberalization, 109, 617, 620–629 Electricity market reform, 44, 178–180 Electricity pool, 125–126 Electricity prices, 90–92 Electricity reform’s aims and assessment, xvii–xviii Electricity-related policies, 376–377 Electricity sector, 1, 7–8, 28–29, 61, 203, 611 Electricity supply, 169, 354, 439 Electricity supply industry (ESI), 35, 42, 77, 78, 81, 96–97, 109, 111, 113, 115, 207, 236, 239, 251–252, 570, 571, 574, 576, 580, 624, 642 Electricity Tariff Equalization Fund (ETEF), 195, 199–200 Electricity Task Force, 209 Electricity transmission, 62–63, 325 Electricity transmission pricing, 87, 93–96 Eletrobr·s System, 570, 571, 572, 586 Eligibility expansion, 627

Emission trading, 245, 255, 307 Emissions trading scheme (ETS), 165–166, 167–168 Empirical/econometric approaches, 69–71 EnBW, 247, 259 ENDESA, 84n, 98 End-to-end regulation, 67 End-user prices, 154, 169, 242, 254 ENEL, 282, 284, 285 Energy Act 2005, 247–251, 261 Energy conservation, 364–365 Energy emergencies, 353, 368, 369 Energy imports and exports, 462–463 Energy market design, 456–458 Energy market performance, 459–461 Energy market structure, 361, 458–459 Energy markets, 196, 401, 456–462, 468, 500, 506 Energy offers and bids, 492–494 Energy-only market, 21, 132–138, 261, 503 Energy-only prices, 257 Energy payment, 80n Energy policies, 45, 99–100, 101, 105, 106, 247 Energy Policy Act (EPACT), 325, 483, 516 Energy policy and investment, 235, 247 Energy pricing, 90, 405, 496–498 Energy Reassignment Mechanism (MRE), 577, 578 Energy recall right, 469 Energy sector, 61–64 Energy transfers, 88 ENRE, 599 Enron, 13, 29, 147, 362, 364, 379, 430 Entry barriers, 5, 136, 139, 140, 247 Environment, 21, 161, 165, 253–254, 256, 261, 327, 376, 545 Environmental issues, 306–307 Environmental limitations, 360–361 Environmental policy, 22–23, 45 Equilibrium price range, 124 ESCJ, 623–626, 638 ESI systems, 79, 80 Essential facility, 54–55, 66 EU E-Directive 2003, 240–241, 242, 246, 247, 250 EU Internal Market Design, 302 EU’s goal, 307–308 European Energy Exchange (EEX), 244 European RECS, 166n European Union (EU), 14, 15, 68, 69, 255, 279, 297, 299, 305 Ex ante mitigation, 510 Ex-ante regulation, 247–248 Ex ante screening, 508–510 Ex ante sector-specific regulation, 67

Index Ex ante spot price, 576 Ex post refunds/settlements, 510–511, 625 Export capacity, 121 Federal Energy Regulatory Commission (FERC), 16, 137, 325, 334, 335, 362, 367, 369n, 372, 384, 454, 472, 483, 508, 511–512 Federal Power Act, 482, 483 Federal regulation, 324, 367 Federal responses, 367, 369 FERC Order 888, 483–485, 472 FERC Order 2000, 485–486 FERC price mitigation order, 368–369 FERC rule making, 367–369 Financial crisis, 349, 361–362, 365–366, 369, 379–380 Financial transmission rights (FTRs), 466–467, 485, 515, 517 Finland, 22, 153, 154–155, 160–161, 167 Firm energy, 88, 466, 577n First Reform Process, 572–581 Flexibility, 43, 44, 52, 189, 305, 327, 328, 341, 378, 401, 406, 452, 507, 554–557, 576, 640, 642 Flexible power, 576 Flexible pricing structure, 66 Florence Forum, 305 Force Majeure, 97, 98, 99 Formula models, see Hub pricing model Fortum, 159 Forward market price model, see Hub pricing model Fossil-fuel-fired electricity generation, 332 Free market entry, 52 FTR market performance, 467, 517 FTR market structure, 467 FTRs allocation and trade, 515–516 Fuel factor model, 538–539 Fuel profile, 174–176 Full-cost uniform pricing, 220–221 Full retail liberalization, 129, 139, 140, 638–639 Game theory, 61, 71 Gas, 50, 61, 62, 64, 104, 118, 131, 165, 175, 177, 180, 198, 245, 247, 259, 287, 303, 309, 589, 590, 601, 603, 612 Gas consumption, 103 Gas-fueled CCGT, 259 Gas industry, 177, 178, 589 Gas infrastructure, 589, 590 Gas law, 589 Gas market, 358, 589 Gas Natural, 284, 301 Gas sector, 61 Gas supply industry (GSI), 251–252

