Developments and Innovation in Carbon Dioxide (CO2) Capture and Storage Technology: Volume 1: Carbon Dioxide (CO2) Capture, Transport and Industrial Applications (Woodhead Publishing Series in Energy)

i Developments and innovation in carbon dioxide (CO2) capture and storage technology © Woodhead Publishing Limited, 2...

Author: M.Mercedes Maroto-Valer

95 downloads 2087 Views 16MB 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

i

Developments and innovation in carbon dioxide (CO2) capture and storage technology

© Woodhead Publishing Limited, 2010

ii

Related titles: Advanced power plant materials, design and technology (ISBN 978-1-84569-515-6) Fossil-fuel power plants generate the majority of the world’s power, but many plants are ageing and cannot meet rising global energy demands and increasingly stringent emissions criteria. To ensure security and economy of supply, utilities are building a new generation of advanced power plant with increased output and environmental performance. This book initially reviews improved plant designs for efficiency and fuel flexibility, including combinedcycle technology and utilisation of lower-grade feedstocks. Coverage extends to advanced material and component use, and the incorporation of alternative energy conversion technology, such as hydrogen production. Environmental and emissions performance issues round off the book. Oxy-fuel combustion for power generation and carbon dioxide (CO2) capture (ISBN: 978-1-84569-671-9) Oxy-fuel combustion is a power generation and carbon dioxide (CO2) capture option for advanced power plant in which fuel is burnt in an oxygen-rich environment instead of in air. This allows for a reduction in NOx and SOx emissions as well as producing a high-purity carbon dioxide (CO2) flue gas stream. This high-purity CO2 stream allows for more efficient and economical capture, processing and sequestration. This book critically reviews the fundamental principles, processes and technology of oxy-fuel combustion, including advanced concepts for its implementation. Details of these and other Woodhead Publishing books can be obtained by:  visiting our web site at www.woodheadpublishing.com  contacting Customer Services (e-mail: [email protected]; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext. 130; address: Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK) If you would like to receive information on forthcoming titles, please send your address details to: Francis Dodds (address, tel. and fax as above; e-mail: [email protected]). Please confirm which subject areas you are interested in.


iii

Woodhead Publishing Series in Energy: Number 8

Developments and innovation in carbon dioxide (CO2) capture and storage technology Volume 1: Carbon dioxide (CO2) capture, transport and industrial applications Edited by M. Mercedes Maroto-Valer

CRC Press Boca Raton Boston New York Washington, DC

Woodhead

publishing limited

Oxford Cambridge New Delhi


iv Published by Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK www.woodheadpublishing.com Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road, Daryaganj, New Delhi – 110002, India www.woodheadpublishingindia.com Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2010, Woodhead Publishing Limited and CRC Press LLC © Woodhead Publishing Limited, 2010 The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the author and the publishers cannot assume responsibility for the validity of all materials. Neither the author nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing ISBN 978-1-84569-533-0 (book) Woodhead Publishing ISBN 978-1-84569-957-4 (e-book) CRC Press ISBN 978-1-4398-3099-4 CRC Press order number: N10185 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elemental chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Cover image © BCS Creative, 88–90 North Sherwood Street, Nottingham NG1 4EE, UK, www.bcscreative.co.uk Typeset by Replika Press Pvt Ltd, India Printed by TJ International Limited, Padstow, Cornwall, UK


v

Contents

Contributor contact details

xiii

Woodhead Publishing Series in Energy

xvii

Foreword by Lord Oxburgh

xix

1

Overview of carbon dioxide (CO2) capture and storage technology

S. Bouzalakos and M. Mercedes Maroto-Valer, University of Nottingham, UK

1.1 1.2 1.3 1.4

Introduction Greenhouse gas emissions and global climate change Carbon management and stabilisation routes Development and innovation in carbon dioxide (CO2) capture and transport technology Development and innovation in carbon dioxide (CO2) storage and utilisation technology Future trends Sources of further information and advice Acknowledgements References

1.5 1.6 1.7 1.8 1.9

1

1 2 8 11 17 19 20 22 22

Part I Carbon dioxide (CO2) capture and storage economics, regulation and planning 2

Techno-economic analysis and modeling of carbon dioxide (CO2) capture and storage (CCS) technologies

J. Ogden and N. Johnson, University of California Davis, USA

2.1 2.2

Introduction Carbon dioxide (CO2) capture


27 27 31

vi

Contents

2.3 2.4 2.5

Carbon dioxide (CO2) transport Carbon dioxide (CO2) injection Carbon dioxide (CO2) capture and storage (CCS) system modeling Future trends References

56 59 61

3

Regulatory and social analysis for the legitimation and market formation of carbon dioxide (CO2) capture and storage technologies

64

H. de Coninck, M. de Best-Waldhober and H. Groenenberg, Energy research Centre of the Netherlands (ECN), the Netherlands

3.1 3.2

Introduction Technological maturity and the carbon dioxide (CO2) capture and storage (CCS) innovation system Legitimation: results and gaps in social scientific research regarding public perception and participation Market formation and direction of search: an enabling regulatory framework for carbon dioxide (CO2) capture and storage (CCS) in the EU Implementation outlook for carbon dioxide (CO2) capture and storage (CCS) technologies Sources of further information and advice References

86 88 88

Energy supply planning for the introduction of carbon dioxide (CO2) capture technologies

93

2.6 2.7

3.3 3.4 3.5 3.6 3.7 4

4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8

36 44

64 67 74 80

A. Elkamel, H. Mirzaesmaeeli, E. Croiset and P. L. Douglas, University of Waterloo, Canada

The emerging energy challenge and a case from Ontario, Canada Overview of supply technologies and carbon capture and storage Future trends Energy conservation strategy Planning model Illustrative case study Conclusions References


93 98 105 113 115 124 149 151

Contents

vii

Part II Post- and pre-combustion processes and technology for carbon dioxide (CO2) capture in power plants 5

155

Advanced absorption processes and technology for carbon dioxide (CO2) capture in power plants

U. Desideri, Università degli Studi di Perugia, Italy

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8

Introduction Absorption processes Description of the technology Advancements in the technologies Advantages and disadvantages Applications and future trends Conclusions References

155 156 161 166 170 172 172 180

6

Advanced adsorption processes and technology for carbon dioxide (CO2) capture in power plants

183

6.1 6.2 6.3 6.4 6.5 6.6 6.7

Introduction Mesoporous and microporous adsorbents Functionalised sorbents Regenerable sorbents Sources of further information and advice Conclusions References

183 184 186 192 197 197 198

7

Advanced membrane separation processes and technology for carbon dioxide (CO2) capture in power plants

203

A. Basile and A. Iulianelli, Italian National Research Council, Italy, F. Gallucci, University of Twente, the Netherlands, P. Morrone, University of Calabria, Italy

7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11

Introduction Cryogenic carbon dioxide (CO2) capture Performance of membrane systems Carbon dioxide (CO2) membrane materials and design Membrane modules Comparing membrane modules Design for power plant integration Cost considerations Future trends and conclusions Sources of further information and advice References

R. M. Davidson, IEA Clean Coal Centre, UK


203 208 213 216 221 223 225 232 234 236 238

viii

Contents

8

243

Gasification processes and synthesis gas treatment technologies for carbon dioxide (CO2) capture

C. Higman, Higman Consulting GmbH, Germany

8.1 8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9

Introduction Basic principles Applications Building blocks for complete systems Power plant as an example for a complete system Advantages and limitations Future trends Sources of further information and advice References

243 244 258 261 270 273 276 277 278

Part III Advanced combustion processes and technology for carbon dioxide (CO2) capture in power plants 9

283

Oxyfuel combustion systems and technology for carbon dioxide (CO2) capture in power plants

P. Mathieu, University of Liège, Belgium

9.1 9.2 9.3 9.4 9.5 9.6

Introduction Basic principles of oxyfuel combustion Technologies and potential applications Advantages and limitations Future trends References

283 285 287 307 313 315

10

Advanced oxygen production systems for power plants with integrated carbon dioxide (CO2) capture

320

10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9

S. C. Kluiters, R. W. van den Brink and W. G. Haije, Energy research Centre of the Netherlands, the Netherlands

Introduction Technologies for air separation Oxygen selective membrane technology for oxyfuel power plants Power generation systems integrated with oxygen selective membrane (OSM) units Advantages and limitations Future trends Sources of further information and advice Conclusions References


320 322 326 331 347 352 352 353 354

Contents

11

Chemical-looping combustion systems and technology for carbon dioxide (CO2) capture in power plants

E. J. Anthony, CANMET Energy Technology Centre-Ottawa, Canada

11.1 11.2 11.3 11.4

Introduction Basic principles Technologies and potential applications Advantages and limitations of chemical-looping combustion (CLC) for natural gas and syngas 11.5 Hydrogen manufacture using chemical-looping combustion (CLC) 11.6 The use of chemical-looping combustion (CLC) technology with solid fuels 11.7 The CaS–CaSO4 system 11.8 Future trends 11.9 Sources of further information and advice 11.10 References

ix

358

358 359 362 364 366 368 371 373 374 374

Part IV Carbon dioxide (CO2) compression, transport and injection processes and technology 12

Gas purification, compression and liquefaction processes and technology for carbon dioxide (CO2) transport

A. Aspelund, The Norwegian University of Science and Technology, Norway

12.1 12.2 12.3

Introduction Selection of transport pressures Carbon dioxide (CO2) quality recommendations for transport in pipelines and by ship 12.4 Overview and basic building blocks in carbon dioxide (CO2) transport processes 12.5 Sensitivity analysis 12.6 The interface between capture and transport 12.7 Ship to pipeline and pipeline to ship processes 12.8 Discussion 12.9 Future trends and future work 12.10 Conclusions 12.11 Acknowledgements 12.12 References


383

383 385 386 387 395 400 402 403 404 405 405 405

x

Contents

13

408

Infrastructure and pipeline technology for carbon dioxide (CO2) transport

P. N. Seevam, J. M. Race, and M. J. Downie, Newcastle University, UK

13.1 13.2 13.3 13.4 13.5 13.6 13.7 13.8 13.9 13.10 13.11 13.12

Introduction Carbon dioxide (CO2) phase properties Transport of carbon dioxide (CO2) by pipeline Transport of carbon dioxide (CO2) by ship Transport economics Large-scale transport infrastructure Discussion Future trends and future work Conclusions Sources of further information and advice Acknowledgements References

408 409 414 423 425 425 428 429 429 430 430 431

14

Carbon dioxide (CO2) injection processes and technology

435

S. Solomon and T. Flach, DNV – Research and Innovation, Norway

14.1 14.2 14.3

Introduction Underground fluid injection Analogues for carbon dioxide (CO2) storage and best practices from other sectors Injection well technologies Controlling parameters for carbon dioxide (CO2) injectivity Carbon dioxide (CO2) injection in different storage formations Carbon dioxide (CO2) injection field operations Injection of carbon dioxide (CO2) and well integrity Technologies for monitoring injection well integrity Future trends Sources of further information and advice Acknowledgements References

14.4 14.5 14.6 14.7 14.8 14.9 14.10 14.11 14.12 14.13

435 436 437 438 441 449 451 453 459 462 462 463 463

Part V Industrial applications of carbon dioxide (CO2) capture and storage technology 15

Carbon dioxide (CO2) capture and storage technology in the cement and concrete industry

S. Ghoshal, McGill University, Canada, F. Zeman, New York Institute of Technology, USA

15.1

Introduction

469

469


Contents

xi

15.2 15.3 15.4 15.5 15.6 15.7 15.8

Basic principles Capture of carbon dioxide (CO2) from cement plants Accelerated carbon dioxide (CO2) curing of concrete Future trends Conclusions Sources of further information and advice References

470 472 479 486 487 488 489

16

Carbon dioxide (CO2) capture and storage technology in the iron and steel industry

492

J-P. Birat, Arcelor Mittal, France

16.1 16.2 16.3

Introduction Carbon dioxide (CO2) emissions of the steel sector Strategies to control carbon dioxide (CO2) emissions from the steel sector 16.4 Carbon capture and storage (CCS) for the steel sector 16.5 Carbon dioxide (CO2) capture technologies for the steel sector 16.6 Carbon dioxide (CO2) storage for the steel sector 16.7 Perspectives on carbon capture and storage (CCS) and carbon dioxide (CO2) abatement in the steel sector 16.8 Conclusions 16.9 Acknowledgements 16.10 Sources of further information and advice 16.11 References

504 509

523

Index


492 493 497 503

515 517 518 518 518

xii

xiii


(* = main contact)

Chapter 1 Dr Steve Bouzalakos and Professor M. Mercedes Maroto-Valer* Centre for Innovation in Carbon Capture and Storage (CICCS) Faculty of Engineering The University of Nottingham University Park Nottingham NG7 2RD UK Email: mercedes.maroto-valer@ nottingham.ac.uk

Chapter 2 Dr Joan Ogden* Professor, Environmental Science and Policy Department Director, Sustainable Transportation Energy Pathways Program Institute of Transportation Studies University of California, Davis One Shields Avenue Davis CA 95616 USA

Mr Nils Johnson Graduate Research Assistant Institute of Transportation Studies University of California, Davis One Shields Avenue Davis CA 95616 USA Email: [email protected]

Chapter 3 Heleen de Coninck*, Marjolein de Best-Waldhober and Heleen Groenenberg Energy research Centre of the Netherlands (ECN) Unit Policy Studies Radarweg 60 1043 NT Amsterdam the Netherlands Email: [email protected]

Email: [email protected]


xiv


Chapter 4

Chapter 7

A. Elkamel, H. Mirzaesmaeeli, E. Croiset and P.L. Douglas* Department of Chemical Engineering University of Waterloo 200 University Avenue West Waterloo Ontario N2L 3G1 Canada

Angelo Basile* and Adolfo Iulianelli Institute on Membrane Technology Italian National Research Council Via Pietro Bucci Cubo 17/C c/o University of Calabria 87030 Rende (CS) Italy Email: [email protected]


Professor Umberto Desideri Dipartimento di Ingegneria Industriale Università degli Studi di Perugia Via Duranti, 93 06125 Perugia Italy

Fausto Gallucci Fundamentals of Chemical Reaction Engineering IMPACT, Faculty of Science and Technology University of Twente P.O. Box 217 NL-7500 AE Enschede the Netherlands



Chapter 5

Chapter 6 Robert M. Davidson IEA Clean Coal Centre Gemini House 10-18 Putney Hill London SW15 6AA UK

Pietropaolo Morrone Department of Mechanical Engineering Via Pietro Bucci Cubo 44/C University of Calabria 87030 Rende (CS) Italy




Chapter 8

Chapter 11

Chris Higman Higman Consulting GmbH Sachsenstrasse 19 D-65824 Schwalbach Germany

Dr E.J. (Ben) Anthony CANMET Energy Technology Centre-Ottawa Natural Resources Canada 1 Haanel Drive Nepean Ontario K1A 1M1 Canada


Chapter 9

xv


Dr Philippe Mathieu Department of Aerospace and Mechanical Engineering University of Liège Chemin des Chevreuils 1 -bâtiment B52/3 Sart Tilman 4000 Liège 1 Belgium Email: [email protected]

Chapter 12 A. Aspelund Department of Energy and Process Engineering The Norwegian University of Science and Technology NO-7491, Trondheim Norway Email: [email protected]

Chapter 10 Steven C. Kluiters, Ruud W. van den Brink and Wim G. Haije* Energy research Centre of the Netherlands, ECN P.O. Box 1 1755 ZG Petten the Netherlands Email: [email protected]

Chapter 13 P. N. Seevam, J. M. Race and M. J. Downie* School of Marine Science and Technology Newcastle University Newcastle upon Tyne NE1 7RU UK Email: [email protected]


xvi


Chapter 14

Chapter 16

Dr Semere Solomon* and Todd Flach Det Norske Veritas (DNV) DNV – Research and Innovation N-1322, Høvik Norway

Dr Jean-Pierre Birat Arcelor Mittal Maizieres Research SA Voie Romaine BP 30320 57283 Maizières-lès-Metz Cedex France


Email: jean-pierre.birat@arcelormittal. com

Chapter 15 Dr Subhasis Ghoshal* Department of Civil Engineering McGill University 817 Sherbrooke Street West Montreal Quebec Canada H3A 2K6 Email: [email protected]

Dr Frank Zeman Center for Metropolitan Sustainability New York Institute of Technology 1855 Broadway New York NY 10023 USA Email: [email protected]


xvii


1 Generating power at high efficiency: Combined cycle technology for sustainable energy production Eric Jeffs 2 Advanced separation techniques for nuclear fuel reprocessing and radioactive waste treatment Edited by Kenneth L. Nash and Gregg J. Lumetta 3 Bioalcohol production: Biochemical conversion of lignocellulosic biomass Edited by K.W. Waldron 4 Understanding and mitigating ageing in nuclear power plants: Materials and operational aspects of plant life management (PLiM) Edited by Philip G. Tipping 5 Advanced power plant materials, design and technology Edited by Dermot Roddy 6 Stand-alone and hybrid wind energy systems: Technology, energy storage and applications Edited by J.K. Kaldellis 7 Biodiesel science and technology: From soil to oil Jan C.J. Bart, Natale Palmeri and Stefano Cavallaro 8 Developments and innovation in carbon dioxide (CO2) capture and storage technology Volume 1: Carbon dioxide (CO2) capture, transport and industrial applications Edited by M. Mercedes Maroto-Valer 9 Geological repository systems for safe disposal of spent nuclear fuels and radioactive waste Edited by Joonhong Ahn and Michael J. Apted 10 Wind energy systems: Optimising design and construction for safe and reliable operation Edited by John D. Sørensen and Jens N. Sørensen © Woodhead Publishing Limited, 2010

xviii


11 Solid oxide fuel cell technology: Principles, performance and operations Kevin Huang and John Bannister Goodenough 12 Handbook of advanced radioactive waste conditioning technologies Edited by Michael I. Ojovan 13 Nuclear reactor safety systems Edited by Dan Gabriel Cacuci 14 Materials for energy efficiency and thermal comfort in buildings Edited by Matthew R. Hall 15 Handbook of biofuels production: Processes and technology Edited by Rafael Luque, Juan Campelo and James Clark 16 Developments and innovation in carbon dioxide (CO2) capture and storage technology Volume 2: Carbon dioxide (CO2) storage and utilisation Edited by M. Mercedes Maroto-Valer 17 Oxy-fuel combustion for power generation and carbon dioxide (CO2) capture Edited by Ligang Zheng


xix

Foreword

In an ideal world, we wouldn’t need carbon capture and storage. But in an ideal world the inhabitants would have been quicker to spot that large-scale burning of fossil fuel could interfere with their planet’s carbon cycle and have serious consequences. By building a world economy over the last 150 years that flourished on the cheap and accessible energy that was available from coal and oil, we short-circuited the natural cycle. We have transferred from the solid Earth to the atmosphere huge quantities of carbon that would not otherwise have seen the light of day for many millions of years. We now know to our cost that the carbon cycle couples into the processes that control the Earth’s climate and that we have triggered rapid climate change. True, the Earth’s climate has always changed, but most natural change has happened sufficiently slowly for plants and animals to migrate or adapt to the new conditions. What we are doing is too fast to allow this. Although fossil fuels are the main cause of the rise in greenhouse gases in the atmosphere, there are also significant contributions from deforestation and changes in land use. Life on Earth depends on the benign greenhouse effect of our atmosphere. It provides surface temperatures that we do not find on neighbouring planets and that allow water to exist as ice, liquid and vapour. By burning fossil fuels, we increase the atmospheric concentration of CO2 which, along with other greenhouse gases, increases the greenhouse effect and increases mean global temperature including in the oceans. The atmospheric consequence of warming the oceans is the same as that of turning up the gas under a pan of gently simmering water – movements become faster and more violent. In weather terms this means more extreme weather conditions – storms, floods, droughts. Continental ice masses, particularly Antarctica and Greenland, begin to melt faster and contribute to a rise in sea level beyond that expected from thermal expansion. None of this is particularly good news and, although it is denied by a few, the evidence for the human influence on observed climate change is overwhelming. The more fossil fuel we burn the greater the rise in atmospheric CO2 and the worse the perturbation of the Earth’s climate. The problem is


xx

Foreword

that virtually all the world energy is supplied by fossil fuels, and weaning ourselves off them will take decades until we make the transition to other energy sources. The best scientific forecasts suggest that massive reductions in emissions have to be achieved by 2050 if there is to be any hope of containing damage to the climate. This is a problem that has been created largely by the developed world which owes its prosperity to the use of cheap and abundant fossil fuel. Although estimates vary, around two thirds of the ‘excess’ atmospheric CO2 is attributable to Europe and the USA. Today, however, we have new players. Developing countries whose economies are growing fast have rising energy requirements and often the cheapest available source of energy is coal, the fuel that carries the heaviest carbon cost per unit energy produced. It is urgent therefore that we find a way of managing emissions while we make the move to a low-carbon economy. The best way of managing emissions is not to produce them in the first place. This means that improved efficiency and energy conservation are vital. However, something has to be done about the ‘essential’ (for the moment) emissions that we cannot avoid. These come partly from vehicles that burn liquid fossil fuels and partly from a range of so-called fixed sources such as power stations, cement factories, oil refineries and a myriad of small local sources from office blocks to domestic houses. This is where carbon capture and storage (CCS) comes in. CCS is a group of technologies that are designed to capture and immobilise emissions from the larger fixed industrial sources. This involves separating the greenhouse gases from the other gases in the industrial exhaust streams and transporting them to a suitable site where they can be contained underground for many tens of thousands of years. This may sound relatively straightforward and indeed all the component technologies needed to achieve CCS have to a greater or lesser extent been demonstrated. The problem is that these technologies have their roots elsewhere and were developed with other ends in mind. It is only relatively recently that moves have been made to harness them together to achieve CCS. There is thus enormous scope for improvement of the systems. In reality there are three different sets of technologies required: the technology for separation of greenhouse gases from the exhaust gas stream at the point source, the pipeline or other means needed to transport the separated gas, and finally the technology for storing it. Most current concepts of storage involve pumping the gas underground into geological traps that have the demonstrated ability to retain gases for many tens of thousands of years. Of the three different activities, the first is likely to be the technically most challenging and the most expensive and currently appears likely to amount to between half and two thirds of the total cost of CCS. Overall, electricity generated in coal-fired power plant with its emissions reduced by