649

Gas tax, 247 Gas turbines, 39 GCE, 582, 583, 584 Generally accepted accounting practice (GAAP), 221, 232 Generation, 10, 14, 26, 191–192, 211–215, 220, 222–224, 225, 230, 232, 238, 239, 240–241, 246, 255, 269, 286, 303, 325, 326–327, 346, 350, 352, 366, 367, 386, 392, 394, 423, 424, 439, 444, 454, 455, 456, 469, 472, 479, 514, 530, 534, 538, 541, 543, 569, 573, 576, 577, 600, 605, 606, 618, 619, 623, 624, 627, 635 Generation adequacy, 235, 236, 257–260 Generation by fuel type, 388, 455–456 Generation capacity, 160, 163, 164, 258–259, 270, 297, 303, 355, 502–507 Generation capacity limitation, 345, 354–355 Generation investment, xxi–xxii Generation market, 288, 388, 404, 489, 492, 507–508, 514, 602 Generation mix, 236, 237, 386–387 Generation owner, 14, 16, 456–457, 458–459, 469, 473 Generation plants offline, 359–360 Generation restructuring, 211–215 Generation supply curve 2004, 222 Generation/wholesale markets, 326–327 Generator market power, 16, 131, 188–190 Generator outage reporting requirement, 470 Generator performance, 456 Generators, 10, 11, 45, 78, 85, 88–90, 98–99, 100, 104, 105, 114, 116, 119, 121, 122, 124, 130, 140, 152, 168, 169, 181, 186, 203, 211, 230, 231–232, 346, 347n, 354, 360, 368n, 369, 375, 469, 492n, 495, 497–498, 506, 585, 587, 600, 602 German Energy Act, 249 Germany, 14, 15, 38, 235, 237, 240–241, 245, 258–259, 260, 261, 280, 287, 300, 301 GOL model, 89n Government, 11, 23, 35, 37, 41, 43, 44, 45, 81, 82, 84, 85, 99, 103, 104, 111, 112, 152, 164, 173, 174, 177, 178, 180, 182, 183, 185, 194, 195–199, 200, 203, 208, 209, 210, 214, 220, 223, 227, 230, 238, 251, 282, 301, 309, 324, 367, 377, 419, 420, 423, 426, 431, 433, 434, 435, 437, 438, 439, 441, 444, 569, 573, 578, 582, 612, 623, 640, 641 Government agency, 35 Government market interventions, 195–199 Government’s financial exposure, 434 Green electricity, 166, 167, 168, 349 Greenhouse gas abatement schemes, 196–198 Greenhouse gas emission, 199, 376 Grid losses, 603 Grid management, 159, 474

650 Grid pricing, 45, 224–226 Gross pool system, 186 Hands-off policy, 257, 258 Hausman–Sidak test, 65n Herfindahl Hirschman index (HHI), 123, 239, 402, 597–598, 598n, 602, 606 HHI calculation, 404 Hidrocantabrico, 284 High-voltage direct current (HVDC), 204 Higher quality reserve, see Quicker response reserve Horizontal integration, 4 Horizontal restructuring, xviii, 4, 8, 11 Hub prices, 497, 535–537 Hub pricing model, 535–537 Hydro-based system, 78, 81, 578, 588, 606 Hydro One, 423, 424, 431 Hydro stock, 150–152 Ideology, 39 Imperfect contestability, 56–58 Implementation problems of restructuring, 427–433 Import capacity, 153, 298–299, 427, 430 Imports/exports development, 270–272 Inadequate investment, 40, 186 Incentive-based regulation, 248–250, 253 Incentive regulation, 9, 17, 249, 251 Independent Market Monitor (IMM), 398n, 401, 412 Independent power producers (IPPs), 39, 118, 327, 346, 370, 375, 378, 483, 623 Independent System Operator (ISO), 4, 12, 116, 331, 388, 392, 394, 398–399, 413, 479, 480, 482, 484–485, 486, 488, 494, 496, 497, 498, 499, 502, 510, 511, 513, 514, 516, 518–519, 521, 537–538 Indicative Works Plan for Generation and Transmission, 88, 90, 99 Indirect component, 289 Industrial customers, 3, 11, 26, 210, 388, 631 Industry regulator, 230 Inflexible power, 576 Information technology, 520–521 Infrastructure, 28, 40, 54, 60, 66, 69, 393, 412 Innovation, 68, 80, 249, 521, 554 Insights, 7, 43–48, 77, 78, 106, 109, 617 Installed capacity shares, 568 Institutional arrangement, xxix, 7, 586–587 Institutional design, 113–117, 642 Institutional framework, 77, 251, 252, 302 Integrated utilities, 14, 38, 63, 112, 177–186, 194, 378, 392, 407, 414, 443, 485–486, 520, 578–579