Foreword

xxi

90 % or more through CCS might be expected to cost between 30 and 50 % more than at present. Although these costs would be very unwelcome in the developed world, they would not be unbearable; in the developing world, however, they would be very difficult to accept. This means that there is an overriding urgency to reduce the cost of CCS. It is to be expected that with time costs will come down as engineers and operators gain experience, but more than slow incremental improvement is needed before CCS becomes deployed worldwide on a scale that can be expected to influence climate change. It would be wrong to assume, however, that the challenges are solely technical. Because CCS is a new activity, a suitable regulatory framework has to be developed and because CCS was never contemplated when the existing wider regulatory framework was established, there will certainly be conflicts to be resolved. The framework will have to cover the legal obligations and rights of all parties and the basis for licensing of all aspects of the operations. Work has already begun on these problems in a number of countries and within the EU. From a business point of view too, there are substantial logistical challenges. All three main elements identified above involve major capital expenditure and have lead times of at least five years. For a reasonably cost-effective system they need to come on line together. Furthermore, they are the responsibility of different consenting authorities for some of which CCS may not be their highest priority. This title addresses a number of the important challenges faced by CCS. They are formidable, but the most important step is to recognize them early and to plan ways of dealing with them. No one should pretend that CCS is a complete answer to the problem of fossil fuels and climate change but, conversely, it is unthinkable that we can manage that problem without CCS. Lord Oxburgh House of Lords Westminster London SW1, UK Email: [email protected]


xxii

1

Overview of carbon dioxide (CO2) capture and storage technology

S. B o u z a l a k o s and M. M e r c e d e s M a r o t o - V a l e r, University of Nottingham, UK Abstract: Carbon dioxide (CO2) capture and storage (CCS) is considered one of the most promising strategies to reduce CO2 emissions while enabling the continued use of fossil fuels and without compromising the security of electricity supply. This chapter first states the global CO2 emissions from power generation and points out that climate change is a serious and urgent issue. The chapter then discusses carbon management options and puts CCS technology into perspective. An account is then given on current plans to deploy large-scale CCS demonstration projects around the world and obstacles that need to be overcome to achieve the current commercialisation target of 2020. Innovation in research, development and deployment is increasingly becoming an important driver of both the mature and developing CCS technologies. This is clearly perceptible throughout the chapters of this book. The chapter closes by offering an outlook of the future trends and recommendations of sources of further information on CCS. Key words: carbon dioxide, CO2 capture and storage, CCS, climate change, fossil fuels, power generation.

1.1

Introduction

Fossil-fuel derived energy presently dominates most aspects of modern human activities and our current way of life, and is projected to remain the main energy source for the foreseeable future. However, the combustion of fossil fuels in stationary and mobile power sources produces large amounts of greenhouse gas (GHG) emissions, including carbon dioxide (CO2) which accounts for approximately 57 % carbon dioxide-equivalent (CO2-eq) of the GHG emissions from fossil fuel use (IPCC, 2007). CO2-equivalent emission is the amount of CO2 emission that would cause the same time-integrated radiative forcing, over a given time horizon, as an emitted amount of a long-lived GHG or a mixture of GHGs such as, for example, a mixture with methane (CH4) and nitrous oxide (N2O). The equivalent CO2 emission is calculated by multiplying a given emission of GHG by its Global Warming Potential (GWP) for the given time horizon, and for a mix of GHGs the CO2-eq is calculated by summing the equivalent CO2 emissions of each gas. It should be noted that while equivalent CO2 emissions is a standard and useful metric for comparing emissions of different GHGs, it does not imply the same climate change responses (IPCC, 2007). 1 © Woodhead Publishing Limited, 2010

2

Developments and innovation in CCS technology

A recent study by McKinsey & Company (2008) states that approximately 47 % (approximately 2 GtCO2 in 2007) of total European CO2 emissions could be addressed by the application of CO2 capture and storage (CCS) technologies. This includes predominantly large stationary sources, with coal power stations accounting for 52 %. On a global scale, various recent reports estimate that CCS could potentially abate between 1.4 GtCO2 (Stern, 2006) and 4 GtCO2 (IEA, 2007) by 2030. With the increasing energy demand witnessed and projected, already soaring atmospheric CO2 emissions will continue to rise. As CO2 emissions have been unequivocally linked to global warming and climate change (IPCC, 2007), mitigation measures are a matter of urgency. A range of technologies, collectively termed CO2 capture and storage (CCS), have been identified as a critical option in the portfolio of solutions available to combat climate change, allowing for the reduction of CO2 emissions while enabling the continued use of fossil fuels (IPCC, 2005). CCS involves three main steps: capture, transportation and storage. Overall, the technologies are fairly mature and plans are underway for their largescale demonstration in the near future. Technological barriers are often a monetary concern, particularly for capture technologies that account for roughly two-thirds of the total cost of CCS. Given the fact that CCS is a relatively new activity for both power plant operators and governments, a suitable regulatory framework has to be put in place to facilitate the wider deployment of the technology. The development and use of CO2 capture technology could take place within existing regulatory frameworks for power stations; however, the main issues for regulation of CCS concern activities offshore, especially geological storage, and transportation.

1.2

Greenhouse gas emissions and global climate change

The greenhouse effect is necessary to sustain life on earth, and in its absence the average temperature on the planet would be around –18 °C. The major GHGs in terms of total emissions in 2004 were CO2 from fossil fuel use (56.6 % CO2-eq), CO2 from deforestation, decay of biomass etc. (17.3 % CO2-eq); methane (CH4) (14.3 % CO2-eq); and nitrous oxide (N2O) (7.9 % CO2-eq) (IPCC, 2007). Carbon dioxide is the most important anthropogenic greenhouse gas, even though it is not as harmful as CH4 which is produced from fossil fuel combustion in smaller amounts (IPCC, 2007). Climate scientists have no doubt that the earth’s climate will warm in response to further release of man-made greenhouse gases into the atmosphere by intensifying the greenhouse effect. However, there are uncertainties about the extent of warming that will occur and what the regional impacts of this will be, precisely. To date, the most credible estimation of future climate states


Overview of carbon dioxide capture and storage technology

3

comes from mathematical climate models based on physical approximations. Despite uncertainties, all climate models predict substantial climate warming under greenhouse gas increases (IPCC, 2007). Taking into account the range of human activities, power stations are the largest contributor of anthropogenic CO2 emissions with levels reaching approximately 0.17 GtCO2 in the UK in 2008 (BERR, 2009). This level of CO2 emissions is further emphasised by Fig. 1.1, which shows that approximately 29 % of CO2 emissions for 2006 in EU-15 countries were attributed to power generation. According to the International Energy Outlook 2008 (EIA, 2008), the total world energy-related CO2 emissions for 2005 were estimated at 28.1 GtCO2, and are projected to increase by an average of 1.7 % per annum from 2005–2030. Concentrations of atmospheric CO2 have been increasing from approximately 280 ppmv in the pre-industrial era (Fig. 1.2a) to 389.47 ppmv, as measured in April 2009 (Fig. 1.2b). The detrimental effects of increasing CO2 levels on global climate have been well documented, and it is clear that there is a need to reduce CO2 levels (Stocker and Schmittner, 1997; Palmer and Räisänen, 2002; Karl and Trenberth, 2003; Stern, 2006). According to the latest United Nations Intergovernmental Panel on Climate Change (IPCC) report (IPCC, 2007), climate change has been proven to be unequivocally linked to human activity from observations of increases in global average air and ocean temperatures, rising global average sea levels and widespread melting of sea-ice in the Arctic (Fig. 1.3).

Petroleum refining 3 %

Cement production 2 %

Iron and steel production 2 %

Other 8 % Public electricity and heat production 29 %

Commercial/ industrial 5 %

Residential 12 %

Manufacturing industries and construction 16 %

Road transportation 23 %

1.1 EU-15 CO2 emissions by source for 2006 (total emissions = 3.46 GtCO2) (EIA, 2008).


4

320 315


310


CO2 concentration (ppmv)

305 300 295 290 285 280 275 270 1000 1060

1120

1180 1240

1300

1360 1420

1480 1540 Year (a)

1600 1660

1720

1780 1840

1900

1960

1.2 (a) Average annual atmospheric CO2 concentrations from Antarctic ice and firn from 1010–1960 (Etheridge et al., 1996) (b) Average annual atmospheric CO2 concentrations based on direct measurements at Mauna Loa Observatory from 1960–2009 (Dr Pieter Tans, NOAA/ESRL, www.esrl.noaa.gov/gmd/ccgg/trends).

400 389.47 ppmv (April 2009)

390 385 380


CO2 concentration (ppmv)

375 370 365 360 355 350 345 340 335 330 325 320 315 310 1960

1965

1970

1975

1980

1985 Year (b)

1990

1995

2000

2005

2010


395

1.2 Continued

5


14.5

0.0

14.0

–0.5

13.5 (a)

50 0 (mm)

Difference from 1961–1990

(°C)

0.5

Temperature (°C)

6

–50 –100 –150 (b)

0

36

–4

1850

(million km2)

40

2

(million km )

4

32 1900

Year (c)

1950

2000

1.3 Observed changes in (a) global average surface temperature; (b) global average sea level; and (c) Northern Hemisphere snow cover for March–April (IPCC, 2007).

Despite the increasing atmospheric CO2 concentrations mentioned in the previous paragraph, energy-related CO2 intensities, expressed as emissions per unit of economic output (Table 1.1), have been projected to improve (i.e., decline) from 2005–2030 as world economies strive to use energy more efficiently. Carbon dioxide intensity by non-OECD countries is projected to decline by an average of 2.6 % per year, from 529 metric tonnes per million dollars of GDP in 2005 to 274 metric tonnes per million dollars of GDP in 2030. For all OECD countries average CO2 intensity in 2030 is projected to be 296 metric tonnes per million dollars of GDP. The average for the entire



7

Table 1.1 Carbon dioxide intensity by region and country, 1980–2030, in metric tonnes per million 2000 US dollars of gross domestic product (GDP) (EIA, 2008) Region History Projections

Average annual percent change

1980 1990 2005 2010 2015 2020 2025 2030 1990– 2005

2005– 2030

OECD United States Canada Mexico Europe Japan South Korea Australia/ New Zealand Non-OECD Europe/Eurasia Russia Other Asia China India Other Middle East Africa Central and South America Brazil Other Total world

732 916 867 394 674 482 942 694

565 701 679 441 508 353 729 679

461 544 607 381 383 358 670 633

411 483 563 337 343 316 580 558

379 439 521 312 318 297 521 500

347 399 486 288 290 284 464 449

319 366 453 266 264 273 424 404

296 339 422 247 241 262 396 365

–1.3 –1.7 –0.7 –1.0 –1.9 0.1 –0.6 –0.5

–1.8 –1.9 –1.4 –1.7 –1.8 –1.2 –2.1 –2.2

694 1019 900 1215 755 1959 295 400 450 398 317

711 1166 1060 1339 624 1242 333 352 854 448 310

529 804 836 762 498 693 287 360 903 421 305

440 615 649 573 411 552 221 313 827 362 290

388 531 554 504 363 478 189 299 747 327 262

344 469 494 440 322 421 165 270 679 292 234

306 410 432 385 289 373 148 246 605 255 209

274 368 392 342 261 334 135 224 539 220 187

–2.0 –2.4 –1.6 –3.7 –1.5 –3.8 –1.0 0.1 0.4 –0.4 –0.1

–2.6 –3.1 –3.0 –3.2 –2.5 –2.9 –3.0 –1.9 –2.0 –2.6 –1.9

212 403

211 219 398 379

224 342

208 303

192 267

175 234

162 205

0.2 –0.3

–1.2 –2.4

716

624 494 427

384

345

311

282

–1.6

–2.2

world is projected to fall from 494 metric tonnes per million dollars of GDP in 2005 to 282 metric tonnes in 2030 (EIA, 2008). The United Nations Framework Convention on Climate Change (UNFCCC) in 1994, through which the Kyoto Protocol entered into full force in 2005, commits nations to achieving a: ‘stabilisation of greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system.’ For instance, the Stern review comments that the worst impacts of climate change could be substantially reduced if greenhouse gas concentrations were to be stabilised between 450 and 550 ppm CO2-eq (Stern, 2006). In November 2008, the UK Climate Change Act became law in the UK, setting up a target of 80 % reduction over 1990 CO2 levels by 2050; making the UK the first country to set such a long-range and significant carbon reduction target into law (DECC, 2009). Further global commitments were discussed at the 15th Conference of the Parties (COP15) under the auspices of the United Nations Framework Convention on Climate


8


Change (UNFCCC), in Copenhagen, in December 2009, with the intention to create a legally binding, international treaty to replace the Kyoto Protocol that expires in 2012. Negotiations in the run-up to COP15 showed disagreement on how to tackle climate change and the expectations for a legally binding agreement were lowered (Climatico, 2010). An accord was reached that, although it has significant elements, is not legally binding. The key elements of the accord include the objective to keep the maximum temperature rise to below 2 °C, commitment to list developed countries emission reduction targets and mitigation actions for developing countries, and finance to kick start action in the developing world to fight climate change (http://unfccc. int/meetings/cop_15/items/5257.php). The success of COP15 will depend on the challenge to develop the Copenhagen Accord into a legally binding treaty in 2010 (COP16 in Mexico).

1.3

Carbon management and stabilisation routes

The management of increasing CO2 emissions typically revolves around three broad (but closely related) strategies as possible solutions, namely: (i) switching to a low-carbon economy, i.e., relying on renewable and/ or alternative sources of energy; (ii) increasing the efficiency and energy conservation of our current fossil-fuel energy generation; and (iii) applying CCS technologies to reduce CO2 emissions in order to bridge the gap presented by working to change from our current fossil-fuel dependency to a fully sustainable, low-carbon future. These strategies are discussed in more detail below. Fossil fuels (mainly coal) account for approximately 86 % of the overall world energy use (IEA, 2007; Orr, Jr, 2009), and are foreseen to remain the dominant energy source for the largest part of the 21st century (McKinsey & Company, 2008). Although there is significant concern about the increasing amount of CO2 that will be emitted (IPCC, 2005, 2007; Bachu, 2008b), alternative or renewable energy sources still have fundamental hurdles to overcome. For instance, there are many security and environmental issues still associated with nuclear energy generation, while on the other hand wind, solar, water, wave and geothermal power cannot currently provide sufficient amounts of base-load electricity generation to displace fossil-fuel power. Many of these technologies also rely on the availability of resources, which depends on the geographical location and attributes of a country (Martinot et al., 2007). Furthermore, the use of biomass leaves open the question of the correct technology to be implemented if heavy energy demand is to be met. Hydrogen is likely to be an important energy-carrier in the future (Edwards et al., 2008), but it requires reliable and high-capacity production that is independent of the decarbonisation of fossil fuels, as well as improved storage technologies, to achieve this role in a sustainable future.



9

For the time being, the reduction of CO2 emissions can be achieved by implementing efficient energy strategies. Innovative technologies for power generation, such as Integrated Gasification Combined Cycle (IGCC), may increase the efficiency of conversion of the fuel’s chemical energy from 28–32 % of the recent past to 52 %. Supercritical and ultra-supercritical coalfired power plant technology may also offer a major option for high-efficiency and low-emission power generation, with efficiency projected in the region of 50 % and approximately 30 % projected CO2 emissions reduction. Fuel flexibility can also contribute to the reduction of emissions. For instance, moving from coal to oil to liquefied natural gas (LNG), the amount of CO2 emitted per kWh goes down from 1 to 0.75 to 0.5 kg, respectively. However, even with increased efficiency and reduction in emissions, the rapid expansion of the worldwide demand for energy will ultimately produce a net increase in CO2 emissions. A net reduction in emissions would require a rigorous carbon management strategy to be applied worldwide. CO2 capture and storage (CCS) is a technically feasible strategy to reduce anthropogenic CO2 emissions from large point sources, and particularly fossil fuel-fired power plants, by up to 90 % (IPCC, 2005). One of the key features of this technology is that it allows for the continued use of fossil fuels, including coal which is relatively cheap and abundant, while simultaneously reducing CO2 emissions to the atmosphere (IRGC, 2008). Overall, CCS consists of three main steps: separating and capturing CO2 from other exhaust gases; compressing the CO2 to supercritical conditions in order to transport it to its storage location; and final isolation from the atmosphere by a variety of methods as illustrated in Fig. 1.4. The carbon mitigation potential of the different strategies mentioned above requires a fixed timeframe. For instance, Pacala and Socolow (2004) propose that a 50-year perspective could be long enough to allow changes in infrastructure and consumption patterns but short enough to be heavily influenced by decisions made today. Assuming the world continues on its current predicted path, i.e., business as usual (BAU), it is predicted that CO2 emissions will roughly double by 2054. Stabilising GHG concentrations in the region of 500 ± 50 ppm has been proposed as the target level to prevent the most damaging climate change. Avoiding the doubling of CO2 levels in the business as usual case, in order to reduce substantially the likelihood of the most dramatic consequences of climate change, would require a monumental effort. Committing to a CO2 emissions trajectory approximating a flat path requires an amount of CO2 emissions reduction in 2054 roughly equal to all CO2 emissions today (Fig. 1.5). To assess the potential of the various carbon mitigation strategies, Pacala and Socolow (2004) introduced the concept of stabilisation wedges (Fig. 1.5). The difference between currently predicted path and flat path from present to 2054 gives a triangle of emissions to be avoided, a total of nearly 200


10

Biomass

Cement/steel/ refineries, etc.

Gas Natural gas + CO2 capture

Oil

Mineral carbonation

Coal

CO2

Petrochemical plants + CO2 capture Future H2 use

Electricity generation

Industrial uses

Ocean storage (ship or pipeline)

1.4 Schematic diagram of possible CCS systems showing the sources for which CCS might be relevant, transport of CO 2 and storage options (IPCC, 2005).



Gas to domestic supply


Billions of tonnes of carbon emitted per year

14

11

7 wedges

C

re ur

nt

ly

pr

ed

ic

d te

pa

are needed to build the stabilisation triangle

th

Stabilisation triangle 1 wedge 1 ‘wedge’

7

Flat path 2004

Year

2054

avoids 1 billion tonnes of carbon emissions per year by 2054

1.5 Stabilisation wedges concept for reducing carbon emissions by 2054. (Socolow et al., 2004).

GtC. The stabilisation triangle can be further divided into seven wedges of equal area each representing a reduction of 1 GtC/year by 2054. This is based on technologies that have the potential to contribute a full wedge to carbon mitigation. CCS technology prevents about 90 % of fossil carbon from reaching the atmosphere, so a wedge would be provided by the installation of CCS at 800 GW of base-load coal plants by 2054 or 1600 GW of baseload natural gas plants.

1.4

Development and innovation in carbon dioxide (CO2) capture and transport technology

Although many of the component technologies for CCS are fairly mature, there are no, as yet, fully integrated commercial applications (Fig. 1.6). There are, however, a number of pilot-scale CCS projects around the world demonstrating confidence in the technology (Table 1.2). The UK government launched a competition in 2007 to build one of the first commercial-scale CCS projects by 2014 (BERR, 2008; APGTF, 2009). As can be seen in Table 1.2, other world governments and energy corporations are focusing on similar incentives to facilitate widespread deployment of CCS technologies in the near future. A programme of 10–12 demonstrations has also been called for in the EU to be operational by 2015, in line with the target of commercialisation of CCS by 2020 (APGTF, 2009). China, which is overtaking the USA in CO2 emissions (approximately 6.0 GtCO2 in comparison to the 5.9 GtCO2 by the USA in 2006) from consumption and flaring of fossil fuels (EIA, 2008), is making significant


Concept

12

Stage of development Lab testing

Commercial refinements needed


First projects are coming online now

Commercial

Component technologies are mature; integrated platform to be proven

Several projects are operational (e.g., Weyburn (Canada)). EU has limited EOR potential

Post-combustion Membranes Chemical looping

Oxyfuel

CO2– EGR

Pre-combustion

Saline aquifers

Sleipner (Norway) field has been operational for around 10 years

Depleted oil and gas fields

Have been used for seasonal gas storage for decades

CO2–EOR Transport Transport offshore onshore

USA has existing CO2 pipeline network of more than 5000 km Capture Transport Storage

1.6 Stage of CCS component technologies (EGR = enhanced gas recovery) (McKinsey & Company, 2008).


Potential future breakthrough technologies

Demonstration

Table 1.2 Global CO2 capture and storage projects (MIT, 2008) Location

Leader

Feedstock

Size (MW)

Capture process

CO2 fate

Start-up

Total Lacq Schwarze Pumpe AEP Alstom Mountaineer Callide-A Oxy Fuel GreenGen Williston Kimberlina NZEC AEP Alstom Northeastern Sargas Husnes Scottish & Southern Energy Ferrybridge Naturkraft Kårstø Fort Nelson ZeroGen Antelope Valley WA Parish UAE Project Appalachian Power Wallula Energy Resource Centre RWE npower Tilbury Tenaska HECA UK CCS Project Statoil Mongstad RWE Zero CO2 Boundary Dam Monash Energy

France Germany USA Australia China USA USA China USA Norway UK

Total Vattenfall AEP CS Energy GreenGen PCOR CES UK & China AEP Sargas SSE

Oil Coal Coal Coal Coal Coal Coal Coal Coal Coal Coal

35 30/300/1000 30 30 250/800 450 50 TBD 200 400 500

Oxy Oxy Post Oxy Pre Post Oxy TBD Post Post Post

Seq Seq/EOR Seq Seq Seq EOR Seq Seq EOR EOR Seq

2008 2008 2008 2009 2009 2009–15 2010 2010 2011 2011 2011–12

Norway Canada Australia USA USA UAE USA USA

Naturkraft PCOR ZeroGen Basin Electric NRG Energy Masdar AEP Wallula Energy

Gas Gas Coal Coal Coal Gas Coal Coal

420 Gas process 100 120 125 420 629 600–700

Post Pre Pre Post Post Pre Pre Pre

TBD Brine res Seq EOR EOR EOR TBD Seq

2011–12 2011 2012 2012 2012 2012 2012 2013

UK USA USA UK Norway Germany Canada Australia

RWE Tenaska HEI TBD Statoil RWE SaskPower Monash

Coal Coal Petcoke Coal Gas Coal Coal Coal

1600 600 390 300–400 630 CHP 450 100 60 k bpd

Post Post Post Post Post Pre Oxy Pre

Seq EOR EOR Seq Seq Seq EOR Seq

2013 2014 2014 2014 2014 2015 2015 2016

13

Notes: Seq = sequestration; EOR = enhanced oil recovery; TBD = to be decided; Brine res = brine reservoir; Gas process = Gas processing; Pre = pre-combustion; Post = post-combustion; Oxy = oxyfuel combustion.



Project name

14


progress in CCS projects. GreenGen is a partnership between the Chinese government, Chinese energy companies and Peabody energy. It is planned to deploy incrementally to a 400 MW IGCC power plant with CCS by 2020. Near Zero Emission Coal (NZEC) is another CCS project between China and the European Union, and has the goal of deploying a coal-fuelled power plant with CCS by 2020. The Australian government in April 2009 formally launched the Global Carbon Capture and Storage Institute (GCCSI), a new initiative aimed at accelerating the worldwide commercial deployment of CCS technologies. The G8 countries have committed to the development of 20 large-scale CCS projects to be operational by 2020, with the GCCSI playing a vital role in developing the partnerships to make these projects a reality. The Department of Energy (DOE) in the USA has formed a nationwide network of Regional Carbon Sequestration Partnerships (RCSP) to help determine the best approaches for capturing and storing GHG. The RCSP initiative is currently in the development phase (2008–2017) to conduct large-volume carbon storage tests.

1.4.1 Carbon dioxide (CO2) capture and storage economics, regulation and planning A major difficulty in deploying CCS technology is the reluctance of corporations to invest given the absence of financial cost associated with greenhouse gas emissions, uncertainty over the future regulations governing coal-burning power plants and CO2 storage, and the need for additional research, development and demonstration (Gibbins and Chalmers, 2008). The absence of governmental regulations and policy frameworks creates additional uncertainty for companies considering investment in CCS (Bachu, 2008b). Volume 1, Chapter 3, deConinck, provides a regulatory analysis and outlook for CCS technologies. Commercial-scale CCS deployment will require a regime to manage risks as well as supporting policies to facilitate technology investment (IRGC, 2008). Public perception and support are also vital for actual implementation of CCS technologies. The main concerns society has over CCS are related to safety issues and the extent to which CCS provides a solution to climate change (Gough and Shackley, 2005; van Alphen et al., 2007). Estimating the cost of CCS technologies involves a high degree of uncertainty over how these costs may develop over time and in terms of potential variations in the technical requirements, scale and application of projects. McKinsey & Company (2008) have recently released a report in which CCS costs are predicted, based on a case-study approach. According to the main findings of the report, early commercial CCS projects, potentially around 2020, are estimated to cost 735–50 per tonne CO2 abated (Fig. 1.7). By far, CO2 capture is the most expensive component and may account for



15

Assumption

1 Capture

2 Transport

25–32

4–6

∑ CO2 capture rate of 90–92 % ∑ CCS efficiency penalty of 7–12 % points ∑ Same utilisation as non-CCS plant (86 %) ∑ CO2 compression at capture site ∑ Transport through onshore/offshore pipeline network of 200/300 km in supercritical state with no intermediate booster station ∑ Use of carbon steel (assumed sufficiently dry CO2) ∑ Injection depth of 1500 m in supercritical state 4–12 ∑ Use of carbon steel (assumed sufficiently dry CO2) ∑ Vertical well for onshore/directional for offshore

3 Storage

35–50*

Total *Ranges are rounded to 5 on totals.

1.7 Total cost of early commercial projects – reference case (7/tonne CO2 abated; ranges include on- and offshore) (McKinsey & Company, 2008).

up to two-thirds of the total cost of a CCS project (725–32 per tonne CO2 abated). Further information on economic analyses of CCS technologies is provided in Vol. 1, Chapter 2, Ogden, and further information on planning and economic modelling for CO2 capture and reduction is provided in Vol. 1, Chapter 4, Elkamel, Mirzaesmaeeli, Croiset and Douglas.

1.4.2 Carbon dioxide (CO2) capture processes and technologies in power plants There are three main technologies for carbon capture from fossil fuel power plants: ∑ after combustion (post-combustion); ∑ decarbonisation of the fuel before combustion (pre-combustion); and, ∑ burning the fuel in pure oxygen (oxyfuel combustion). Post- and pre-combustion processes include chemical and physical capture of CO2 by absorption (Vol. 1, Chapter 5, Desideri) and adsorption (Vol. 1, Chapter 6, Davidson), as well as CO2 separation by membranes (Vol. 1, Chapter 7, Basile, Gallucci, Morrone and Iulianelli) and gasification of fuels syngas/hydrogen for combustion and CO2 for capture (Vol. 1, Chapter 8, Higman). Under oxyfuel combustion conditions, fuel is burnt in pure oxygen rather than air, resulting in more complete combustion and producing a flue gas constituting approximately 90 % CO2 for easier separation (oxyfuel


16


combustion) (Vol. 1, Chapter 9, Mathieu). Advanced oxygen separation and generation systems (Vol. 1, Chapter 10, Kluiters, van den Brink and Haije), and chemical looping combustion systems (Vol. 1, Chapter 11, Anthony) are also being developed as promising alternatives to the more common post- and pre-combustion capture processes.

1.4.3 Carbon dioxide (CO2) compression, transport and injection processes and technologies Transport of pressurised supercritical CO2 from the point of capture to storage sites through an extensive pipeline network is considered to be the most cost-effective and reliable method for onshore CCS (Svensson et al., 2004). Although there may be a risk associated with potential leakage through infrastructure failure or third-party intrusion, there is a lot of experience from the petroleum industry in the USA and Canada in the transport of natural gas, hydrocarbon liquids and CO2 for enhanced oil recovery (EOR) that could be applicable to CO2 transport for CCS. Nevertheless, long-distance CO2 pipeline networks, both onshore and offshore, have technical challenges (e.g., safety and reliability) that need to be faced in order to minimise risks to the environment and human health. Further innovation is focused on pipeline materials, infrastructure and modelling studies to minimise the risk of failure and to yield a better understanding of the consequences and behaviour of a pipeline failure (Mazzoldi et al., 2008). Theses issues and further details on the various compression, transport and injection technologies applicable to CCS systems are respectively discussed in Vol. 1, Chapter 12, Aspelund; Chapter 13, Downie, Race and Seevam; and Chapter 14, Solomon and Flach.

1.4.4. Industrial applications of carbon dioxide (CO2) capture and storage technologies Wider implementation of CCS is being encouraged in other industries responsible for significant contribution to global CO2 emissions, e.g., the cement and concrete industry (Vol. 1, Chapter 15, Ghoshal and Zeman), and the iron and steel industry (Vol. 1, Chapter 16, Birat). It is estimated that the cement industry is responsible for approximately 5 % of global CO2 emissions (IPCC, 2005). The reduction of CO2 emissions from cement production is currently being addressed by looking into post-combustion and oxygen combustion capture, and using the CO2 for accelerated curing of concrete products and cement-based waste stabilisation/solidification. The iron and steel industry is also responsible for another 6–7 % of global CO2 emissions (IPCC, 2005). Strategies to control CO2 emissions have focused on energy conservation measures. However, further development and incorporation of CCS systems



17

need to be adopted to further reduce CO2 emissions. This involves post- and pre-combustion capture processes in the core of the blast furnace, with the potential of retrofitting existing steel mills from the 2020s onwards. New technologies may also be developed with CCS, and possibly without relying on CCS, through the use of hydrogen, electricity or biomass.

1.5

Development and innovation in carbon dioxide (CO2) storage and utilisation technology

1.5.1 Geological sequestration of carbon dioxide (CO2) Various options are possible for final storage of CO2. At present, injection into underground geological formations is the most promising and developed method (Holloway, 2005; IPCC, 2005; Bachu, 2008a,b), although these formations naturally need to be characterised and screened to ensure longterm sequestration (Vol. 2, Chapter 2 Bachu). There are three main types of proposed underground storage site: deep saline aquifers (Vol. 2, Chapter 3, Rosenbauer and Thomas); depleted oil/gas reservoirs and enhanced oil recovery (EOR) (Vol. 2, Chapter 4, Kovscek and Vega); and deep unmineable coal seams (Vol. 2, Chapter 5, Mazzotti, Pini, Storti and Burlini). Geological storage combines a number of engineering processes to ensure safe and long-term isolation of CO2 from the atmosphere. Deep saline aquifers are likely to be the most promising of other geological options, but there is still uncertainty regarding their capacity and geological/geochemical properties. To address the issue, innovative research is being focused to better understand the geochemical reactions between CO2, impurity gases, formation brine, host rocks and cap rocks. Depleted oil and gas reservoirs are frequently said to be the likely first category of geological formation to inject CO2 owing largely to the added benefit of EOR. It is estimated that 80 % of oil reservoirs worldwide might be suitable for CO2 injection for EOR. Enhanced Coal Bed Methane (ECBM) recovery is a technique under investigation for storing CO2 in unmineable coal seams with the added benefit of methane production.

1.5.2 Maximising and verifying carbon dioxide (CO2) storage in underground reservoirs Petrographical studies and the established body of knowledge concerning CO2 storage/migration mechanisms in geological media from the oil industry combine to improve our understanding of the CO2 injection design approaches that can be adopted to maximise CO2 storage and/or EOR in underground reservoirs (Vol. 2, Chapter 6, Blunt, Qi and LaForce). Similarly, improved methods of sealing underground reservoirs for CO2 trapping have been


18


informed by our understanding of gas leakage mechanisms from geological media (Vol. 2, Chapter 7, Meckel). Improved measurement, monitoring and verification (MMV) (Vol. 2, Chapter 8, Chadwick) and modelling (Vol. 2, Chapter 9, Pruess, Birkholzer and Zhou) techniques are being developed to verify storage and to prove long-term storage security. The environmental impact and safety of CO2 storage is constantly being developed to quantify risk, and better understand and minimise potential leakage of CO2 into freshwater aquifers or back into the atmosphere.

1.5.3 Terrestrial and ocean sequestration of carbon dioxide (CO2) and environmental impacts Another option for CO2 storage is terrestrial sequestration which, along with CO2 emissions reduction, has such ancillary benefits as increased agronomic productivity through CO2 use by plants (Vol. 2, Chapter 10, Lal). While CO2 sequestration has been noted to present benefits to terrestrial ecosystems, up to a point, there remain risks of CO2 leakage from underground reservoirs, and increased levels of CO2 present greater impacts and risks, including the potential for phytotoxicity whereby affected plant species will no longer grow (Vol. 2, Chapter 12, Steven, Smith and Colls). Injection into the deep ocean has also been proposed (Vol. 2, Chapter 11, Golomb and Pennell) but is associated with many more risks to marine ecosystems (including risks of CO2 leakage from subsea reservoirs) that have not yet been adequately researched (Vol. 2, Chapter 13, Blackford, Widdicombe, Lowe and Chen).

1.5.4 Advanced concepts for carbon dioxide (CO2) storage and utilisation Alternative, less conventional, routes for CO2 storage and utilisation are receiving increased attention, including CO2 utilisation by industry, mineralisation/mineral carbonation, biofixation of CO2 by microorganisms and photocatalytic reduction of CO2. The majority of these technologies are still in the research phase, apart from mineralisation/mineral carbonation which is almost ready for pilot-scale demonstration within the next few years. The technological barriers that usually render the process financially non-viable are where further innovation and breakthroughs are needed if these technologies are to have any chance of being integrated into the CCS agenda in the near future. Industrial fixation in inorganic carbonates (mineral carbonation) could have an increasingly important role in future CCS operations, particularly if waste materials are put to use (Vol. 2, Chapter 16, Zevenhoven and Fagerlund).



19

Some industrial processes might also utilise and store small amounts of captured CO2 in manufactured products (Vol. 2, Chapter 14, Aresta). In biofixation of CO2, microorganisms use CO2 for cell growth, which can be achieved both in natural habitats as well as in controlled systems such as microalgal farming, while the potential for CO2 conversion to biofuels by microorganisms is also undergoing further research (Vol. 2, Chapter 15, Wang and Lan). The future development of photocatalytic reduction processes for CO2 presents the potential for a similar benefit through the conversion of CO2 to hydrocarbons using artificial light or sunlight, in a process similar to photosynthesis (Vol. 2, Chapter 17, Wu).

1.6

Future trends

As previously mentioned in this chapter, it is highly likely that coal and other fossil fuels will dominate worldwide power generation for the foreseeable future. In order to maintain current standards of living with the ever increasing demand for energy, CCS must play a vital role in reducing CO2 emissions to the atmosphere to mitigate the potential for further global warming and climate change until a low-carbon economy can be fully implemented. At present, however, CCS cannot justify itself on an economic basis alone owing to the high energy penalty and associated costs of installation, and it may remain unprofitable until policy issues are decided and technologies further developed. More research, development and deployment is clearly necessary to make CCS more realistic (in terms of economic costs) to occur on a large scale. Industries are increasingly focusing their attention on CCS technologies, with a large number of pilot-scale projects currently operational and many more being planned to begin within the next five years. International collaboration is now considered a key element in delivering CCS commercialisation targets. It is anticipated that CCS will be demonstrated on a large scale by around 2012, and by the 2020s there will be some tens of CCS plants and some hundreds by the 2030s around the world. Although in comparison to the total number of fossil-fuel fired power stations, these figures are quite low, these developments can be regarded as significant progress towards implementing effective global warming mitigation measures. Legal and regulatory frameworks are underway and are planned to be enforced from about 2010–2012; these will help to justify more robust investment decisions and to minimise barriers for deployment in the near future. For example, the European Commission has recently proposed a Directive on CCS (COM; EC, 2008) to enable environmentally-safe capture and geological storage of CO2 in the EU as part of a major legislative package of measures to achieve the EU’s emissions targets, mitigate the potential effects of climate change and promote renewable energy up to


20


2020 and beyond. Although this Directive provides a regulatory framework specifically for geological storage, additionally, CO2 capture is regulated under Commission Directive 96/61/EC concerning integrated pollution prevention and control (IPPC) and Commission Directive 85/337/EEC on the effects of certain public and private projects on the environment. The market for CCS is huge, and by about 2020 CCS should have the potential to develop into a significant and competitive option to simultaneously reduce CO2 emissions and promote economic development, energy security and air quality in many countries for decades to come. Compared to a business-as-usual case, CCS could reduce energy-related CO2 emissions by up to a half by 2050. This largely relies on the commercial value of carbon exceeding the cost of CCS technologies and eventually on other incentives for implementation. CCS could also present increased revenue opportunities in the future, particularly where CO2 can be sold to nearby oil field operators for EOR. Other future industrial applications are likely to develop, and business opportunities may include managing CCS applications for hydrogen production, fuel cell applications, emission trading and CO2 storage (WEC, 2007).