Index Interconnection, 13, 46, 65, 92, 99, 267, 272, 278, 284, 297, 300, 301, 302, 323, 388, 451, 452, 463, 480, 571, 607, 619, 620 Interconnection capacity, 272, 284, 619n, 637 International gas pipelines, 102 Investment, 10, 11, 12, 18, 19, 20, 21, 29, 38, 40, 45, 68, 90, 93, 95n, 96, 99, 102, 105, 109–110, 137, 138, 140, 152, 160, 161, 163–165, 167, 169, 200, 236, 246, 247, 255–256, 258–259, 261, 307, 309, 326, 327, 327–328, 335, 370, 375, 388, 432, 434, 439, 440, 482–483, 507, 511, 512, 520, 521, 557, 557–558, 565, 569, 571, 572, 581, 583, 585, 589, 601, 603, 611, 613, 629, 633n, 640 Investor-owned utilities (IOUs), 320, 323, 334, 339, 340, 341, 345, 348, 362, 366, 373, 376, 385, 388, 390, 391, 392, 402, 618, 627, 630 IOU generating asset, 338 IOU status, 370–371 IOU wholesale electricity procurement, 374–375 Irreversible cost, 52, 54, 58, 59, 60, 63 ISO boundaries, 486, 510 ISO demand-response programs, 500–502 ISO formation and developments, 485, 487 ISO market redesign, 375 ISO performance, 519–522 ISO real-time markets, 336–337 ISO spot markets, 510 ISO transmission markets design, 513–518 ISOs and reliability, 518–519 Japan, 38n, 47–48, 617, 618, 619–620, 621, 622, 624, 625–626, 627–629, 631, 632, 634, 635, 637, 638, 639, 640, 641 JEPX, 623, 626–627, 633–634, 637, 638, 641 Kelman Report, 583, 588 Laissez faire, 43, 199, 387 Last Resort Service, 530, 532, 533, 539, 541 Law 8631, 573 Law 8987, 573 Law 9074, 573 Law 9427, 574, 575 Law 9478, 575–576 Law 9648, 575 Law 10433, 580 Law 10847, 586 Law 24065, 599 Learning, 22–23, 77, 92, 105, 614 Legal unbundling, 139, 277, 624, 627 Legislative actions, 387–391 Legislative foresight, 390 Liberalization, 1, 8, 22, 23, 29, 41, 47, 48, 81, 109, 139, 140, 239, 242, 244, 251, 261, 266,

Index 267, 274, 287, 290, 308, 309, 599, 605, 617, 620, 627, 638–639, 640, 641 Licences, 112, 118, 140, 587, 589 Light-handed regulations, 203, 219, 221, 231, 484 Lignite, 237–238, 260 Liquefied natural gas (LNG), 62, 68, 104, 232, 259, 377, 620 Litigation, 371 Load aggregation, 410 Load pockets, 6, 19, 402, 510 Load-serving entities (LSEs), 394, 468, 469, 470, 506, 517 Local distribution companies, 23, 26, 63, 331, 339, 420, 424, 431, 433 Locational marginal pricing (LMP), 120, 413, 452, 460, 463, 464, 465, 497 Long-term contract, 37–38, 46, 88, 103, 105n, 118, 139, 218, 285, 299, 319, 323, 325, 331n, 338, 348, 365, 375, 505, 588, 589 Long-term energy procurement (LEP), 195 Loss pricing methods, 517–518 MacDonald Report, 423 Madrid Forum, 305 MAE, 575, 576, 578, 579, 580, 586 Management, 181, 187, 221, 302, 330, 332, 333, 349, 365, 380, 384, 406, 463, 474, 513, 521, 634–637 Mandatory insurance schemes, 195 Mandatory renewable energy target (MRET), 196, 200 Manipulation, 37, 189, 402, 485, 494, 606 Marginal units by fuel type, 456 Market Achievement Plan (MAP) auction, 444 Market characteristics, 125, 140, 285–286 Market-clearing price (MCP), 335, 337, 342, 343, 345, 346–347, 360, 369 Market concentration, 153, 286–288, 301, 401–403, 458, 508, 510, 597–598 Market design, 8, 9, 23–27, 109, 113–117, 137, 169, 301–303, 367, 375, 398–399, 406, 451, 454, 456–458, 461, 475, 480, 487, 487–489, 495, 498, 500, 508, 541, 552 Market design choice, 398–399 Market Design Committee (MDC), 423 Market design debates, 513–515 Market design decisions, 521–522 Market design flaw, 319, 401, 406, 489 Market distortions, 199–200 Market equilibrium, 123–126 Market flaws, 371–372 Market force, 195, 266, 328, 334, 350, 371, 466, 546

651

Market governance, 180–181 Market integration, 155, 187, 301–302 Market issues, 341–348 Market monitoring, 16, 384, 398, 412, 542 Market opening, 14, 56, 57, 130–131, 156, 240, 247, 252, 274, 277, 280, 397–398, 424–426, 430, 458, 459–461, 467, 470–472, 479 Market performance, 47, 130, 285, 459, 460, 462, 467, 470–472, 473, 479 Market power, 16, 65, 159, 342, 363–364, 401–403 Market power and gaming, 362–364 Market power mitigation, 16, 461–462, 472, 499, 503, 507–512, 522 Market Power Mitigation Agreement (MPMA), 424, 430 Market power monitoring, 507–512 Market power sources, 507–508 Market profile, 174 Market re-regulating proposals, 377–378 Market risk, 345–348 Market rules, 180, 637, 432, 452, 467, 472–473, 482, 486, 488 Market structure, 122, 128–130, 154, 158–159, 285, 288, 332, 361, 362, 363, 371, 393, 405–406, 413, 458–459, 467, 470, 626 Market structure demand, 431–432 Market structure supply, 430–431 Market versus regulation, xviii Massachusetts, 538, 541 Memorandum of Understanding (MOU), 331 Merchant plant, 389 Merchant transmission, 19, 184, 434n Mergers, 18, 169, 181n, 286–289 Mergers and acquisitions, 239, 252, 287 Million solar Roofs, 376 Monopolistic bottlenecks, 50, 53–54, 55–56, 56–69 Monopoly, 36, 51–52, 54, 59, 60, 116, 140, 173, 180, 181, 189, 211, 267, 320, 329, 361, 423, 479–480, 483 Multi-settlement system, 500 Municipal utilities, 239, 320, 341,366, 380 National blocks, 302, 305 National champion, 251, 282–283, 284, 287–288 National Electricity Market (NEM), 44, 173, 176, 201 National Grid, 111, 112 National Grid Company, 112, 116, 119, 120, 132, 137 National market operations, 169, 180, 186–190, 508 National Power, 112, 118, 128, 129, 138, 199

652

Index

National System of Electric Transmission (SINTREL), 573 Natural gas, 62, 78, 80, 81, 92, 97, 102, 103, 104, 106, 265, 345, 354, 357–359, 360, 370, 371, 377, 387, 408, 410, 445, 589, 631 Natural Gas Corporation, 211, 218 Natural gas deficit, 103, 104 Natural gas pipelines, 63, 218, 345, 483 Natural gas transmission, 62–63 Natural monopoly, 50, 51n, 52, 53, 54, 59, 60, 122, 140, 339, 392 NECA, 180 Negotiated rate model, 534, 535, 541 NEMMCO, 180, 187 Net revenue, 459, 462, 468, 469, 475 Network access, 50, 56, 66, 93, 235, 242, 246, 261 Network competition, 50, 59, 63–64 Network industries, 50, 70 Network investment, 252–253, 261 Network price setting, 180, 185–186 Network regulation, xxvi–xxvii, 235, 236, 259 Network sectors, 49, 50, 54, 55, 57, 59–61, 67 Network-specific market power, 49, 50, 53–54, 55, 56, 57–59, 61, 66, 68, 69 New capacity and prices, 200–202 New Electricity Trading Arrangements (NETA), 44, 109, 119, 126–138, 141 New Jersey, 27 New market entrants, 407–408, 631 New South Wales (NSW), 173, 176, 178, 184, 188, 192, 193, 194–195, 196, 199 New Zealand, 18, 27, 45, 203, 207, 208, 210, 211, 212–213, 214, 215, 217–218, 221–222, 223, 229, 230, 231, 232 New Zealand Electricity Division (NZED) (see also Electricity Corporation of New Zealand (ECNZ)), 207, 208, 209, 210, 211 Nodal market design, 398–399 Nodal market structure, 413–414 Nodal pricing, 121, 215–216 Non-discriminatory access, 275–277 Non-fossil Fuel Obligation (NFFO), 116 Non-price regulation, 542, 561 Nord Pool, 146, 147, 148, 154, 155, 157 Nordic electric market, 145 Nordic market, 20, 22, 44, 145, 146, 147, 166–167, 169 Normative values (NV), 577 Norsk Hydro, 159 North America, 38, 46, 383, 451 Norway, 21, 27, 111, 121, 153, 155, 157, 159, 164, 165, 166, 168, 578 Nuclear costs, 197–198, 198 Nuclear energy, 37