1.7

Sources of further information and advice

There is now a plethora of sources for information on CCS. This includes a range of reports, articles published in academic journals, international symposia, research organisations/institutions, professional bodies, popular science books, and the World Wide Web. The ones mentioned below are by no means exhaustive, but merely indicative of the variety available to anyone interested in this field. Each chapter of the book provides additional sources of information. Probably one of the most well-known publications on CCS is the Special Report by the IPCC in 2005. This was the first major reporting of a consortium of leading professionals in the form of an intergovernmental scientific body presenting the urgency to act upon the global warming phenomenon and the role of CCS. This was followed by another renowned publication, The Stern Review (Stern, 2006), on the economics of climate change. Two recent books presenting a holistic account of CCS technologies are available by Wilson and Gerard (2007) and Shackley and Gough (2006). An earlier account on the science and technology of CO2 mitigation is available by Halmann and Steinberg (1999). Marini (2007), Baines and Worden (2004) and Wanty and Seal II (2004) provide extensive information on geological sequestration of CO2. A series of reports are available from many leading professional and research organisations/institutions and government bodies/departments. Some recent examples are mentioned below and other reports have been previously



21

referenced in this chapter. The World Resources Institute (WRI) has released a publication providing guidelines for CO2 capture, transportation and storage (WRI, 2008). McKinsey & Company (2008) address the economics of CCS and predict the cost of CCS technologies in their early days of commercialisation. The International Energy Agency Greenhouse Gas Programme (IEA GHG) holds a large collection of technical reports on CCS and it is advisable to check their website for further information (www.ieagreen.org.uk). The Energy Information Administration (EIA) produces annual international energy outlook reports (EIA, 2008) providing vital information on global energy trends and predictions to 2030. Finally, the UK Advanced Power Generation Technology Forum (APGTF) produced an excellent document in April 2009 providing guidance and advice on research, development and deployment needs of carbon abatement technologies (CATs) for fossil fuels (APGTF, 2009). Academic peer-reviewed articles on CCS are becoming increasingly popular in many journal publications. Elsevier launched a journal in 2008 (International Journal on Greenhouse Gas Control) specifically targeting papers on CCS and related topics, and it has been well received by the scientific community. The Carbon Capture Journal is another recent magazine that aims to inform on developments in CCS and related government policy. Other journals popular amongst researchers include: Energy Conversion & Management (Elsevier); Fuel Processing Technology (Elsevier); Fuel (Elsevier); Energy Policy (Elsevier); Environmental Science & Technology (American Chemical Society); Energy & Fuels (American Chemical Society); Energy & Environmental Science (Royal Society of Chemistry) and Proceedings of ICE – Energy (Institution of Civil Engineers). Several symposia are available to exchange ideas and present the state of the art on research and developments in CCS. The bi-annual International Greenhouse Gas Technologies conference (GHGT) is probably the main venue attracting thousands of delegates. Other well-known venues include the Annual Conference on Carbon Capture and Sequestration (USA); the Platts Annual European Carbon Capture and Storage Conference; and the International Conference on Carbon Dioxide Utilisation (ICCDU). The above mentioned events are dedicated to CCS; however, there are many more conferences in related fields that cover themes on CCS. The main organisations/institutions and centres involved with CCS research and development include the: Centre for Innovation in Carbon Capture and Storage (CICCS) (UK); Scottish Centre for Carbon Storage (UK); UK Energy Research Centre (UKERC); British Geological Survey (BGS); Energy Technologies Institute (ETI) (UK); Princeton University Carbon Mitigation Initiative (USA); Columbia University Earth Engineering Centre (USA); Stanford University Global Climate & Energy Project (USA); US Geological Survey (USGS); Lawrence Berkeley National Laboratory (USA); Los Alamos


22


National Laboratory (USA); National Energy Technologies Laboratory (USA); Global CCS Institute (Australia); CO2 Cooperative Research Centre (CO2CRC) (Australia); Australian Commonwealth Scientific and Research Organisation (CSIRO); CO2NET; Carbon Sequestration Leadership Forum (CSLF); Research Institute of Innovative Technology for the Earth (RITE) (Japan); Bureau de Recherches Géologiques et Minières (BRGM) (France); SINTEF (Norway); Canadian Geological Survey and; Bellona (Norway/ Russia).

1.8

Acknowledgements

The financial support of the Centre for Innovation in Carbon Capture and Storage (CICCS, EP/F012098/1) is gratefully acknowledged by the authors.

1.9

References

APGTF (2009) Cleaner Fossil Power Generation in the 21st Century: A Technology Strategy for Carbon Capture and Storage. UK Advanced Power Generation Technology Forum (APGTF), April. Bachu S (2008a) CO2 storage in geological media: role, means, status and barriers to deployment. Progress in Energy and Combustion Science, 34: 254–273. Bachu S (2008b) Legal and regulatory challenges in the implementation of CO2 geological storage: an Alberta and Canadian perspective. International Journal of Greenhouse Gas Control, 2: 259–273. Baines SJ and Worden RH (2004) Geological Storage of Carbon Dioxide. Special Publication 233, Geological Society, London, UK. BERR (2008) Towards Carbon Capture and Storage: A Consultation Document. Department for Business, Enterprise & Regulatory Reform (BERR), London, UK. BERR (2009) Energy Trends. Department of Energy and Climate Change, London, UK, March 2009. Climatico (2010) Copenhagen De-briefing, An analysis of COP15 for long-tem cooperation, available at: http://www.climaticoanalysis.org/post/copenhagen-de-briefing-an-analysisof-cop15-for-long-term-cooperation/ (accessed January 2010). DECC (2009) Climate Change Act 2008: Impact Assessment. Department of Energy and Climate Change (DECC), London, UK. Edwards PP, Kuznetsov VL, Brandon David WIF and Brandon NP (2008) Hydrogen and fuel cells: towards a sustainable energy future. Energy Policy, 36(12): 4356–4362. EC (2008) COM (2008) 18 final, Proposal for a Directive of the European Parliament and of the Council on the geological storage of carbon dioxide and amending Council Directives 85/337/EEC, 96/61/EC, Directives 2000/60/EC, 2001/80/EC, 2004/35/EC, 2006/12/EC and Regulation (EC) No 1013/2006. Brussels, 23.1.2008. EIA (2008) International Energy Outlook 2008. Energy Information Administration (EIA), US Department of Energy (DOE), DOE/EIA-0404(2008), Washington, DC. Etheridge DM, Steele LP, Langenfields RL, Francey RJ, Barnola J-M and Morgan VI (1996) Natural and anthropogenic changes in atmospheric CO2 over the last 1000



23

years from air in Antarctic ice and firn. Journal of Geophysical Research, 101(D2): 4115–4128. Gibbins J and Chalmers H (2008) Carbon capture and storage. Energy Policy, 36: 4317–4322. Gough C and Shackley S (2005) An Integrated Assessment of Carbon Dioxide Capture and Storage in the UK, Technical Report 47. Tyndall Centre for Climate Change Research, Manchester, UK. Halmann MM and Steinberg M (1999) Greenhouse Gas Carbon Dioxide Mitigation: Science and Technology. CRC Press, Boca Raton, FL. Holloway S (2005) Underground sequestration of carbon dioxide – a viable greenhouse gas mitigation option. Energy, 30: 2318–2333. IEA (2007) World Energy Outlook 2007. International Energy Agency, OECD/IEA, Paris, France. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, de Coninck H C, Loos M and Meyer L A (eds), Cambridge University Press, Cambridge, UK. IPCC (2007) Climate Change 2007: Mitigation, Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Metz B, Davidson O R, Bosch P R, Dave R, Meyer L A (eds), Cambridge University Press, Cambridge, UK. IRGC (2008) Policy Brief: Regulation of Carbon Capture and Storage. International Risk Governance Council (IRGC), Geneva, Switzerland. Karl TR and Trenberth KE (2003) Modern global climate change. Science, 302: 1719–1723. Marini L (2007) Geological Sequestration of Carbon Dioxide: Thermodynamics, Kinetics and Reaction Path Modelling. Elsevier, Amsterdam, the Netherlands. Martinot E, Dienst C, Weiliang L and Qimin C (2007) Renewable energy futures: Targets, scenarios and pathways. Annual Review of Environment and Resources, 32: 205–239. Mazzoldi A, Hill T and Colls JJ (2008) CO2 transportation for carbon capture and storage: Sublimation of carbon dioxide from a dry ice bank. International Journal of Greenhouse Gas Control, 2: 210–218. McKinsey & Company (2008) Carbon Capture and Storage: Assessing the Economics. McKinsey & Company, available at: http://www.mckinsey.com/clientservice/ccsi/pdf/ CCS_Assessing_the_Economics.pdf (accessed December 2009). MIT (2008) Carbon capture and storage projects. Massachusetts institute of Technology, Cambridge, MA, Available at; http://sequestration.mit.edu/tools/projects/index.html (accessed December 2009). Orr Jr FM (2009) CO2 capture and storage: are we ready? Energy & Environmental Science, 2: 449–458. Pacala S and Socolow R (2004) Stabilisation wedges: Solving the climate problem for the next 50 years with current technology. Science, 305: 968–972. Palmer TN and Räisänen J (2002) Quantifying the risk of extreme seasonal precipitation events in a changing climate. Nature, 415(6871): 512–514. Shackley S and Gough C (2006) Carbon Capture and its Storage: An Integrated Assessment. Ashgate Publishing, Aldershot, UK. Socolow R, Hotinski R, Greenblatt JB and Pacala S (2004) Solving the climate problem: technologies available to curb CO2 emissions. Environment, 46(10): 8–19.


24


Stern N (2006) The economics of climate change: The Stern review. Cambridge University Press, Cambridge, UK. Stocker TF and Schmittner A (1997) Influence of CO2 emission rates on the stability of the thermohaline circulation. Nature, 388(6645): 862–865. Svensson R, Odenberger M, Johnsson F and Strömberg L (2004) Transportation systems for CO2 – application to carbon capture and storage. Energy Conservation and Management, 45: 2343–2353. van Alphen K, van Voorst tot Voorst Q, Hekkert M and Smits REHM (2007) Societal acceptance of carbon capture and storage technologies. Energy Policy, 35: 4368– 4380. Wanty RB and Seal II RR (2004) Water–Rock Interaction: Proceedings of the Eleventh International Symposium on Water–Rock Interaction, WRI-11, Taylor & Francis, London, UK. WEC (2007) Carbon Capture and Storage: A WEC ‘Interim Balance’. World Energy Council (WEC) Clean Fossil Fuels Systems Committee. Wilson EJ and Gerard D (2007) Carbon Capture and Sequestration: Integrating Technology, Monitoring and Regulation. Wiley-Blackwell, Oxford, UK. WRI (2008) CCS Guidelines: Guidelines for Carbon Dioxide Capture, Transportation and Storage. World Resource Institute (WRI), Washington, DC.


2

Techno-economic analysis and modeling of carbon dioxide (CO2) capture and storage (CCS) technologies

J. O g d e n and N. J o h n s o n, University of California Davis, USA Abstract: This chapter presents an overview of the estimated costs of implementing carbon dioxide (CO2) capture and sequestration (CCS) for fossil energy systems. Cost models for the components of a CCS system are reviewed and developed, including CO2 capture, transport via pipeline and ship, and injection into various reservoirs for storage and/or enhanced fossil fuel recovery. The chapter closes with a discussion of CCS system models and future trends. Key words: carbon capture and sequestration (CCS) economics, CO2 injection, CO2 capture, CO2 transport, CCS system modeling.

2.1

Introduction

Implementing carbon capture and sequestration (CCS) on a large scale will involve major investments. Existing fossil energy conversion plants will be retrofitted for carbon capture or replaced with new plants that integrate carbon capture into the process. A new transport infrastructure will be needed to bring carbon dioxide to suitable storage sites, such as underground geological formations. The storage site will include injection wells, piping, and monitoring equipment. In some cases, a new energy delivery system could be required to bring decarbonized energy carriers such as hydrogen to markets. Understanding the design, performance, and costs throughout the whole system is crucial to evaluating the economic viability of fossil energy with CCS, as compared to fossil energy without CCS, and in comparison to other low-carbon energy supply options such as renewables and nuclear power. In this chapter, we review current state of knowledge on technoeconomic modeling of fossil energy systems with CCS. We develop idealized technical and cost models for the entire system and discuss methods and metrics for economic evaluation. A central question for techno-economic studies is how much it will cost to implement CCS for a particular situation. To assess costs and carbon dioxide (CO2) savings with CCS, we need to compare to a reference fossil energy system without CCS. The modeller also needs to specify the system boundaries for the particular analysis, the technology status (e.g., is the 27 © Woodhead Publishing Limited, 2010

28


system based on current or advanced technology, are costs projected for the nth plant?), economic and feedstock cost assumptions (the engineering/ economic model is embedded in a larger context of energy and policy), and policy constraints such as carbon taxes. These issues are illustrated through a simplified diagram of a ‘generic’ fossil energy system with carbon capture and storage (Fig. 2.1). Most studies consider the total system as three ‘components’, (i) CO2 capture from a source such as a fossil energy plant or industrial complex, (ii) CO 2 transport from the source to a sequestration site, and (iii) the CO2 storage site itself (e.g., an underground deep saline aquifer, a depleted oil field, or a deep unmineable coal formation). We have sketched a fourth ‘component’, the energy demand. Clearly, the characteristics of the demand (the type of energy carriers needed, the scale, location, and time dependence of demand) determine what is required from a fossil energy system with CCS. Moreover, fossil power plants with CCS are part of a larger energy system, which must be considered in economic evaluation of scenarios for CCS deployment. The design of each part of the system depends on a host of factors, some of which are shown in italics. The size and location of the different components of the system depend on regional factors such as the proximity of energy supply to good sequestration sites, and costs depend on local energy prices (for example, the price of natural gas versus coal). System models can be either static or time dependent, generic or geographically specific. Several metrics are commonly used for assessing the economics of fossil energy systems with CCS (IPCC, 2005), relative to a reference case without CCS. First is the capital cost of the system. Second is the levelized cost of energy (for example $/kWh of electricity), which includes the time value of money, and all fixed and variable operating costs. Third is the cost of CO2 in $/tonne of CO2 captured, or $/tCO2 emissions avoided. Various levels of modeling can be used to understand the performance and economics of the whole system. These are loosely grouped as ‘component level engineering/economic models’, ‘system level engineering/economic/ spatial models,’ and ‘macro-economic and integrated assessment models’. Most engineering/economic modeling studies to date focus on one component such as the fossil energy conversion plant, the CO2 transport system or the underground CO2 storage site. Table 2.1 shows the types of inputs and outputs and modeling tools used in component models for different parts of the system. These are described in detail in later sections of this chapter. Some studies combine one or more components into system level engineering/economic/spatial models. Most focus on either the fossil energy complex or the CO2 transport and storage infrastructure (Hendriks, 1994; Ogden, 2003; Herzog, 2006; McCoy, 2008). These models allow sensitivity studies to understand the major costs at the system level. They also enable use of geographic specific data for realistic regional case studies. With


Fossil energy complex Fossil feedstock (coal, NG, petroleum, coke, etc.)

CO2 transport

Distance, geography

Energy demand Electricity and fuel energy demands; end-use technologies; geographic layout and time dependence of demand; fuel storage and delivery system

Distance, geography

CO2 sequestration site Well depth; reservoir characteristics (e.g., permeability, thickness, pressure, temperature, capacity)

Recovered oil or gas in the case of EOR, EGR, or ECBM

29

2.1 Fossil energy system with carbon capture and sequestration. Key variables are shown in italics.

Techno-economic analysis and modeling


Plant design, scale, cost, energy efficiency, and capture efficiency; product mix (e.g., electricity, H2, liquid fuels); pressure, temperature, and purity of CO2

Electricity, H2, liquid fuel transport

30


Table 2.1 Component level engineering/economic models Fossil energy plant w/CCS (Section 2.2) Inputs: Plant type (post-combustion capture, pre-combustion capture, oxyfuel combustion, capture technology type, new or retrofit); feedstock (coal, natural gas, pet coke, biomass mix); plant capacity; technology status (current or advanced; first plant or nth plant); equipment performance and costs; output product mix (electricity, hydrogen, liquid fuels); CO2 capture rate; plant equipment selection (e.g. type of capture system). Model type: Process simulation for steady-state mass and energy flows (ASPEN, HYSYS); specialized models for particular equipment; equipment cost and performance, models based on existing supplier costs and projections for future technology. Outputs: Plant process flow diagram with mass and energy flows; feedstock and energy inputs, product outputs; plant energy conversion efficiency; rate of CO2 capture; overall plant capital and operating cost as a function of scale; capital cost, cost of carbon capture $/tonne CO2, cost of energy $/kWh electricity or $/GJ fuel. CO2 transport (section 2.3) Inputs: Transport mode (pipeline, ship, truck); CO2 flow rate; CO2 pressure, purity, temperature; distance to sequestration site; pipeline costs per meter of length for various diameters. Model type: Pipeline fluid flow models that find CO2 flow rate as a function of pressure (inlet and outlet), pipeline diameter and length. Cost models based on existing pipeline costs and projections for future technology. Outputs: Pipeline diameter, capital cost, booster compressor requirements, capital and operating cost as a function of scale, levelized cost of transport $/tonne CO 2. CO2 storage (section 2.4) Inputs: CO2 flow rate; CO2 pressure, purity, temperature; sequestration site reservoir characteristics (type of reservoir, pressure, permeability, layer thickness, depth; storage capacity); injection well characteristics (depth, pipe diameter). Model type: Model for diffusion of injected CO2 in porous media. For wells: pipeline fluid flow models that find CO2 flow rate as a function of pressure (inlet and outlet), pipeline diameter and length and well depth. Cost models based on existing CO2 injection well costs and projections for future technology. Outputs: Number of wells, well spacing, piping, and other sequestration site equipment; capital and operating costs for wells and injection site. Levelized cost of storage $/tonne CO2.

the increasing availability of geographic information system data, it has become possible to do geographic specific case studies of fossil energy systems with CCS (Herzog, 2006; Johnson and Ogden, 2008; Johnson et al., 2008; Middleton and Bielicki, 2009). Existing CO2 sources and power plant locations, sequestration sites, and potential pipeline locations can be identified for a given region. The model develops an optimized design for CO2 pipelines to connect sources of CO2 and storage sites. Most technoeconomic system models are static, and assume a fixed demand. However, several scenarios for building up regional CCS-based systems over time have been developed (Kuuskraa, 2007; IEA, 2008; Johnson et al., 2008). The costs for making a multi-decade transition to a fossil energy system with CCS can be estimated. © Woodhead Publishing Limited, 2010


31

Finally, models of CCS systems have been incorporated into larger models of the energy economy such as MARKAL or NEMS or into integrated assessment models such as MESSAGE. In these models, various energy supply options compete to supply specified demands for energy services such as lighting, cooling, and vehicle miles travelled. These models yield insight into the potential role of fossil energy systems with CCS under different technology assumptions or policy constraints. The next three sections deal with modeling components, as a basis for discussing integrated system models.

2.2

Carbon dioxide (CO2) capture

There are a variety of options for capturing CO2 from fossil energy plants. These are classified as post-combustion, oxyfuel combustion, and precombustion decarbonization (Fig. 2.2). Post-combustion capture involves separation of CO2 from combustion flue gases and is appropriate for retrofits to existing fossil energy plants. In oxyfuel combustion, the fuel is burned in pure oxygen, allowing ready capture of CO2. Pre-combustion capture involves a gasification or reforming stage to produce a syngas, with subsequent processing to allow capture of a pure stream of CO2. As shown in Fig. 2.3, pre-combustion decarbonization adds flexibility, as a variety of products can be made including electricity and fuels such as hydrogen, syngas, SNG (synthetic natural gas), or methanol. N 2, O 2, H 2O

Post-combustion capture Flue gas Fuel Air

CO2 separation

Power and heat

CO2

Pre-combustion capture

Fuel

CO2

Gasification or partial oxidation shift + CO2 separation

H2 Air

Power and heat N 2, O 2, H 2O

O2 Air

Air separation

CO2 compression, transport, and storage

N2

Oxyfuel combustion capture Fuel

O2 Air

CO2(H2O)

Power and heat

Air separation

Recycle (CO2, H2O) N2

2.2 CO2 capture processes (IPCC, 2005).


32

Developments and innovation in CCS technology Flexibility for alternative products

Natural gas Coal Biomass Waste

CO2 Gasifier, reformer

Gas treatment

CO2 capture

CCGT

Alternatives or additionalities

Electricity Heat H2 Synthesis gas (CO + H2) SNG Methanol Motor fuels

2.3 Pre-combustion capture options (CCGT = combined cycle gas turbine) (IEA, 2008).

2.2.1 Plant level modeling First, we define the system boundaries and the plant configuration. To assess the incremental cost of implementing CO2 capture, we also specify a reference case fossil energy plant without CO2 capture. Then process simulation software such as ASPEN or HYSYS is used to estimate steady-state mass and energy balances for the fossil energy plant. The process model sizes and characterizes all the equipment that is needed in the plant, and energy and material inputs and outputs. This forms the basis for estimating plant capital and operation costs, energy conversion efficiency or heat rate, and plant product mix. Generally, CO2 capture involves both extra capital costs and additional energy use, for example, for CO2 compression. Table 2.2 shows results from a recent review paper by the IEA (2008) that estimates the conversion efficiency for fossil power plants with and without CCS.