Nuclear fuel cycle, 641 Nuclear plants, 22, 238 Nuclear power, 112, 117, 160–161, 198, 303, 336, 600, 639–641 Nuclear stations, 112 NZED (see also ECNZ), 207, 208, 209, 210, 211 Oil company, 590, 632 Oil shocks, 620 Old supplier, 157, 158 One-price auction system, 335 ONS, 575, 581 Ontario, 146, 420–423, 424, 430, 431, 433, 434, 435, 440 Ontario Electricity Financial Corporation (OEFC), 434 Ontario electricity sector, 436 Ontario Energy Board (OEB), 424, 435, 437 Ontario experience, 146, 440–441, 427 Ontario Hydro, 423, 429, 438 Ontario market, 424, 430, 432, 438n Ontario Power Authority (OPA), 436, 441 Ontario Power Generation (OPG), 422, 423–424, 430–431, 439 Optimization–simulation model, 605 Optimized depreciated replacement cost (ORDC), 222 Optimized deprival value (ODV), 18, 225, 226, 227 Order 888, 11n, 242n, 325, 472, 483–485 Organizational models, 531, 532, 543–546 Organized markets, 22, 286, 338, 394, 487 Out-of-merit-order, 395 Over-investment, 36, 164 Over-regulatory avoidance, 68 Pacific Northwest drought, 323, 355–357, 364 Para-NETA, 131 Pay-as-bid pricing, 494, 496 Pennsylvania, 24–25, 541 Pennsylvania–New Jersey–Maryland (PJM), 2, 46, 339, 438, 452, 541, 553, 605, 606 Perfect storm, 364, 379, 380 Performance after privatization, 121–123 Performance assessments, 6–7 Performance evaluation, 1, 511–512, 519-522 PG&E, 320, 373 Physical withholding, 508 PJM capacity markets operation, 470 PJM history, 452–454 PJM market, 451, 454 PJM’s market power mitigation, 461–462 Plant capacity, 114 Playing field, 42

Index PLP model, 89n Poland, 45, 267, 268, 275, 280, 281, 285 Policy goals, 397 Policy-makers, 44, 48, 209, 257 Polluter-pays principle, 641 Pool purchase price (PPP), 114, 120 Pool selling price (PSP), 114 Portugal, 45, 267, 268, 275, 281, 284 Potential competition, 50, 51, 52, 55, 56–57, 58, 59 Power crisis 2001, 47, 565, 581–585 Power exchange (PX), 15, 124, 149, 331, 335–336, 617, 626–627, 637, 638 Power generating organizational issues, 377 Power generation, 61, 77, 159, 197, 237, 626–627 Power Generation Companies (PGCs), 386, 387, 389–390, 392, 393, 394, 403 Power generation costs, 197 Power generators, 63, 401 Power producers and suppliers (PPSs), 623, 632 Power Purchase Arrangements (PPAs), 444 Power transmission networks, 63 PowerGen, 112, 118, 128, 138, 552 Pre-2004 restructuring, 423–427 Price cap, 46, 66, 67, 122, 253, 319, 344, 502 Price-cap regulation, 67, 250, 253 Price components, 531, 560 Price convergence, 272, 274, 285, 494 Price–cost margin, 9, 44, 109, 126, 228, 511 Price–cost markup index, 459 Price–cost relationship, 540–541 Price discrimination, 159–160, 363–364 Price disparities, 36–37 Price divergence, 215, 246 Price equilibration, 344 Price freeze results, 433–434 Price frequency, 531, 533, 560 Price impacts, 546, 551 Price-level regulation, 65–66 Price outcomes, 190, 193 Price setting mechanism, 531, 533, 560 Price subsidies, 37 Price-to-beat (PTB), 384, 395, 396, 410, 411, 414, 451 Price trends, 410–411 Price volatility, 403, 438n, 613 Pricing, 88–96, 219–230, 435–439 Pricing attributes, 554 Private investment discouragement, 434 Private ownership, 194–195 Private sector businesses, 194 Privately owned, regulated monopolies, 35, 36 Privatization, 3, 9, 41, 86, 111–123, 431

653

Product customization and flexibility, 554–557 Product innovation, 554–557 Production, 61, 177, 237, 238, 254 Production cost, 496 Productive efficiency, 487 Profit-sharing rule, 250 Profitability, 219–230 Proposition 9, 334–335 Proposition 80, 378, 379 Providers of last resort (POLRs), 384, 393, 395, 396 Public contest method, xxviii–xxix Public debt, 40 Public disclosure, 38 Public ownership, 194–195 Public perceptions, 378–379 Public service, 500, 590, 599 Public Utility Commission of Texas (PUCT), 385, 390, 396, 397, 402n, 410, 412, 413, 414, 454, 552 Public Utility Regulatory Policies Act (PURPA), 324, 325, 326, 388, 480, 483, 552, 623 PX-bidding strategies, 343–345, 366 PX MCP, 342, 344, 350, 362 Qualifying facilities (QFs), 324, 326, 360, 366 Quantity requirements, 505–506 Queensland, 176, 178, 182, 183, 187, 192, 193, 195, 196, 200 Queensland Electricity Commission (QEC), 178 Quicker response reserve, 495 Rate base, 36, 250, 557 Rate-of-return regulation, 36, 249, 250, 253, 469 Ratepayers, 36, 327–328, 366 Rational buyer model, 496 Rationing, 162, 582 Rationing cost, 89n Reactive power, 499–500 Real-Time Energy Market, 456–457 Referee, 42, 305 Reference price, 510, 511 Reforms of the reforms, 2, 28–29, 370, 378, 586–587 Reform process, 14, 16–18, 177–178, 226, 284 Reformed electricity market, 191–194, 423 Regional Clean Air Incentives Market (RECLAIM), 360–361 Regional Electricity Companies, 112, 113, 116, 117, 118, 119, 128, 129, 138, 139 Regulated monopoly model, 29, 36 Regulated nodal prices, 90, 92–93, 97, 99 Regulated rate option, 446, 447