2.2.2 Economics of fossil energy plants with CO2 capture The main economic metrics used to assess fossil energy systems with CO2 capture are the plant capital investment cost given in $ or $/kW, the cost of electricity (or electricity and fuels), the avoided cost of carbon in $/tonne CO2, and the capture cost for carbon in $/tonne CO2. Plant capital investment cost Once the plant mass and energy flows have been determined, and plant equipment is sized, the total capital investment can be found. Plant investment cost is typically given in $/kW. Technology maturity is an important consideration in estimating plant capital costs. For current CO2 capture technologies, costs are moderately well


Table 2.2 Characteristics of electric power plants with and without CO2 capture (IEA, 2008) Investment costs ($/kW)

Efficiency (LHV %)

Electricity cost ($/MWh)

Technology Start

CO2 capture rate (%)

With CO2 capture

No CO2 capture

With CO2 capture

No CO2 capture

With CO2 capture

Reference plant (no CO2 capture)

Coal steam cycle Coal steam cycle, oxyfuel Coal IGCC, Selexol1 Biomass IGCC Gas, CC, California Gas CC oxyfuel

85 85 90 90 85 85 85 85 85 95

2250–3200 1850–2500 2500–3100 2100–2600 2300–2800 1800–2400 2600–3000 1000–1200 800–1000 1250–1400

1500–2200 1300–2000 1900–2400 1500–2100 1600–2300 1300–2000 1900–2400 660–750 550–650 700–850

38 44 37 44 35 48 26 49 56 48

47 52 47 52 44 54 34 57 63 58

74–83 59–68 77–87 60–69 76–86 58–65 110–130 59–88 49–75 51–79

39 27–29 41–44 28–31 40–41 26 64–73 33–59 30–53 30–53

2010 2030 2020 2030 2010 2030 2025 2010 2030 2020

CC = combined cycle; IGCC = integrated gasification combined cycle. TM Dow Chemical Company.

1



33

34


known, because many of the processes such as gasification, reforming, and CO2 compression are used commercially. However, for advanced or future CO2 capture technologies, costs are more uncertain and projections must be made. Here it is important to specify whether the advanced technology plant is assumed to be the first of kind or the ‘nth’ plant (assuming the advanced technology has been built many times and has met its goals). Technology learning curves have been employed in some analyses to assess the transition costs for bringing advanced CO2 capture technology from the first plant to maturity (Kuuskraa, 2007; IEA, 2008). The conventions for estimating the plant investment cost are not completely consistent among studies in the literature. Indirect costs for installation, site preparation, engineering, project contingency, and interest during construction, are not always added to the equipment costs. Recent studies by the IPCC (2005) and the IEA (2008) attempted to adjust for these differences to put estimates from the literature on a consistent basis. Table 2.2 illustrates the investment costs for a variety of coal power plant options with and without CO2 capture in the 300–1000 MWe range (IEA, 2008). CO2 capture adds 40–100 % to the plant capital cost. Cost of electricity A second important metric is the cost of electricity from a fossil energy plant with CO2 capture, as compared to the cost of electricity for a similar type of plant without capture. The levelized cost of electricity (COE) is estimated for a power plant as follows: COE ($/kWh) =

(CRF*Ccap + COMf + COMv + Cfeed – CRco ) CO2 * CO2 )

[2.1]

where CRF is the capital recovery factor, Ccap is the total capital) investment cost ($), COMf is the annual fixed operation and*8760) maintenance (O&M) cost ($/ (CF*Cap year), COMv is the annual+variable O&M cost excluding feedstock ($/year), (Tax Cfeed is the annual feedstock cost ($/year) (e.g., coal, natural gas, biomass), CRco is the annual credit for co-products ($/year) (e.g., fuels), CF is the annual average plant capacity factor (%), Cap is the plant capacity (kW), 8760 is the number of hours in one year, TaxCO2 is the carbon tax, if applicable ($/ x E rate of the plant (tCO2/kWh). The tCO2), and ECO2 is the CO2 emissions feedstock cost per kWh of electrical output can be expressed as the product of the plant heat rate (kJ feed/kWh electricity) and the price of feedstock ($/ kJ). The variable O&M costs depend on required energy inputs and the prices of these inputs. The fixed O&M costs are often expressed as a fraction of the total plant capital cost, typically 3–5 % of the capital cost per year.



35

The cost of electricity with and without CO2 capture is estimated in Table 2.2, based on studies by the IEA (2008) for current and future technologies. The efficiency is generally lower for power plants with CO2 capture and capital costs are higher, yielding an electricity cost premium of 2–6 cents/ kWh or 40–100 %. Avoided cost of CO2 emissions To compare carbon capture with other options for reducing carbon emissions, it is useful to estimate the avoided cost of CO2 emissions. For electricity plants, this is defined as the difference in the electricity cost per kWh divided by the difference in CO2 emissions per kWh compared to a reference system without CO2 capture (IPCC, 2005):

LCavoided =

(COEcapture – COEref ) ) (ECO2ref – ECO2c capture

[2.2]

where LCavoided is the levelized cost of avoided CO2 emissions ($/tCO2 avoided), COEcapture is the cost of electricity for the plant with capture ($/ kWh), COEref is the cost of electricity for a reference plant without capture ($/kWh), ECO2ref is the CO2 emissions rate of the reference plant (tCO2/kWh), and ECO2capture is the CO2 emissions rate of the capture plant (tCO2/kWh). Cost of CO2 capture A similar metric is the cost of CO2 capture, also in $/tCO2. This is useful for comparing to the market price of CO2, for example, for enhanced oil recovery:

LCcapture =

(COEcapture – COEref ) CO2 capture

[2.3]

where LCcapture is the levelized cost of captured CO2 emissions ($/tCO2) and CO2capture is the quantity of CO2 captured at the plant (tCO2/kWh). The cost of captured CO2 is always less than the avoided cost of CO2 emissions because the energy required to operate the CO2 capture system adds to the amount of CO2 emitted per kWh. Table 2.3 summarizes the emissions rates and costs for power plants with and without CO2 capture (IPCC, 2005).

2.2.3 Production of hydrogen and other fuels with CO2 capture Although much of the emphasis is on CO2 capture at fossil fuel electric power plants, pre-combustion decarbonization opens the possibility of making


36


Table 2.3 Emissions rates and costs associated with power plants with and without CO2 capture (IPCC, 2005)

CO2 emissions rate (kgCO2/MWh)

Electricity cost ($/MWh)

Technology With CO2 No CO2 With CO2 No CO2 Cost of Cost of capture capture capture capture captured CO2 avoided CO2 ($/tonne CO2) ($/tonne CO2) Coal PC plant 92–145 NGCC 40–66 Coal IGCC 65–152

736–811 62–86 344–379 43–72 682–846 54–79

43–52 31–50 41–61

23–35 33–57 11–32

29–51 37–74 13–37

IGCC = integrated combined cycle; NGCC = natural gas combined cycle; PC = pulverized coal.

hydrogen or other syngas-derived fuels from fossil resources with CO2 capture (Fig. 2.3). CO2 capture adds perhaps 20 % to the cost of producing hydrogen from coal (Ogden, 2003).

2.2.4 Cost of electricity: sensitivity to the value of carbon In the previous sections we presented results for the cost of electricity for fossil power systems with and without CO2 capture. In general, the cost of electricity is higher with CO2 capture, but carbon emissions are reduced by over 80 %. Figure 2.4 shows how the costs of electricity compare at different valuations of carbon (e.g., through a carbon tax). Plants with CCS are preferred for carbon taxes in the range of $40–70/tCO 2.

2.3

Carbon dioxide (CO2) transport

This section describes the economics of transporting CO2 from the point of capture at large-scale industrial point sources to onshore or offshore storage sites. Three options are possible for transporting CO2: tanker trucks (onshore), pipelines (onshore and offshore), and ships (offshore). Tanker trucks have been deemed impractical for transporting the enormous quantities of CO 2 produced by large-scale industrial point sources. Consequently, this section focuses on the economics of pipeline transport for onshore applications and pipeline and ship transport for offshore applications. Onshore pipeline transport is a proven technology with approximately 2400 km of large CO2 pipelines in operation globally (Gale and Davison, 2004; Zhang et al., 2006). The majority of these pipelines are used to supply enhanced oil recovery operations in the USA. Several studies identify pipeline transport as the most economical method for moving large volumes of CO2



37

0.25

Cost of electricity ($/kWh)

0.2

NGCC with CO2 capture

NGCC no capture PC plant no capture

PC plant w/CO2 capture

IGCC plant no capture

IGCC plant w/CO2 capture

0.15

0.1

0.05

0 0

25

50 Carbon tax ($/tonne CO2)

75

100

2.4 Cost of electricity from fossil power plants (c. 2010 technology) for different values of carbon tax.

overland (Svensson et al., 2004; Doctor et al., 2005). Pipeline transport has also been identified as the most economical method for offshore transport of CO2 for distances less than 500–1000 km (Bock et al., 2003; Doctor et al., 2005; Wildenborg et al., 2005). For longer distances, ship transport may be preferable.

2.3.1 Pipeline transport The cost of pipeline transport is affected by the characteristics of both the pipeline route and the pipeline itself. Elements of the route that can impact cost include both the physical geography (e.g., river and road crossings, parks, terrain) and social geography (e.g. population density, regional labor and land costs, local acceptance) (Hendriks et al., 2003). In addition, pipeline characteristics, such as length, diameter, materials, the number of bends, and the need for booster stations, are important determinants of cost. Pipeline design Fluid flow models that use hydraulic equations for turbulent flow (Woodhill Engineering Consultants, 2002; Bock et al., 2003; Hendriks et al., 2003; Ogden et al., 2004; McCoy and Rubin, 2008; Vandeginste and Piessens, 2008) are useful in estimating pipe diameter since they include the parameters relevant to pipeline design (i.e., inlet and outlet pressures, pipeline length,


38


CO2 mass flow rate, pipeline roughness factor, elevation change along the segment, and CO2 density, viscosity, and average temperature) and are thus flexible in their application. There are practical constraints on pipeline operating conditions. To avoid two-phase flow, CO2 should be transported in the supercritical phase, which occurs at a pressure greater than 7.38 MPa (Skovholt, 1993). It is recommended that it is transported at pressures greater than 8.6 MPa where changes in compressibility can be avoided at a range of temperatures used for pipeline operation. Table 2.4 lists common design parameters for CO2 pipeline transport. Since pipelines are not available in continuous diameters, the pipeline diameter used for a particular project is constrained by the available discrete sizes, or nominal pipe size (NPS). The NPS generally corresponds to the external diameter in inches of commercially available pipelines (McCoy and Rubin, 2008). In some cases, it may be economically advantageous to design the pipeline with booster compressor stations every 150–300 km. If CO2 is recompressed at intervals, this allows use of smaller diameter, lower cost pipelines. However, there is a trade-off between lower pipe costs and the added costs of compression. These issues are discussed in Zhang et al. (2006), McCoy (2008), and Bock et al. (2003). Comparison of actual and calculated pipeline diameters The capacity, length, elevation difference, and actual diameter of several existing CO2 pipelines are known (Table 2.5) (Vandeginste and Piessens, 2008). Using this information, it is possible to compare the actual diameters with those calculated from the pipeline models provided by McCoy and Rubin (2008) and Vandeginste and Piessens (2008). Table 2.5 indicates that both models accurately estimate the diameters of the existing CO 2 pipelines.

Table 2.4 Common design parameters for pipeline transport (Bock et al., 2003) Parameter

Value

Units

Inlet pressure Minimum outlet pressure Average CO2 temperature Average CO2 density Average CO2 viscosity Pipeline roughness factor Pipeline capacity factor CO2 purity in pipeline Change in elevation

15.2 10.3 25 884 6.06 ¥ 10–5 4.57 ¥ 10–5 100 100 0

MPa MPa °C kg/m3 N–s/m2 meters % % meters



39

Table 2.5 Comparison between actual and calculated pipeline diameter for existing CO2 pipelines Pipeline Capacity Length Elevation Actual NPS McCoy NPS Vandeginste (Mt/year) (km) difference diameter diameter NPS (m)1 (in) (in) diameter (in) Transpetco Sheep Mountain part 1 Sheep Mountain part 2 Bravo Weyburn

3.4 6.4

193 296

1094 893

12.6 20

12 18

14 18

9.3

360

464

24

24

24

7.4 1.8

351 330

955 46

20 14

20 14

20 14

1

Elevation difference represents a decrease in the elevation between the start and end points of the pipeline. Note: For the calculated values, the flow rate (capacity), length, and elevation change given in the table for each specific pipeline are used. The inputs listed in Table 2.4 are used for the unknown parameters.

Offshore pipeline design Woodhill Engineering Consultants (2002), Dahowski et al. (2005), and Bock et al. (2003) examine offshore CO2 pipeline design and cost and note two significant differences between onshore and offshore pipeline design. First, booster stations are impractical for offshore pipelines so long offshore pipelines may require larger diameters than equivalent onshore pipelines in order to maintain pipeline pressure. Second, in designing offshore pipelines one must consider not only the change in pressure between the inlet and outlet, but also the gravity head gain due to the large decrease in elevation from the shore to the outlet at –2000 to –3000 m. Hence, the allowable pressure drop is much greater for offshore pipelines, which may reduce the diameter required for these pipelines. Pipeline cost For onshore CO2 pipelines, the costs include capital costs (engineering, installation, and materials) and operations and maintenance (O&M) costs (monitoring, inspection, and repair). If booster stations are required, additional capital costs and O&M costs are incurred. The energy cost for operating the booster pump or compressor can be a significant annual expense. Onshore pipeline capital costs We surveyed several studies that estimate onshore CO2 pipeline capital costs, including (Woodhill Engineering Consultants, 2002; Bock et al., 2003; © Woodhead Publishing Limited, 2010

40


Hendriks et al., 2003; Parker, 2004; McCoy and Rubin, 2008). Based on these studies, average pipeline capital cost functions were developed for six discrete pipeline lengths from 50–500 km and several nominal pipe sizes from 4–30 inches. These functions were used to derive Equation 2.4 for calculating pipeline capital cost as a function of pipeline NPS and length:

C = 32.086L–0.033D

[2.4]

where C is the capital cost (2005$/m), L is the pipeline length (km), and D is the nominal pipe size (in). The capital cost per meter shows very little dependence on overall pipeline length L and is approximately linear in D. Capital cost estimates given by Equation 2.4 represent average CO2 pipeline construction costs in the USA. Several project-specific factors can influence the actual cost of construction, including population density, physical terrain, regional location, and river and road crossings. In particular, regional cost variations in the USA can exceed 30 % (McCoy and Rubin, 2008) and construction in urban areas or rocky, marshy, or mountainous terrain can significantly increase the installation cost. Another issue that can affect estimates of construction costs is the observed rapid escalation of the costs of the labor and materials required for building pipelines over the past few years. An equipment cost index (e.g. Marshall and Swift) that accounts for this escalation should be used to adjust costs of older studies. The literature provides a range of fixed O&M costs for onshore pipelines between 0.5 and 4 % of the total capital cost with an average of 2.2 %. The levelized cost of onshore pipeline transport ($/tCO2 transported) is shown in Fig. 2.5 as a function of CO2 flow rate and pipeline length. The cost of CO2 transport scales approximately linearly with the pipeline length and shows strong scale economies with flow rate, varying approximately as the flow rate to the –0.6 power. For a 100 km pipeline, the levelized cost is between $1 and $2.50/tCO2 for plants producing > 5000 tCO2 per day. The levelized cost increases substantially for longer pipelines, but is still between $4 and $10/tCO2 for a 500 km pipeline transporting >5000 tCO2/day. Because the technology associated with onshore pipeline transport is mature, costs are not expected to benefit from technological learning in the future. Offshore pipeline capital costs Several studies estimate offshore pipeline costs for scenarios in which CO2 is injected directly into the ocean at depths exceeding 2000 m (Sarv, 2001; Woodhill Engineering Consultants, 2002; Bock et al., 2003; Doctor et al., 2005). Fig. 2.6 compares the levelized cost of onshore and offshore pipelines for a transport distance of 500 km. Total offshore costs are 2.3–2.6 times larger than equivalent onshore costs.


Techno-economic analysis and modeling 35.00

100 km

41

500 km

Pipeline levelized cost ($/tonne CO2)

30.00 25.00 20.00 15.00 10.00 5.00 0.00 0

5000

10 000 15 000 20 000 CO2 mass flow rate (tonnes/day)

25 000

30 000

2.5 Levelized cost for onshore pipeline transport.

Levelized cost ($/tonne CO2)

40.00

Offshore

35.00

Onshore

30.00 25.00 20.00 15.00 10.00 0.00 0.00 0

5000

10 000 15 000 20 000 25 000 CO2 mass flow rate (tonnes/day)

30 000

35 000

2.6 Comparison of onshore and offshore pipeline costs for a 500 km transport distance.

2.3.2 Ship transport Presently, there are only four ships worldwide that operate specifically for CO2 transport. These ships are small-scale vessels used primarily for distributing food-grade CO2 with capacities between 1000 and 1500 m3 (Aspelund et al., 2006). In order to handle the large quantities of CO2 associated with CCS, much larger vessels would be required. As these vessels do not currently exist, cost estimates are largely based on existing LPG ships that operate


42


under similar conditions to those required for CO2 transport (Aspelund et al., 2006). The processes involved in ship transport include liquefaction, intermediate storage, loading, ship transport, and unloading at either an onshore port or an offshore injection platform. Transport system design Several studies examine the cost of CO2 transport by ship (Sarv, 2001; Bock et al., 2003; Svensson et al., 2004; Doctor et al., 2005; Aspelund et al., 2006). We follow the discussion in Aspelund et al. (2006) which includes a detailed description of each system component and, thus, provides a good template for system design. Prior to ship transport, CO2 is compressed, dried, cleaned of unwanted components, and liquefied. Liquid CO2 is then stored in either caverns or semi-pressurized steel tanks and a loading system transfers CO2 from storage to the ship and includes piping, a marine loading arm, pumps, and an export building. The ship then transports CO2 to an offshore injection platform, where it is unloaded. Offshore unloading of a cryogenic liquid has not been performed for any previous process, but Aspelund et al. (2006) describes a novel technology for this system, which includes a submerged turret loading system, piping, and pumping, expansion, and heating systems. It is assumed that CO2 will be injected in batches while the ship is connected to the platform. Ship transport costs Given the number of liquefiers, storage tanks, ships, and loading and unloading systems, the capital and O&M costs for a particular project can be identified. Cost estimates are listed in Table 2.6. Figure 2.7 shows the calculated levelized cost of ship transport for CO2 mass flow rates from 4000–30 000 tonnes/day (1.5–11 Mt/year) given the reference parameters in Table 2.6. In this case, the levelized cost of ship transport is $22–43/tCO2. Liquefaction accounts for the largest portion of the total transport cost (50–64 %) and the cost does not decline with scale since the liquefier capacity in this case is fixed and the variable cost of electricity is significant. In fact, liquefaction accounts for ~88 % of total energy use in the transport chain and ~93 % of the total cost of energy. Consequently, any reduction in this cost can significantly decrease the cost of transport. Aspelund et al. (2006) discusses several methods for reducing the energy requirement for liquefaction, including increasing the inlet pressure and decreasing the impurities in the feed. Storage, loading, and unloading system costs benefit from economies of scale since these systems are better utilized as the CO2 flow rate increases. Ship costs account for 17–23 % of total costs and exhibit cost spikes when additional ships are required.