654

Index

Regulation, 10, 38n, 41–43, 66, 117–120, 163, 204, 252–253, 261, 390, 435–439, 472, 473 Regulation and operating reserves, 494–495, 495–496 Regulation market, 49, 454, 473, 554 Regulator, 5, 15, 26, 37, 39, 40, 42, 46, 67, 126, 139n, 140, 186, 200, 248, 249, 251, 253, 259, 319, 326, 511, 607, 614 Regulatory agency, 5, 69, 585 Regulatory authorities, 66, 248, 250, 277, 279, 301, 468 Regulatory compact, 327 Regulatory complexity, 40 Regulatory environment, 5, 217, 277–278, 388, 436, 443 Regulatory frameworks, 17, 19, 92, 177, 209, 329, 612 Regulatory governance, 303–306 Regulatory institution, 18 Regulatory instruments, 65, 67 Regulatory paradigm, 36, 39–41 Regulatory reform, 2, 7, 8–14, 17, 47, 145, 160, 161, 479, 480, 482 Reliability unit commitment, 494 Renewable energy, 22–23, 196, 376 Renewable Energy Certificates, 116, 118, 128, 138, 196, 197, 198 Renewable Energy Credit, 394, 395 Renewable resource owner, 395 REPs, 393, 395, 396, 397, 407–408 Reserve margin, 132, 136, 258, 414, 468, 505n, 600, 610 Reserve trader, 187 Reserves pricing, 496–498 Residential market, xxii–xxv, 559 Residual supply index (RSI), 458–459 Resource adequacy, 13, 20–22, 46, 319, 384, 412, 414, 523 Resources, 320, 336, 337, 372, 395, 403, 412, 447, 454, 472 Restructured system, 335–339, 345 Restructuring, 1, 2, 4–6, 8, 14, 19, 41, 46, 111, 121, 122, 203, 208–219, 226, 265, 274, 283, 345, 370, 379, 380, 397, 419, 421, 423, 433, 441, 547, 551, 552 Restructuring and legislative actions, 387–391 Restructuring implementation, 391–397 Restructuring process, 19, 23, 226, 292, 328, 431 Restructuring utilities, 300–301 Retail assessment, 411 Retail competition, xxvi, xxii–xxiv, 3, 8, 9, 11, 23, 26, 27, 27–28, 138, 155, 156–157, 182, 183, 219, 241, 328, 341, 372–373, 390–391, 392, 395–397, 406–411, 613, 617, 639

Retail competition program, 9–10, 23, 27 Retail electric commodity products, 555 Retail electric providers (REPs), 391, 393, 394, 395, 396, 397, 400, 403, 406, 407, 407–408, 410, 412, 546 Retail electricity evolution, 290–291 Retail electricity prices, 157, 629–633 Retail liberalization, 140 Retail market design, 23, 542 Retail market price, 155, 538, 539 Retail market risk, 348 Retail markets, 24, 29, 150, 158–159, 169, 218, 328, 339–341, 425, 468, 534, 612, 623 Retail price, 10, 26, 27, 153, 154, 155, 157, 200, 277, 325, 326, 333, 334, 340, 361, 366, 371, 372, 380, 395, 408, 531, 552 Retail price control, 182, 193, 195–196, 361, 372, 380 Retail price freeze, 426–427, 433, 434 Retail regulation, 182–183 Retail restructuring, 216–219, 341 Retail supply market, xxii Retailer business model, 558–559 Retailing, 23, 139–140, 159, 159–160, 181 Revenue shortfalls, 516–517 Revenue sufficiency guarantee, 498 RFP model, see Auction model Risk management, 181, 187, 330–331, 333, 339, 380 RWE, 238, 246, 260 Scarcity pricing, 257, 414, 498–499, 503, 506 Scotland, 44, 109, 110–111, 111–112, 120, 121, 140 Scottish system, 112 SDG&E, 320, 348, 366, 373 Second Generation Reforms, 613–614 Sector-specific market power regulation, 49, 51–56, 57, 58, 67, 68 Sector structuring, 236 Self-schedule, 15, 466, 497n Senate Bill 7, 391–392, 402 Senate Bill 20, 395 Senate Bill 373, 388, 389–390 Senate Bill 408, 412 Sequential market clearing, 496 Short-run elasticity, 345, 359 Short-term AS, 494–495, 495 Short-term security plan, 101 SIC, 78, 81, 82, 83, 85, 90, 91, 92, 93–94, 96, 97, 103–104 SIC electric energy crisis, 97 Simultaneous feasibility test, 516 Simultaneous market clearing, 496