43

Table 2.6 Reference parameters for cost analysis (Bock et al., 2003; Svensson et al., 2004; Doctor et al., 2005; Aspelund et al., 2006) Parameter

Value

Offshore distance Electricity price Bunker oil price Natural gas price

500 0.05 309 8

Units km $/kWh $/tonne oil $/GJ

Capital costs 50 Liquifier (1 Mt CO2/year) Steel storage tank (3000 m3) 10 Loading system 50 Semi-refrigerated ship (20 000 m3) 54 Unloading system 60

Million Million Million Million Million

O&M costs Non-fuel O&M for ships 4.7 Non-energy O&M for other components 2 Liquefier energy use 110 Loading energy use 0.2 Ship energy use 0.6 Unloading energy use 0.65 Harbor tax 40 000

% of capital % of capital kWh/t CO2 kWh/t CO2 tonnes oil/hr kWh/t CO2 $

50.00

$ $ $ $ $

Liquefaction

45.00

Storage Loading


40.00

Unloading

35.00

Tanker ships

30.00 25.00 20.00 15.00 10.00 5.00 0.00 4000

6000

8000 10 000 12 000 14 000 16 000 18 000 20 000 25 000 30 000 CO2 mass flow rate (tonnes/day)

2.7 Levelized cost of ship transport.

2.3.3 Comparison of pipeline and ship transport costs for offshore injection When transporting CO2 to offshore injection sites, studies indicate that pipelines are less expensive than ships at distances less than 500–1000 km (Sarv,


44


1999; Bock et al., 2003; Doctor et al., 2005). Based on the cost estimates in this chapter, ship transport becomes more economical than pipelines at distances greater than 750 km (Fig. 2.8). Since pipeline transport benefits from economies of scale more than ship transport, the breakeven distance is expected to shift to longer distances as the CO2 mass flow rate increases. However, since booster compression is difficult with offshore pipelines, the maximum length of these pipelines may be limited. Given that early offshore CCS projects will likely be sited in preferred sites near shore, pipelines are likely to be the preferred mode of transport in the near term unless ship transport costs can be significantly reduced. Ship transport may be favored for long-distance transport (> 500 km) and in cases where the location of injection sites changes and flexibility in routing is required.

2.4

Carbon dioxide (CO2) injection

This section describes the economics of injecting CO2 into geologic reservoirs for long-term storage. CO2 can be injected for ‘storage-only’ into deep saline aquifers or depleted oil and gas reservoirs or it can be used to enhance recovery of valuable fossil fuels, such as oil, natural gas, and coalbed methane. This section begins with an overview of models for injection site infrastructure requirements and then discusses the costs of the various components for each reservoir type.

120.00

Ship transport Offshore pipeline


100.00 80.00 60.00 40.00 20.00 0.00 0

200

400

600

800 1000 1200 1400 Offshore distance (km)

1600

1800

2.8 Comparison of pipeline and ship costs at various offshore distances (8000 tCO2/day).


2000


45

2.4.1 Injection site design To estimate the cost of CO2 injection, it is necessary to calculate the required number of injection wells and the extent of the underground CO2 plume. The number of injection wells can be derived from the maximum injection rate per well, which is dependent on the specific characteristics of a target reservoir (e.g., permeability, porosity, pressure, and thickness). Ahmed (2006) provides a detailed review of the reservoir fluid flow equations that can be used to model CO2 injection rates. These equations address steady-state, pseudo-steady-state, and transient (unsteady-state) flow regimes (Ahmed, 2006). Several studies use the steady-state equation for radial flow of an incompressible fluid to model the maximum injection rate from a single well into a reservoir (Hendriks, 1994; Ogden et al., 2004; Middleton and Bielicki, 2009): qwell =

2p kh(pe – pw ) r m ÈÍln e ˘˙ r Î w˚

[2.5]

where qwell is the injection rate (m3/s), k is the absolute permeability (m2), h is the reservoir thickness (m), re is the radius of influence of the well (m), pe is the pressure at re and is commonly the initial reservoir pressure (Pa), pw is the bottomhole injection pressure (BHIP) (Pa), rw is the wellbore radius (m), and m is the viscosity of CO2 at reservoir pressure and temperature (Pa–s). This steady-state model makes the simplifying assumption that flow is constant throughout the reservoir and that pressure at every location in the reservoir remains constant in time. McCoy (2008) has extended the steady-state model to include the effects of multiple wells and finds that injectivity decreases as the number of wells at a site increases. Consequently, the maximum injection rate of a well is likely to be smaller than predicted by models that don’t consider the effects of neighbouring wells. The maximum extent of the underground CO2 plume over the life of the project is important since it provides the radius of influence (re) used in calculating the injection rate and also identifies the area of influence used to identify site characterization costs. Various studies have developed models for calculating plume radius for different reservoir types (Hendriks, 1994; Bock et al., 2003; Nordbotten et al., 2005; McCoy, 2008).

2.4.2 Carbon dioxide (CO2) injection for storage only Two geologic reservoir types are commonly considered for long-term storage of CO2 without enhanced recovery: deep saline aquifers and depleted gas reservoirs. A typical CO2 injection site includes injection wells and surface


46


equipment (e.g., distribution pipelines and headers). In some cases, compressors or pumps may be required to achieve the required wellhead injection pressure at each well. Costs include well drilling, surface equipment, permitting, and site characterization. Deep saline aquifers Given reservoir and injection parameters, flow equations can be used to determine the maximum plume radius and maximum injection rate for a target aquifer. Table 2.7 provides parameters for two representative saline aquifers and lists the calculated plume radius and injection rate for each aquifer. The injection rate is highly sensitive to changes in certain reservoir parameters like permeability and thickness and can range from tens to thousands of tonnes CO2 per day. Several studies limit the injection rate per well to 2500–3400 t/day (Hendriks et al., 2004; Ogden et al., 2004; McCoy, 2008). Site characterization costs Given the injection rate and plume radius, the number of required injection wells and area of the CO2 plume (or ‘area of review’) can be calculated. Site characterization costs include 3-D seismic imaging of the area of review, drilling of characterization wells, and data processing and modeling services. McCoy (2008) estimates these costs as $38 610/km2 for 3-D imaging, $3 000 000 per well for drilling of characterization wells, and 30 % of the total cost for data processing, modeling, and other services. One well is required for every 65 km2 of the area of review. Equation 2.6 describes

Table 2.7 Common aquifer and injection parameters (Nordbotten et al., 2005; Stauffer et al., 2009) Input parameter

Cold/shallow

Hot/deep

Units

Wellhead injection pressure Reservoir pressure (pe) Maximum BHIP (pw) Reservoir temperature Reservoir CO2 viscosity (m) Reservoir CO2 density (r) Reservoir depth (d) Reservoir thickness (h) Porosity (F) Permeability (k)

10.3 10 15 35 5.77 ¥ 10-5 714 1000 30 0.15 1.97 ¥ 10–14

10.3 30 45 155 3.95 ¥ 10–5 479 3000 30 0.15 1.97 ¥ 10–14

MPa MPa MPa °C Pa–s kg/m3 m m

5.9 5300

km/well tonnes/day/well

Calculated parameter Plume radius (re) @ 20 years 4.8 Max injection rate (qwell) 1850


m2


47

the total site characterization cost (corrected to 2005$) as a function of the area of review:

Csite = 117 344A + 2.70 ¥ 106

[2.6]

where A is the area of review (km2). In developing an aquifer site for CO2 injection, permitting studies must be conducted and land, mineral leases, and permits must be purchased. The estimated average cost is approximately $23 000 per well (Bock et al., 2003; Bank and Kuuskraa, 2006). Well drilling costs Well drilling costs include the costs of drilling and completing wells for CO2 injection. Since limited data is available on the costs of drilling wells specifically for CO2 injection, most studies assume that the costs will be similar to those for drilling onshore oil wells in the USA (Bock et al., 2003; McCoy, 2008; Eccles et al., 2009; Stauffer et al., 2009). Based on 2004 Joint Association Survey (JAS) data, Equation 2.7 describes the well drilling cost as a function of well depth (API, 2007):

Cdrill = (–3.9 ¥ 10–8d3 + 4.00 ¥ 10–4d2 – 0.84d + 903)d

[2.7]

where d is well depth (m) and Cdrill is the drilling cost (2005$/well). For typical well depths of 1–3 km, the cost is $0.4–$2.7 million per well. Surface equipment Several studies assume that surface equipment costs for CO2 injection are similar to the costs of equipping water injection wells for secondary oil recovery (Bock et al., 2003; McCoy, 2008; Eccles et al., 2009). These studies use data published by the Energy Information Administration (EIA, 2007) on lease equipment costs for secondary oil recovery in West Texas. Equations 2.8 and 2.9 were developed from 2005 EIA data using the approach employed by McCoy (2008):

Cequip = 0.0121d2 – 21.745d + 53 818 Cequip,N

ÈCequip, N wells > 20 Í 0.5 =Í Ê 21 ˆ C ÍÎ equip ÁË N wells ˜¯ , N wells ≤ 20

[2.8]

[2.9]

where Cequip is the surface equipment cost (2005$/well). O&M costs include normal daily expenses, surface maintenance, and subsurface maintenance and are derived from EIA (2007) data for secondary oil recovery:


48


COM = 8.76d + 13 267

[2.10]

where COM is the O&M Cost (2005$/well/year). McCoy (2008) reports the monitoring cost to be ~$0.02/tCO2 injected. Assuming a project life of 20 years and capital recovery factor (CRF) of 11.75 %, the levelized cost of CO2 injection can be calculated. Figure 2.9 shows the levelized cost for CO2 injection into a cold/shallow aquifer for several annual CO2 flow rates. The contribution of each cost component to the overall cost is also shown. The most surprising finding is that site characterization contributes ~90 % of the total levelized cost while drilling costs contribute less than 7 %. In the hot/deep case, well drilling costs are significantly larger so well drilling and site characterization costs account for ~30 % and ~60 % of the total cost, respectively. The levelized cost of CO2 injection varies from $1.70 to $2.90/tonne CO2 for the cold/shallow case and from $1.40 to $3.30/tonne for the hot/deep case. In both cases, site characterization costs are significant and consequently methods to reduce these costs could substantially reduce the cost of CO2 injection. Two potential methods are targeting aquifers that generate smaller plumes (e.g., thick reservoirs) and using sites that have been previously characterized.

$3.50

Land, permitting, and mineral cost O&M and monitoring cost

$3.00

Site characterization cost


Injection equipment cost Well drilling cost

$2.50 $2.00 $1.50 $1.00 $0.50 $0.00 1000

4000

8000 12 000 16 000 CO2 mass flow rate (tonnes/day)

20 000

30 000

2.9 Contribution of each cost component to the total levelized cost of CO2 injection into a cold/shallow aquifer.



49

Depleted gas reservoirs Most studies of CO2 injection into depleted gas reservoirs focus on type I reservoirs, where natural gas is produced via pressure depletion (Oldenburg et al., 2001, 2004; Bock et al., 2003; Zhang et al., 2007; Hughes, 2009). These reservoirs are attractive because they are proven long-term geologic traps for gas storage, have already been extensively characterized, and have existing gas production infrastructure that may be converted for CO 2 injection. Unlike saline aquifers where only about 2 % of the pore space is available for CO2 storage, 80–95 % of this space may be available in type I reservoirs. This means that less reservoir volume is required for storing a given amount of CO2. These reservoirs are also favored since they are appropriate for enhanced gas recovery (EGR) (see Section 2.4.3). Although depleted gas reservoirs have several advantages over saline aquifers for CO2 storage, there are also some disadvantages. First, many gas production wells have been drilled into these reservoirs and there is a risk of leakage if the abandoned wells are not capped appropriately. Second, pressures in depleted type I gas reservoirs are typically 2–5 MPa (Oldenburg et al., 2001), which is significantly smaller than the projected bottom hole CO2 injection pressure (15–35 MPa). Hughes (2009) examines techniques for reducing BHIP, but concludes that a high-tech flow control device will be required at the base of the tubing in order to maintain flow. The need for this device could present a serious technical challenge to CO2 injection into depleted gas reservoirs. The equipment required for CO2 storage is almost identical for depleted gas reservoirs and saline aquifers. The exception is the flow control devices that will likely be required for the former. The characteristics of depleted gas reservoirs fall within a large range and thus widely different injection rates are possible (Bock et al., 2003; Oldenburg et al., 2004; Hughes, 2009). In order to capture the impacts of different injection rates on cost, two injection rates are modeled: 150 and 1500 tonnes per day. Based on these rates, the number of injection wells is estimated and storage costs can be derived. Storage costs in depleted gas fields differ from those in saline aquifers since these sites have been previously characterized and developed for gas production. Consequently, the permitting cost is reduced to ~$9000 per well (Bank and Kuuskraa, 2006) and site characterization cost is eliminated. Furthermore, it is assumed that existing gas production wells can be converted to CO2 injection wells at one third of the cost of drilling new wells (Oldenburg et al., 2004). However, new injection equipment must be installed since existing gas distribution equipment may not be sufficient for the quantities and pressures of supercritical CO2. Consequently, injection equipment, O&M, and monitoring costs are consistent with injection into aquifers (see p. 47). The cost of flow control devices is not known as this technology has not yet been developed. © Woodhead Publishing Limited, 2010

50


The levelized costs for the two injection rates (150 and 1500 tonnes/day/ well) are estimated to be ~$1.40/tonne and ~$0.20/tonne CO2, respectively. The smaller injection rate requires additional wells and thus has a much larger cost. It is evident that the cost of CO2 storage in depleted gas reservoirs is substantially lower than storage in saline aquifers because costly site characterization is unnecessary and injection wells cost less since they are converted from existing production wells. If new injection wells are drilled, the levelized cost increases 50–60 % for both injection rates, but is still lower than the cost for storage in saline aquifers.

2.4.3 Carbon dioxide (CO2) injection for enhanced fossil fuel recovery In some cases, CO2 injection can enhance the recovery of valuable fossil fuels and thus generate revenue that can be used to offset the costs of CCS. In this section, the costs associated with enhanced gas recovery (CO2–EGR), enhanced oil recovery (CO2–EOR), and enhanced coalbed methane recovery (CO2–ECBM) are reviewed. Although storage capacities associated with these pathways are relatively small, they represent near-term opportunities for reducing the cost of CCS while advancements in cost-effective capture technologies are developed. Site design For enhanced fossil fuel recovery, the required number of production and injection wells can be estimated using a ‘rule of thumb’ method described in Bock et al. (2003) and given by Equation 2.11.

È(Q /eff )˘ N iwells = Í CO2 ˙ (I : P ) Î qproduct ˚

[2.11]

where Niwells is the number of injection wells, QCO2 is the total design mass flow rate of CO2 (scm CO2/day), eff is the CO2 effectiveness factor [scm CO2/scm methane (CH4) or scm CO2/bbl oil], qproduct is the average gas or oil production rate per well (scm CH4/day/well or bbl oil/day/well), and I:P is the injector to producer well ratio. The CO2 effectiveness factor represents the quantity of CO2 stored for every scm CH4 or barrel of oil produced. Common ranges for the input parameters are listed for each of the reservoir types (Table 2.8). In addition to injection and production wells, all enhanced fossil fuel recovery sites must include surface equipment for distributing CO2 and processing the produced fuel for sale. In some cases, water disposal equipment and/or CO2 separation and recycling equipment are required.



51

Table 2.8 Common ranges for input parameters for Equation 2.11 (Bock et al., 2003; Pashin and McIntyre, 2003; Reeves et al., 2003; Oldenburg et al., 2004; EIA, 2007; Durucan and Shi, 2009; Robertson, 2009) Reservoir type

CO2 effectiveness factor

Average gas or oil production rate

EGR EOR ECBM

1.5–3 1400–140 000 85–227 40–70 1.5–10 1700–14 000+

Injector to producer well ratio 1:1 1:1.1 1:1

Enhanced gas recovery (CO2–EGR) In type I gas reservoirs, gas production declines over time as reservoir pressure is depleted. Pressure depletion recovers 80–90 % of original gas in place (OGIP), leaving about 10–20 %. It is postulated that CO2 injection into these reservoirs could allow for enhanced recovery of the remaining gas through repressurization of the reservoir and displacement of methane by CO2. Consequently, CO2–EGR could potentially generate revenue through sale of natural gas while storing CO2. With CO2–EGR, it is assumed that all of the production equipment is already in place so incremental capital costs include conversion of existing production wells to CO2 injection wells, installation of CO2 injection and distribution equipment, and permitting. Equipment capital costs are derived using the equations for storage in depleted gas reservoirs outlined on p. 49. O&M costs include O&M for production and injection equipment, monitoring, and gas processing, treatment, and compression. The O&M cost for injection equipment is given by equation 2.10 and the production equipment O&M cost is derived from EIA (2007) data for gas production in West Texas:

COMp = 405.81d0.5517

[2.12]

where COMp is the annual O&M cost of gas production (2005$/well) and d is well depth (m). Monitoring is estimated to be $0.02/tCO2 and the cost of natural gas treatment and compression is estimated by Bank and Kuuskraa (2006) to be ~$0.025 per scm of CH4. CO2 separation and recycling equipment may be necessary if substantial CO2 breakthrough occurs at the production wells and can add substantial cost. Incremental gas production associated with CO2–EGR is estimated to be 80–90 % of annual total production (Oldenburg et al., 2001, 2004). The incremental revenue from gas sales is calculated by multiplying the incremental gas production by the wellhead gas price and assuming a royalty tax of 12.5 %. A sensitivity analysis given the baseline parameters in Table 2.9 and wellhead natural gas price of $4/GJ indicates that the levelized storage cost is most sensitive to the CO2 effectiveness factor, average gas production


52


Table 2.9 Reservoir characteristics for a baseline case and at three depths (EIA, 2007; Hughes, 2009) Scenario

Average depth (m)

Average gas production rate (scm/day/well)

CO2 effectiveness factor (scm CO2/scm CH4)

Shallow Medium Deep Baseline

1220 2439 3659 1000

7000 7000 14 000 7000

2.8 1.7 1.4 3

$20.00


$10.00 $0.00 –$10.00 –$20.00 –$30.00 CO2 effectiveness factor Average gas production rate Natural gas price Well depth

–$40.00 –$50.00 –$60.00 –100 %

–50 %

0 % 50 % 100 % 150 % Percentage change in parameter

200 %

250 %

2.10 Sensitivity of levelized cost to various parameters.

rate, and the natural gas price (Fig. 2.10). This figure is generally applicable to all enhanced fossil fuel recovery operations. In particular, high prices for the produced gas or oil improves project economics through increased revenue while a reduction in the CO2 effectiveness factor results in improved economics since more product (i.e., revenue) is produced for each unit of CO2. Reservoirs with high average production rates also have better economics since they generate more revenue per well. The levelized cost of CO2 storage is calculated for three reservoir depths given the parameters in Table 2.9 and a wellhead natural gas price of $4/ GJ. The storage cost ranges from –$8.00 to –$11.00/tCO2, which suggests that revenue generated from methane recovery can offset CO2 storage costs. However, if a CO2 capture and transport cost of $50/tonne is assumed, CO2–EGR would require a $40/tCO2 carbon credit or a wellhead natural gas price >$7/GJ in order to be economic.



53

Enhanced oil recovery (CO2–EOR) Although CO2–flood enhanced oil recovery is a proven technology that is currently used at over 80 sites worldwide (Bock et al., 2003), no CO2–EOR projects currently utilize CO2 from power plants. CO2–EOR is attractive since it can improve the recovery of domestic oil while storing CO2 and producing substantial revenue from oil sales. This technique is employed at sites that have already undergone primary and secondary recovery so most of the infrastructure for oil production and CO2 injection is already in place. Capital costs for converting a site to CO2–EOR include: (i) conversion of water injection and/or production wells to CO2 injection wells; (ii) workover of existing production wells and equipment; (iii) installation of equipment for distributing and injecting CO2; and (iv) installation of CO2 separation, compression, and recycling equipment. The equations for calculating the capital costs of converting existing wells to CO2 injection wells, installing CO2 distribution and injection equipment, and permitting are summarized on pages 47–49. An equation for the cost of upgrading existing production wells and equipment for CO2–EOR is derived from EIA (2007) estimates of additional lease equipment costs for water flooding in West Texas:

Cprod = 91 510ln(d) – 567 823

[2.13]

where d is well depth (m). CO2 separation and recycling equipment costs are determined using the design and costs given in Bock et al. (2003). O&M costs include O&M for production and injection equipment, monitoring, CO2 processing, and fluid lifting. The O&M cost for production and injection equipment is given by Equation 2.10 and the monitoring cost is assumed to be $0.02/tCO2. The CO2 processing O&M cost is estimated to be 1 % of the oil price per Mscf CO2 (e.g., $0.25/Mscf at $25/barrel) and the cost for fluid lifting is assumed to be $0.25/barrel of total liquid production (Advanced Resources International, 2006). Annual revenue from the sale of oil is calculated by multiplying annual oil production by the first purchase oil price and assuming a royalty tax of 12.5 %. The first purchase oil price is the price received at the well and can be derived from the marker oil price (Advanced Resources International, 2006). Based on the annual revenue and costs, the levelized cost of CO 2 storage is determined (Table 2.10). Table 2.10 indicates that the storage cost associated with CO2–EOR is negative even at a relatively low oil price of $20/bbl. For marker oil prices of $35 and $60/bbl, the breakeven cost of CO2 is estimated to be $50/tonne CO2 and $108/tonne CO2, respectively. This is consistent with McCoy & Rubin (2009), which estimates the breakeven cost of CO2 to be $117–195/ tonne CO2 for a marker oil price of $60/bbl. Given that most estimates of CO2 capture and transport costs for power plants are less than $50/tonne CO2, this model suggests that CO2–EOR should be economically viable for © Woodhead Publishing Limited, 2010

54


Table 2.10 Reservoir characteristics and levelized cost for three oil fields (Bock et al., 2003; McCoy and Rubin, 2009) Oil field Average depth (m)

Average oil production rate (bbl/day/well)

CO2 effectiveness Storage cost1 factor (scm ($/tonne CO2) CO2/barrel oil)

Baseline 1220 SACROC, Kelly 2042 Snyder Field Northeast 2499 Purdy Unit

40 35

170 103

–$13.82 –$11.35

41

82

–$5.89

1

Assuming $20/bbl marker oil price, maximum CO2 recycle ratio of 3, and design CO2 mass flow rate of 8000 tonnes/day.

most CO2 sources at oil prices greater than ~$35/bbl. Despite the positive economics of CO2–EOR, the limited locations and low storage capacity of these sites suggest that opportunities for storing power plant CO2 emissions via CO2–EOR will be limited. Enhanced coalbed methane (CO2–ECBM) Deep unmineable coal beds are valued for the significant quantities of methane that they contain. Generally, this methane is extracted by pressure depletion and dewatering of the reservoir. However, CO2 can effectively enhance methane recovery since its higher affinity for coal results in desorption and displacement of adsorbed methane, resulting in additional recovery of 10–20 % of OGIP (Reeves et al., 2004; Robertson, 2009). CO2–ECBM can be employed as a secondary recovery technique at sites that have previously undergone primary recovery or as part of primary recovery at undeveloped sites. Reeves et al. (2004) find that undeveloped sites provide better economics since there is a larger methane resource with which to offset CO2 storage costs. Capital costs for new CO2–ECBM developments include permitting, mineral rights, and land, drilling of CO2 injection and methane production wells, installation of equipment for water disposal, distributing CO2, methane, water, and treating and compressing methane for sale. CO2 recycling equipment may be necessary if substantial CO2 breakthrough occurs at the production wells. However, several studies suggest that CO2 breakthrough is negligible with appropriate well spacing (Reeves, 2001; Robertson, 2009). Bock et al. (2003) and Bank and Kuuskraa (2006) provide per-well costs for permitting, mineral rights, and land costs. The average of these values is approximately $23 000 per well. Drilling costs for both production and injection wells are given by Equation 2.7 and CO2 injection equipment costs are given by Equations 2.8 and 2.9. Methane production equipment costs include gas gathering pipelines, production well equipment, and electric



55

power infrastructure. Gas gathering and electric power equipment costs are estimated by Bank and Kuuskraa (2006) as $58 000/well and $16 000/well, respectively. Production well equipment costs are reported by the EIA (2007) for the four primary coalbed methane sites in the USA (Table 2.11). Water disposal equipment design and costing for shallow underground re-injection are given in Bank and Kuuskraa (2006). O&M costs are divided into four components: well maintenance, monitoring, water disposal, and natural gas treatment, processing, and compression. O&M costs for injection wells are determined using Equation 2.10 and O&M costs for CBM production wells are estimated using EIA (2007) data for CBM projects in each of the four basins. Monitoring is estimated to be $0.02/tonne CO2 and water disposal O&M is estimated by Bank and Kuuskraa (2006) to be $0.10/barrel of water. The cost of gas treatment and processing is estimated by Bank and Kuuskraa (2006) to be $0.025 per scm CH4. Annual revenue from the sale of natural gas is calculated by multiplying annual gas production by the wellhead gas price and assuming a royalty tax of 12.5 %. The levelized cost of CO2 storage for the four primary coal bed methane basins in the USA is estimated using the input parameters from Table 2.11 and a wellhead gas price of $4/GJ. These estimates range from –$12.00/tCO2 in the San Juan Basin to $26.00/tCO2 in the Appalachian Basin. In fact, the only basin in which methane recovery offsets CO2 storage costs is the San Juan basin which has relatively high methane production. In this ‘best case’ basin, the breakeven cost of CO2 for wellhead gas prices of $6/GJ and $10/ GJ are $23/tCO2 and $46/tCO2, respectively. This analysis suggests that the potential for economic CO2–ECBM projects will be limited to sites with very high gas production and, even in these cases, either high wellhead gas prices or significantly reduced CO2 capture costs will be required. Since gas production and CO2 injection rates have a large impact on storage costs, technologies that can improve these rates (e.g., horizontal wells and Table 2.11 Reservoir characteristics of primary coalbed methane basins in USA and baseline case (Pashin and McIntyre, 2003; Reeves et al., 2003; Bank and Kuuskraa, 2006; EIA, 2007; Durucan and Shi, 2009) Coal basin Average Average gas Average water CO2 Production depth production rate production rate effectiveness well (m) (scm/day/well) (bbl/day/well) factor (scm equipment CO2/scm CH4) cost ($/well) Appalachian Black Warrior Powder River San Juan Baseline case

610 610 457 945 600

1700 3000 3000 14 000 7000

3 20 50 100 100

3 2 5 3 3


$26 000 $30 000 $16 000 $66 000 –

56


nitrogen injection) could potentially reduce future costs (Durucan and Shi, 2009). Several studies have examined the injection of flue gas rather than pure CO2 since it would eliminate the need for CO2 capture and nitrogen improves gas production (Reeves et al., 2004; Durucan and Shi, 2009; Robertson, 2009; Zarrouk and Moore, 2009). These studies find improved overall economics, but minimal CO2 storage benefits and higher costs per tCO2 since much smaller quantities of CO2 are injected (Reeves et al., 2004; Robertson, 2009).

2.5

Carbon dioxide (CO2) capture and storage (CCS) system modeling

In this section, we discuss the cost of a complete fossil energy system with CCS, drawing on the individual models described in Sections 2.2–2.4 for CO2 capture, transport, and storage (Table 2.12). We consider power plants in the range of 300–1000 MW, which produce 4000–25 000 tonnes of captured CO2 per day; CO2 transport systems capable of moving this amount of CO2 up to 500 km, and various schemes for CO2 injection and storage in geological formations such as saline aquifers and depleted oil and gas fields, including EGR, EOR, and ECBM. CO2 capture is the largest single cost involved in implementing CCS, accounting for $15–75/tCO2 captured. The transport cost of CO2 shows strong scale economies, decreasing with flow rate, and increasing with distance, with costs of perhaps $1–8/tCO2 for onshore pipelines, and three times that for offshore pipelines. Geological storage costs $0.5–5/tCO2. With enhanced recovery of fossil fuels, the cost of CO2 injection and storage can become negative, depending on the value of the produced oil or gas. Generic cost ranges are interesting because they give an order of magnitude on costs. However, more realistic system models are needed to look at the prospects for CCS for a particular region or situation. With the increasing availability of geographic information system data, it has become possible to do geographic specific case studies of fossil energy systems with CCS (Herzog, 2006; Johnson and Ogden, 2008; Johnson et al., 2008; Middleton and Bielicki, 2009). Existing CO2 sources and power plant locations, sequestration sites, and potential pipeline locations can be identified for a given region. The model develops an optimized design for CO2 pipelines to connect sources of CO2 and storage sites. Most techno-economic system models are static and assume a fixed demand. However, several scenarios for building up regional CCS-based systems over time have been developed and help to estimate the costs for making a multi-decade transition to a fossil energy system with CCS (Kuuskraa, 2007; IEA, 2008; Johnson et al., 2008). An example of an optimization model for a coal-based hydrogen production



57

Table 2.12 Range of costs for each component of a CCS system CCS system component

Capital cost

Levelized cost $/tCO2

CO2 capture Incremental capital cost vs Cost of CO2 capture plant w/o CO2 capture ∑ 300–1000 MWe Power Plant 29–51 Coal PC $750–1000/kWe (+40–50 %) 37–74 Coal IGCC $500–700/kWe (+25–40 %) 13–37 $300–500/kWe (+50 %) NGCC ∑ 300–1000 MWth H2 Plant Coal gasification $–50 to +50 /kWth 2–8 NG reformer $150–200/kWth (+60 %) 30 CO2 transport Onshore pipeline ∑ 100 km 4000 t/d 25 000 t/d ∑ 500 km 4000 t/d 25 000 t/d Offshore pipeline ∑ 500 km 4000 t/d 25 000 t/d Ship ∑ 500 km 4000 t/d 25 000 t/d

Total installed capital cost

2.6 0.8

$130 million ($260 000/km) $260 million ($520 000/km)

12.5 4.0

$335 million ($710 000/km) 36 $680 million ($1 360 000/km) 11

$360 million $970 million

CO2 injection1 ∑ Injection wells 1000 m 3000 m ∑ Deep saline aquifer 4000 t/d 25 000 t/d ∑ Depleted gas reservoir 4000 t/d 25 000 t/d ∑ Enhanced gas recovery 4000 t/d 25 000 t/d ∑ Enhanced oil recovery 4000 t/d 25 000 t/d ∑ Enhanced coalbed methane 4000 t/d 25 000 t/d Monitoring

$28 million ($280 000/km) $55 million ($550 000/km)

43 23

$420 000/well $2 700 000/well $17–20 million $81–102 million

1.8–2 1.4–1.7

$1–7 million $5–44 million

0.2–1.4 0.2–1.4


–11 to –8 –11 to –8


–14 to –6 –14 to –6


–12 to 26 –12 to 26

0.2

1

Injection costs are based on reference cases in Section 2.4, but could be much larger or smaller depending on reservoir characteristics and, in the case of enhanced fossil fuel recovery, oil/gas prices.


58


system with CCS in the state of Ohio is illustrated in Fig. 2.11 (Johnson et al., 2008). Beginning with a GIS database, locations of demand (large polygons), coal plant locations (white/black circles), and rights of way are indentified. From this, a set of possible pathways between demand centers (cities) and coal plants is found (Fig. 2.11a), and mathematical programming techniques are used to select the spatially optimal (lowest cost) supply network (Fig. 2.11b). A build-out scenario is postulated, assuming an introduction rate for hydrogen vehicles. In Figure 2.12, we show the growth of the network as the hydrogen demand rises from 5 % of vehicles to 75 %. A more extensive hydrogen pipeline network and more coal to hydrogen plants are added as

(a)

(b)

2.11 Coal-based hydrogen system optimization with CCS (a) shortest distance pathways between all coal plants and design centres; (b) optimal hydrogen pipeline network (Johnson et al., 2008).

(a) Pipeline

(b) Interstate Coal plant Intercity station

(c) Sequestration site

Demand center

2.12 Build-out scenario for coal-based hydrogen system with CCS (a) 5%; (b) 25%; (c) 75% (Johnson et al., 2008).



59

demand grows. In this study, the hydrogen delivery system (an extensive network of pipelines from the power plant to hydrogen refuelling stations) is found to be a major cost factor. By comparison, the pipeline system for transporting CO2 is relatively small since power plants are located near good sequestration sites. Finally, models of CCS systems have been incorporated into larger models of the energy economy. In these models, various energy supply options compete to supply specified demands for energy services such as lighting, cooling, and transportation. These models yield insight into the potential role of fossil energy systems with CCS under different technology assumptions or policy constraints. A recent study by the IEA (2008) used the MARKAL energy/economics model to examine the potential role of various technologies in reaching low-carbon futures by 2050. Two future scenarios with deep carbon emissions cuts were explored (the ACT Map and BLUE Map Scenarios). They find that coal with CCS plays a major role in meeting a goal of reducing energy-related carbon emissions by 50 % compared to a business as usual scenario. As shown in Fig. 2.13, 14–19 % of the emissions reduction is due to use of CCS in fossil power plants, industry, and fuel production. CO2 capture and sequestration is seen as a critical technology for meeting future climate goals.

2.6

Future trends

Models for fossil energy systems with CCS have been developed at the individual component and system level. There remain many uncertainties in the inputs to such models and in the equations themselves, especially those governing injection and storage in underground reservoirs. In the future, as there is more experience with building and operating CCS systems, the uncertainties in cost and performance of power plants, fuel production systems, and industrial plants with CCS should decline and cost estimation should become more accurate. This is important as CO2 capture is the single largest cost associated with a CCS system. Another major source of uncertainty in current cost models is the need to better understand how CO2 behaves in underground storage reservoirs over a period of time. Characterization of reservoirs to estimate injectivity and storage capacity a priori is an active area of research. Progress in this area will improve the quality of cost models for injection sites. The design of a fossil energy system with CCS is regionally dependent. In recent years, several studies have made use of GIS-based data sets to design CO2 pipeline systems in realistic regional case studies. Similar techniques are used to design and route conventional oil and gas pipelines and are proving useful for understanding the possibilities for locating CO2 pipelines. In addition, they allow development of CCS system designs that


60

BLUE Map 48 Gt CO2 reduction

ACT Map 35 Gt CO2 reduction Nuclear, 6 %


CCS power, 8 %

Power fossil fuel switching and efficiency, 17 % End-use fuel switching, 3 %

Nuclear, 6 % CCS industry and transformation, 9 %

Hydrogen FCVs, 4 %

Power fossil fuel switching and efficiency, 7 % End-use fuel switching, 3 %

CCS power, 10 %

End-use fuel efficiency, 24 % End-use fuel efficiency, 28 % Total renewables, 16 % Electrification, 2 % Electricity enduse efficiency, 16 %

Total renewables, 21 % Electrification, 6 %

Electricity enduse efficiency, 12 %

2.13 Reduction in CO2 emissions for the ACT Map and BLUE Map scenarios relative to the Baseline scenario by technology area, 2050 (FCV = fuel cell vehicle) (IEA, 2008).


CCS industry and transformation, 6 %


61

optimize for low cost and incorporate land use constraints. There has been some innovative work on build-out strategies as well. In the future, such tools will be very valuable to examine different strategies for implementing CCS on a large scale.

2.7

References

Advanced Resources International (2006) Basin oriented strategies for CO2 enhanced oil recovery: Permian Basin, prepared for U.S. Department of Energy. Ahmed T (2006) Reservoir Engineering Handbook. Burlington, MA, Elsevier. API (2007) 2004 Joint Association Survey (JAS) on drilling costs. Washington, DC, American Petroleum Institute. Aspelund A, Mølnvik M J and De Koeijer G (2006) ‘Ship transport of CO2: technical solutions and analysis of costs, energy utilization, exergy efficiency and CO2 emissions.’ Chemical Engineering Research and Design, 84(9): 847–855. Bank G C and Kuuskraa V A (2006) The Economics of Powder River Basin Coalbed Methane Development, Advanced Resources International, Inc., prepared for U.S. Department of Energy. Bock B, Rhudy R, Herzog H J, Klett M, Davison J, De La Torre Ugarte D G and Simbeck D R (2003) Economic Evaluation of CO2 Storage and Sink Enhancement Options. Muscle Shoals, AL, TVA Public Power Institute. Dahowski R T, Dooley J J et al. (2005) Building the Cost Curves for CO2 Storage: North America, report number 2005/3. Cheltenham, UK, IEA Greenhouse Gas R & D Programme. Doctor R, Palmer A, Coleman D, Davison J, Hendriks C, Kaarstad O and Ozaki M (2005) Transport of CO2, in Metz B, Davidson O, de Coninck H, Loos M and Meyer L (eds), IPCC Special Report on Carbon Dioxide Capture and Storage. Cambridge, UK, Cambridge University Press, 173–194. Durucan S and Shi J-Q (2009) ‘Improving the CO2 well injectivity and enhanced coalbed methane production performance in coal seams.’ International Journal of Coal Geology 77(1–2): 214–221. Eccles J K, Pratson L, Newell R G and Jackson R B (2009) ‘Physical and Economic Potential of Geological CO2 Storage in Saline Aquifers.’ Environmental Science & Technology, 43(6): 1962–1969. EIA (2007) Costs and indices for oil and gas field equipment and production operations – 1988 through 2006. Washington, DC, Energy Information Administration – Office of Oil and Gas. Gale J and Davison J (2004) ‘Transmission of CO2 – safety and economic considerations.’ Energy, 29(9–10): 1319–1328. Hendriks C (1994) Carbon Dioxide Removal from Coal-fired Power Plants, PhD Dissertation. Department of Science, Technology, and Society. Utrecht, the Netherlands, Utrecht University. Hendriks C, Wildenborg T, Feron P, Graus W and Brandsma R (2003) EC-Case Carbon Dioxide Sequestration, report number M70066. Utrecht, the Netherlands, Ecofys. Hendriks C, Graus W and van Bergen F (2004) Global Carbon Dioxide Storage Potential and Costs, report number EEP-02001. Utrecht, the Netherlands, Ecofys. Herzog H J (2006) ‘A GIS-based model for CO2 pipeline transport and source-sink matching optimization’. WESTCARB Annual Business Meeting. 8 November, Phoenix, AZ.


62


Hughes D S (2009) ‘Carbon storage in depleted gas fields: key challenges.’ In: Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 3007–3014. IEA (2008) Energy Technology Perspectives 2008: Scenarios & Strategies to 2050. Paris, France, International Energy Agency. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, de Coninck H C, Loos M and Meyer L A (eds). Cambridge, UK, Cambridge University Press. Johnson N and Ogden J (2008) ‘Moving towards a national assessment of coal-based hydrogen infrastructure deployment with carbon capture and sequestration.’ Seventh Annual Conference on Carbon Capture and Sequestration. 5–8 May, Pittsburgh, PA. Johnson N, Yang C and Ogden J (2008) ‘A GIS-based assessment of coal-based hydrogen infrastructure deployment in the state of Ohio.’ International Journal of Hydrogen Energy, 33(20): 5287–5303. Kuuskraa V A (2007) A program to accelerate the deployment of CO2 capture and storage (CCS): rationale, objectives, and costs. Arlington, VA, Pew Center Coal Initiative White Paper Series. McCoy S T (2008) The Economics of CO2 Transport by Pipeline and Storage in Saline Aquifers and Oil Reservoirs. PhD Dissertation, Engineering and Public Policy. Pittsburgh, PA, Carnegie Mellon University. McCoy S T and Rubin E S (2008) ‘An engineering-economic model of pipeline transport of CO2 with application to carbon capture and storage.’ International Journal of Greenhouse Gas Control, 2(2): 219–229. McCoy S T and Rubin E S (2009) ‘The effect of high oil prices on EOR project economics.’ In: Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 4143–4150. Middleton R S and Bielicki J M (2009) ‘A scalable infrastructure model for carbon capture and storage: SimCCS.’ Energy Policy, 37(3): 1052–1060. Nordbotten J M, Celia M A and Bachu S (2005) ‘Injection and storage of CO2 in deep saline aquifers: analytical solution for CO2 plume evolution during injection.’ Transport in Porous Media, 58: 339–360. Ogden J (2003) Modeling infrastructure for a fossil hydrogen energy system with CO2 sequestration, in: Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6. Oxford, UK, Elsevier (Pergamon), Vol. 2, 1069–1074. Ogden J, Yang C, Johnson N, Ni J and Johnson J (2004) Conceptual Design of Optimized Fossil Energy Systems with Capture and Sequestration of Carbon Dioxide, Final report for USDOE Award Number DE-FC26-02NT41623, report no. 41623R04. Oldenburg C M, Pruess K and Benson S M (2001) ‘Process modeling of CO2 injection into natural gas reservoirs for carbon sequestration and enhanced gas recovery.’ Energy & Fuels, 15(2): 293–298. Oldenburg C M, Stevens S H and Benson S M (2004) ‘Economic feasibility of carbon sequestration with enhanced gas recovery (CSEGR).’ Energy, 29(9–10): 1413–1422. Parker N (2004) Using Natural Gas Transmission Pipeline Costs to Estimate Hydrogen Pipeline Costs. Davis, CA, UC Davis Institute of Transportation Studies.