Index SING, 78, 79–80, 82, 85, 91, 92, 93–94, 96, 99, 100, 101, 103–104, 105 SING’s crisis, 86, 100, 101 Slovak Republic, 282 Smart meter, 186–187, 257, 307–308, 437 Social programs, 397 Sovereign risk, 198–199 Spain, 284, 301, 303, 505 Spinning energy premium, 473 Spinning reserve, 128, 472, 473, 474 Spot energy pricing, 494 Spot market, 9, 10, 22, 116, 117, 136, 137, 146, 186–187, 214, 224, 282, 331, 345, 346, 347, 348, 361, 365, 367, 371–372, 439, 457, 458, 489–500, 510, 566, 577, 600, 626, 637 Spot market design, 487–489 Spot market model, 537–538 Spot price, 148, 150, 153, 157n, 187, 205, 215, 222–224, 348, 349, 350, 497, 510, 576–577, 600, 601, 602, 612 Spot-price contracts, 153–154 Spot wholesale energy market, 15, 27–28 Stadtwerke, 239 Standard market design, 137, 141, 487 Standard offer model, see Negotiated rate model State legislation, 332, 377 State-owned enterprises (SOEs), 35–36, 37, 40 State owned integrated utilities, 177–186 State regulation change, 367–369 State regulators, 180, 185, 542 Statkraft, 152, 159 Statoil, 159, 164 Status quo change, 35–39 Status Quo Service, 529, 532, 533, 539, 541 Stranded cost, 254, 327–328, 330, 333, 334, 341, 392, 484, 531, 640 Structural developments, 181–182 Sunk cost, 56, 326, 327, 432 Supplier switching rate, 629–633 Supply adequacy, 37, 38, 160–162, 163–165, 169, 257, 261, 266 Supply/demand balance, 207–208, 370 Supply function equilibrium, 123–124 Supply obligation, 13, 14, 531, 540, 542, 543, 544 Supply profile, 174–176 Supply security, 20, 145, 160–163, 256, 257 Supply shock, 150–155, 168 Supra-regional transmission gas pipelines, 50, 63–64 Sweden, 155, 156, 157, 158, 159, 160, 166, 167 Switches, 407 Switching activity, 408–409

655

Switching costs, 158 Switzerland, 267, 269, 277, 285, 299 Sydkraft, 159 Sympathetic regulators, 39 System buy price (SBP), 127, 135, 136 System operator, 85–87, 121, 127, 162, 169, 214, 251, 505, 520, 600 System price, 146, 155n System sell price (SSP), 127 Tacit co-ordination, 125–126, 140 Take-or-pay contracts, 116, 589 Takeovers, 286–288 Tariff revenue, 94 Taxation by regulation, 4, 22, 29 Taxpayers, 35, 434, 438 Telecommunication, 60, 68, 69, 253n Texas, 9–10, 28, 46, 383, 384, 385, 386, 388, 390, 391, 392, 393, 402, 403, 408, 412, 413, 538, 539, 540, 546, 547–551, 552–553, 561 Texas Price to Beat (PTB), 395, 408, 410, 538, 540, 541, 551 Texas Public Regulatory Act (PURA), 388 Textbook model, xviii–xix, 8–14, 16, 480 TGCs, 166 Thermal generation, 79–80, 81, 92, 203, 207, 583, 590 Thermal power plants, 91, 103, 104, 576, 578, 618 Third Party Access (TPA), 241–242, 261, 285, 300 Tier 1 spinning payments, 472–473 Tractebel, 268, 284, 578 Tradable Green Certificates (TGC), 166–167, 167–168 Trading arrangements, 129, 130–131 Traditional rate of return regulation, 36, 542, 557 Transaction cost, 27, 291, 300, 302 Transco, 4, 578 Transfer prices, 88–89 Transfer rule, 236, 256 Transition time, 413 Transmission and Distribution Service Providers (TDSPs), 392, 393, 396, 397, 407, 412 Transmission business, 178, 282 Transmission capacity, 152–153, 169, 303, 315, 352, 376–377, 484, 514 Transmission charges, 94, 120, 226 Transmission congestion, 19, 394, 508 Transmission congestion rights (TCRs), 394–395 Transmission control, 183–185 Transmission grid, 205, 211, 224, 255, 439, 452, 513

656 Transmission investment, 18–20, 164 Transmission loading relief procedures (TLRs), 462–463 Transmission networks, xxvii–xxviii, 62–64, 90, 627, 638 Transmission pricing, 289 Transmission restructuring, 211–215 Transmission system, 19, 96, 225, 302, 393, 395, 516, 585 Transmission System Operator (TSO), 114, 145–146, 254, 267, 300, 484 Transportation, 62, 64, 102, 103, 259, 576 Unbundling, 122, 140, 251, 277, 392, 581, 624, 625, 627, 629 Unhedgeable congestion, 466 Uniform market clearing prices, 494 United Kingdom, 17, 19, 328, 330, 551–552, 605, 606, 638 Unregulated price model, 539 Unresponsive demand, 503 Unrestrained consumption, 433–434 US antitrust law, 50, 54 US regulatory reforms, 480, 482–487 USA, 10, 12, 16, 20, 21, 22, 24, 47, 55, 111, 257, 324, 385, 388, 402, 408, 412, 479, 480, 498, 500, 506, 507, 508, 529, 638 Utility manager, 344 Utility separation, 531 Value chain, 44, 53, 61, 77 Variable-price contracts, 153–154, 157 Vattenfall, 159, 247, 287 Vertical integration, 27, 140, 141, 219, 267–268 Vertical reintegration, 283, 287 Vertically integrated monopolies, 7–8, 36, 323

Index Vertically integrated utility, 14, 24, 38, 112, 392, 407, 414, 443 Victoria, 176, 177, 182, 183, 185, 189, 192 Water supply, 606 Western electricity crisis, 348 Wheeling price, 623 White Paper, 301, 423 Wholesale assessment, 406 Wholesale competition, 38, 388–390 Wholesale contracts, 338–339 Wholesale design, 532 Wholesale electricity price evolution, 288–289 Wholesale electricity spot market, 214, 366, 367 Wholesale generation market, xx–xxi Wholesale market, 8, 11, 14–15, 18–19, 29, 38, 122, 190, 252, 326–327, 334, 335–339, 393, 397–406, 444, 534, 600 Wholesale Market Oversight (WMO), 398, 402, 412 Wholesale market price, 13, 26 Wholesale market risk, 345–348 Wholesale power market, 637–638, 641 Wholesale price, 38, 155–156, 164, 168, 218, 244, 252, 330, 345, 348, 350–353, 368, 445 Wholesale price mitigation, 375 Wholesale price volatility, 345 Wholesale spot price, 222–224 Wholesale trading, 19, 239 Wind power, 138, 185, 198 Yardstick assessment, 623 Yellow Book, 330 Zonal market design, 398–399 Zonal pricing, 496, 497, 514 Zonal system problems, 399–401

Electricity Market Reform An International Perspective

Power Generation Investment in Electricity Markets (Energy Market Reform)

Multiracial Identity: An International Perspective

Public Finance An International Perspective

Information Modelling: An International Perspective

Market or Government Failures?: An Asian Perspective

International Comparisons of Electricity Regulation

The Amazon from an International Law Perspective

Learning and Teaching Mathematics: An International Perspective

The Origins of Agriculture: An International Perspective

Expertise in Mathematics Instruction: An International Perspective

Russian Electricity Reform: Emerging Challenges And Opportunities

Healthy and Productive Work: An International Perspective

Violence Against Women: An International Perspective

Textbook of Pediatric Asthma: An International Perspective

Computers in applied linguistics: an international perspective

Becoming a Mathematician : An international perspective

Psychology of Grandparenthood: An International Perspective

International comparisons of electricity regulation

Electricity Market Reforms: Economics and Policy Challenges

ELECTRICITY

Electricity

Market Leader: International Management

The Reform of International Institutions

Globalization and Summit Reform: An Experiment in International Governance

Biopharmaceuticals, an Industrial Perspective

Biopharmaceuticals, an Industrial Perspective

Electricity Market Reform An International Perspective

Power Generation Investment in Electricity Markets (Energy Market Reform)

Multiracial Identity: An International Perspective

Public Finance An International Perspective

Information Modelling: An International Perspective

Market or Government Failures?: An Asian Perspective

International Comparisons of Electricity Regulation

The Amazon from an International Law Perspective

Learning and Teaching Mathematics: An International Perspective

The Origins of Agriculture: An International Perspective

Expertise in Mathematics Instruction: An International Perspective

Russian Electricity Reform: Emerging Challenges And Opportunities

Healthy and Productive Work: An International Perspective

Violence Against Women: An International Perspective

Textbook of Pediatric Asthma: An International Perspective

Computers in applied linguistics: an international perspective

Becoming a Mathematician : An international perspective

Psychology of Grandparenthood: An International Perspective

International comparisons of electricity regulation

Electricity Market Reforms: Economics and Policy Challenges

ELECTRICITY

Electricity

Market Leader: International Management

The Reform of International Institutions

Globalization and Summit Reform: An Experiment in International Governance

Biopharmaceuticals, an Industrial Perspective

Biopharmaceuticals, an Industrial Perspective

Recommend Documents