63

Pashin J C and McIntyre M R (2003) ‘Temperature-pressure conditions in coalbed methane reservoirs of the Black Warrior basin: implications for carbon sequestration and enhanced coalbed methane recovery.’ International Journal of Coal Geology, 54(3–4): 167–183. Reeves S R (2001) Geologic sequestration of CO2 in deep, unmineable coalbeds: an integrated research and commercial-scale field demonstration project, SPE 7174. SPE Annual Technical Conference and Exhibition, 30 September – 3 October, New Orleans, LA. Reeves S R, Taillefert A and Pekot L (2003) The Allison Unit CO2–ECBM pilot: a reservoir modeling study. Advanced Resources International, Inc., prepared for U.S. Department of Energy. Reeves S R, Davis D W and Oudinot A Y (2004) A technical and economic sensitivity study of enhanced coalbed methane recovery and carbon sequestration in coal. Advanced Resources International, Inc., prepared for U.S. Department of Energy. Robertson E P (2009) ‘Economic analysis of carbon dioxide sequestration in powder river basin coal.’ International Journal of Coal Geology, 77(1–2): 234–241. Sarv H (1999) Large-scale CO2 transportation and deep ocean sequestration – Phase I final report, McDermott Technology, Inc., prepared for U.S. Department of Energy, DE-AC26-98FT40412. Sarv H (2001) Large-scale CO2 transportation and deep ocean sequestration – Phase II final report, McDermott Technology Inc., prepared for US Department of Energy, DE-AC26-98FT40412. Skovholt O (1993) ‘CO2 transportation system.’ Energy Conversion and Management, 34(9/11): 1095–1103. Stauffer P H, Viswanathan H S, Pawar R J and Guthrie G D (2009) ‘A system model for geologic sequestration of carbon dioxide.’ Environmental Science & Technology, 43(3): 565–570. Svensson R, Odenberger M, Johnsson F and Strömberg L (2004) ‘Transportation systems for CO2–application to carbon capture and storage.’ Energy Conversion and Management, 45(15–16): 2343–2353. Vandeginste V and Piessens K (2008) ‘Pipeline design for a least-cost router application for CO2 transport in the CO2 sequestration cycle.’ International Journal of Greenhouse Gas Control, 2(4): 571–581. Wildenborg T, Holloway S, Hendriks C, Kreft E, Lokhorst A, Brook M, Brandsma R, Egberts P and Larsen M (2005) Building the Cost Curves for CO2 Storage: European Sector, report number 200512. Cheltenham, UK, IEA Greenhouse Gas R&D Programme. Woodhill Engineering Consultants (2002) Transmission of CO2 and energy, report number PH4/6. Cheltenham, UK, IEA Greenhouse Gas R&D Programme. Zarrouk S J and Moore T A (2009) ‘Preliminary reservoir model of enhanced coalbed methane (ECBM) in a subbituminous coal seam, Huntly Coalfield, New Zealand.’ International Journal of Coal Geology, 77(1–2): 153–161. Zhang Y, Oldenburg C M, Finsterle S and Bodvarsson G S (2007) ‘System-level modeling for economic evaluation of geological CO2 storage in gas reservoirs.’ Energy Conversion and Management, 48(6): 1827–1833. Zhang Z X, Wang G X, Massarotto P and Rudolph V (2006) ‘Optimization of pipeline transport for CO2 sequestration.’ Energy Conversion and Management, 47(6): 702–715.


3

Regulatory and social analysis for the legitimation and market formation of carbon dioxide (CO2) capture and storage technologies

H. d e C o n i n c k, M. d e B e s t - W a l d h o b e r and H. G r o e n e n b e r g, Energy research Centre of the Netherlands (ECN), the Netherlands Abstract: Deployment of carbon dioxide (CO2) capture and storage (CCS) depends on regulatory and public perception issues as much as on technological developments. This chapter assesses, from an innovation system perspective, the technological, regulatory and social situation for CCS, and clarifies the current plans for regulation in various countries. Besides estimating the impact prospective regulation is likely to have on CCS implementation, this chapter reviews the experience in the social sciences with public perception of CCS. We conclude that legitimation and market formation of CCS, and to a lesser degree (and in some countries) institutions to influence direction of search and entrepreneurial experimentation have to be further developed for CCS implementation. A global demonstration programme as well as guidelines for public engagement and communication would be instrumental in achieving this. Key words: carbon dioxide capture and storage, CO2, technological innovation systems, regulation, climate policy, innovation.

3.1

Introduction

As climate change is gaining ground on the political agenda and carbon dioxide (CO2) capture and storage (CCS) is emerging as an essential element in any climate mitigation portfolio (IPCC, 2007; IEA, 2008a), the question of making mature CCS technology a reality has become urgent. This chapter assesses, from an innovation system perspective, the prospects for CCS technology to mature quickly. In comparison with other low-carbon energy supply technologies, CCS has a number of distinguishing characteristics relevant to its regulation and implementation. First, CCS is paradoxically both a new and a known technology. It is relatively new because development in a small number of focussed research groups started in the early 1990s, but the technology has only really come on the broader agenda fairly recently, since the publication of the IPCC Special Report on CCS (IPCC, 2005). CCS, however, also starts from a high base. The technology comprises relatively known components: 64 © Woodhead Publishing Limited, 2010

Regulatory and social analysis

65

CO2 separation from flue gases is routinely done in the fossil fuel industry and refineries; CO2 transport by pipeline is a mature technology; and injection of CO2 in geological reservoirs has been done for decades for various purposes, but primarily for enhanced oil recovery (EOR). A full CCS chain can be built from mature components. However, a significant role for CCS is only possible if capture options, in particular, are demonstrated at scale. Hence, CCS is both a mature and a demonstration technology and its journey through the innovation chain does not develop according to the regular sequence of R&D, demonstration, deployment and diffusion. Second, CCS is not carbon-neutral and does not reduce emissions to nearzero levels, like renewable and nuclear energy do. It can be characterised as providing significant but imperfect mitigation. The energy penalty, remaining and additional upstream emissions of post-combustion capture, for instance, add up to an emission reduction of 70–80 % compared to a conventional coal-fired power plant (Viebahn et al., 2007). This has consequences for the long-term viability of CCS, as low stabilisation levels require emission reductions beyond 80 % below 1990 levels after 2050 (IPCC, 2007), and hence for long-term policies for CCS. Finally, CCS is an end-of-pipe technology that is more compatible with the current energy system than other low-carbon electricity supply technologies. Rather than replacing coal-fired power production, CCS reduces its climate impacts. This has two important and countervailing consequences. On the one hand, the lack of the co-benefit of a transition to a fully sustainable, non-fossil energy system and the potential risk of CO2 seepage from reservoirs mobilises some environmental movements against CCS (Greenpeace, 2008), although others take an advocative position on CCS (Stangeland, 2007). On the other, CCS is the only option that provides a future perspective for the fossil fuel industry in a carbon-constrained world, as it allows for the continued use of coal and gas for power production, and it provides opportunities for the oil industry in EOR and as storage reservoir operators. These principal characteristics of CCS have implications for the way it is perceived and deployed by actors, and for the way it evolves. In order to evaluate the evolution of a technology, the concept of a Technological Innovation System (TIS) will be used in this paper. The TIS approach (Hekkert et al., 2007; Bergek et al., 2008) is grounded in the theory of National Innovation Systems, coined by Lundvall (1992) but goes beyond that by providing policymakers and other stakeholders with means to advance a technology by changing specific aspects of the system. Bergek et al. (2008) define a TIS as a ‘socio-technical system focused on the development, diffusion and use of a particular technology (in terms of knowledge, product or both).’ A TIS can be national or international, and its main elements are structural components and functions. Structural components can comprise actors, networks or institutions, both specific for a technology or with a more


66


generic applicability. As structural components are less flexible, attention is often more focussed on improving functions. The functions used in this paper are introduced briefly in Table 3.1. Table 3.1 Functions of a technological innovation system according to Bergek et al. (2008) Function

Description

Knowledge development and diffusion

How well the TIS performs with respect to its knowledge base and how knowledge is diffused and combined. Knowledge includes not only scientific and technical knowledge, but also product, market and logistic knowledge.

Resource mobilisation

The extent to which the TIS is able to mobilise competence/human capital through education in specific scientific and technological fields as well as in entrepreneurship, management and finance, financial capital (seed and venture capital, diversifying firms, etc.), and complementary assets such as complementary products, services, network infrastructure, etc.

Market formation

The phase and shape of the market as well as the institutional stimuli for the technology to proceed. Three market formation phases are distinguished: nursing markets, in which a ‘learning space’ is opened up, in which the TIS can find a place to form; bridging markets, which allow volumes to increase as well as number of actors; and mass markets in terms of volume of the technology roll-out, often several decades after the formation of the initial market.

Influence on the direction The balance of incentives and pressures for firms and of search or guidance organisations to enter the TIS. Indicators include the belief in growth potential, incentives from price factors, regulatory pressure and visions of leading customers. Legitimation

Legitimacy is a matter of social acceptance and compliance with relevant institutions. The technology (and its proponents) needs to be considered appropriate and desirable by relevant actors in order for resources to be mobilised, for demand to form and for actors in the new TIS to acquire political strength.

Entrepreneurial experimentation

The number and variety of experiments taking place in the field of the technology, particularly in the later phases of market formation. Indicators include the number of new entrant firms and different types of applications.

Development of positive externalities

The extent to which the technology can create positive effects in other systems. This includes not only positive economic spill-over effects of the technology itself, but also the opportunities it provides to existing firms to enter new markets.



67

The TIS method has been applied to CCS for the case of Norway earlier (Alphen et al., 2009). Discussing a global TIS for CCS can unravel the structural components and functions, and provide a basis for the choice of international policy intervention. This chapter therefore expands the national TIS analysis to a global level, thereby necessarily losing some of the detail that was present in Alphen et al. (2009). In view of the priorities identified above, the overall question to be answered in this chapter is: How may a global TIS for CCS be enhanced, particularly with regard to market formation and legitimation? The chapter starts with an assessment of the current status of CCS technology and the CCS technology innovation system in Section 3.2. As the outcome of this evaluation points specifically at weaknesses in the market formation and public legitimation, Section 3.3 goes into public perceptions of CCS technology and in Section 3.4, current and prospective regulation are discussed, with a view to shedding light on the market formation and influence on direction of search for CCS. Section 3.5 draws policy conclusions and provides an outlook for CCS.

3.2

Technological maturity and the carbon dioxide (CO2) capture and storage (CCS) innovation system

3.2.1 Technological maturity of CCS Various components of CO2 capture, transport and geological storage are in different stages of technological maturity (see Fig. 3.1). Some capture technologies, particularly in industrial sectors such as gas processing and ammonia, are commercial, but in the power sector, where the largest potential for CCS deployment resides, full-scale demonstrations remain to be built. The main barrier to full-scale demonstrations of CO2 capture in the power sector are the high upfront capital costs of an additional 500–1200 US$/kW (IEA, 2008b), or an additional US$ 250–600 million for a 500 MW power plant. CO2 transport does not pose technological challenges, although only a few CO2 ships currently exist. Current pipelines are mostly in sparsely populated areas, but regulation and risks in urban environments seem manageable (EC, 2008; IEA, 2008b). In the field of geological storage, storage in depleted oil and gas fields and saline formations has been implemented on a few occasions but is not completely understood in all types of reservoirs. EOR using CO2 is a commercial technology for oil recovery, but for permanent CO2 storage has only been demonstrated in one large-scale project; the Weyburn project in Canada. Enhanced CoalBed Methane recovery has not yet seen a demonstration at full scale.


68

Capture

Pre-combustion Oxyfuel combustion

Transport Geological storage


Industrial separation

Pipeline Shipping Depleted oil and gas reservoirs Saline formation Enhanced oil recovery enhanced coalbed methane

Research phase

Demonstration phase

Economically feasible under specific conditions

3.1 Stages of maturity of CCS capture, transport and storage components (Coninck et al., 2009).

Mature market


Post-combustion


69

3.2.2 Economic incentives for CCS Since Fig. 3.1 was first drawn in 2005, and despite the extensive talk on CCS after the IPCC Special Report on CCS was published, the maturity of CCS components has changed little and progress on CCS demonstration has been limited. Although a few new CCS projects at scale have been started in the gas processing sector, lack of incentives, rising commodity prices and the investment risk associated with first-of-a-kind installations have been shown to be important barriers to implementation of power-sector CCS. CCS seems prone to run into the ‘valley of death’ (Murphy and Edwards, 2003) associated with technological demonstration, representing the phase between full public funding when a technology is in the R&D phase, and primarily private financing when a technology is more mature and responds to market signals. The valley of death looms specifically with CCS in the power sector, as the capital costs of that option are highest, and the technical complexity of demonstrating the capture process at full scale is large. There is an ongoing debate on how to finance CCS demonstrations in this sector (ZEP, 2007). Given that full-scale projects should function normally in the power market and should be operated by a power company that can benefit from the learning in the first CCS plant, it can be expected that the private sector should cover most of the investment. However, as CCS demonstrations are not viable because of high costs, low and volatile CO2 prices and technological risks, a funding role for the public sector is also required. A public–private partnership, for instance through a combination of a carbon market price with capital subsidies, or a guaranteed carbon price that would be sufficient to cover the carbon abatement cost of CCS would therefore be most obvious to fund full-scale demonstration of CCS in the power sector (Gibbins and Chalmers, 2008; Groenenberg and Coninck, 2008). For sectors where CO2 capture technology is mature, such as for highconcentration or gas processing CO2 sources, the situation is quite different. As the technological risks are limited, a market signal should suffice for deployment. However, these market signals are often not in place, or other barriers prevent further roll-out of CCS (IEA, 2008b). Norway provides a stable and significant market signal for its offshore industry through a CO2 tax, which has delivered the Sleipner and Snøhvit projects. The European Union provides a market-based incentive for CCS in several sectors via its EU Emissions Trading Scheme (ETS), but prices are currently low and volatile. The United States, Canada and Australia do not currently have structural incentives for CCS. As long as CCS is not allowed in the Clean Development Mechanism, it is unlikely that the developing-country potential for low-cost mitigation in the gas processing, refinery and ammonia sectors will be realised (Zakkour et al., 2008; Bakker et al., 2009).


70


3.2.3 The CCS ‘technological innovation system’ This section uses the TIS methodology to characterise the general situation around CCS, the main differences between countries and where international cooperation is most pronounced. Table 3.2 provides a discussion of the TIS structural elements for CCS with examples from various countries and internationally. Table 3.2 Structural elements of the CCS TIS Structural elements

Description for CCS (examples of elements)

Actors

Private sector: ∑ Oil and gas industry (Shell, Saudi Aramco) ∑ Coal industry (AngloCoal, Rio Tinto) ∑ Power companies (RWE, Duke Energy) ∑ Technology manufacturers (GE, Bosch, ALSTOM, Mitsubishi Heavy Industries) ∑ Engineering companies (Schlumberger) ∑ (Re)insurance companies (SwissRe, AIG) ∑ CCS industry associations (CCSA) Government: ∑ Energy departments ∑ Environment departments ∑ Regulators and mining authorities ∑ International organisations (IEA and its implementing agreements, UNFCCC, IPCC, Global CCS Institute) Civil society ∑ National or international environmental NGOs (Greenpeace, EDF) ∑ Local organisations/organisers of local resistance (Environmental Justice; local political parties) Non-governmental, independent: ∑ Academia: research institutes and universities (engineering and social science)

Networks Institutions

National research programmes (US Regional Sequestration Partnerships, Netherlands CATO) EU research programmes (CO2-Geonet, GeoCapacity, CO2REMOVE and many others) International conferences (GHGT-conferences, Green Power conferences) Carbon Sequestration Leadership Forum IEA GHG R&D Programme networks and meetings (summer schools) UNFCCC Conferences of Parties and other meetings Legal: existing and emerging legislation Policy: EU Emissions Trading Scheme, Norwegian offshore carbon tax Culture/routine: some countries have strong links between the private sector, government and academia (Japan, Australia, the Netherlands)



71

Almost all industrialised countries have an interested government and an active civil society. In the private sector, there is more variation between countries. Factors include whether the country has strong manufacturing industry, has fossil fuel resources, a power sector based on fossil fuels, or an insurance industry. Networks are both national and international. The CCS networks show strong links between academia and industry in most countries as well as internationally. Institutions differ more between countries. Both policy and legislation are in various stages of development, as the urgency of climate change and hence the need for CCS are not similarly acknowledged. There is little exchange between governments on legislative issues. In some countries, there is a strong innovation system: industry, research and government cooperate in a relatively harmonised way and there is much mutual trust and cooperation. In other countries, this institution is less strongly developed, but this is compensated by stronger entrepreneurial experimentation in the private and finance sectors. Table 3.3 describes how the functions of the CCS TIS can currently be characterised. The function knowledge development and diffusion related to CCS receives attention from industrial, scientific and public funding parties, as indicated for instance by the increase in R&D budgets, although diffusion of knowledge is in its infancy. The European Commission had good reasons to make any funding to CCS demonstrations from its Economic Recovery Plan (Council of the EU, 2009) contingent on knowledge sharing by the beneficiary. The need for information sharing is also one of the objectives of the Global CCS Institute (GCCSI, 2009), recently founded to advance CCS globally. Resource mobilisation is crucial for advancing any technology, both in earlier phases of development as well as when the technology is diffused widely. For research, the EU and more recently the USA make available financial resources to a selection of CCS demonstrations, but financial institutions are only moderately interested in CCS. The human capacity of CCS is increasing rapidly but most likely insufficient to enable full rollout. Market formation in the bridging market that CCS currently is could be further improved. The institutional embedding of CCS is most progressed in the EU; in most other developed countries, legal systems are still under development, whereas in developing countries work has been very limited in this area. Guidance on the direction of search for further CCS development so far is limited to the targeted number of CCS demonstrations that would need to be realised. Although there is a great belief in the potential of the technology, and the G8 supported the ambition to have 20 large-scale demonstrations by 2010, with the intention to have broad deployment by 2020 (G8, 2008), structural incentives are still largely absent or insufficient. Although CCS has been included in emissions trading schemes in the EU


72


Table 3.3 Functions in the CCS TIS Function

Description

Knowledge development ∑ and diffusion ∑ ∑

Resource mobilisation

Market formation

In most industrialised countries, R&D capacity is relatively good and growing rapidly. Publications on CCS are plentiful and growing. For existing and new technologies, patents are being filed, but there is no assessment of how many.

∑ The human capacity on CCS is rising, both in quality and in quantity. Existing training for geologists or engineers are applicable to CCS. Additional specific training is provided on some occasions. ∑ Mobilising of financial resources from the financial sector has been modest, although governments (EU, Australia, Canada, USA) have raised significant amounts in their economic stimulus packages and through other means. Also equity has been invested into CCS. ∑ The market stage for CCS can be characterised as a ‘bridging market’ – it is beyond the ‘nursing market’ stage, the number of players growing rapidly but it is too soon to speak of a ‘mass market’. ∑ Institutional stimuli are being put into place in the EU and other countries considering strong climate policy while planning new coal-fired power plants or maintaining CO2-intensive industries. ∑ Potential market size in the longer term (IEA, 2008a) is significant.

Influence on the direction ∑ CCS is clearly still awaiting regulatory pressure. In of search or guidance the EU and Norway this is probably strongest with the EU ETS and the offshore carbon tax. ∑ There appears to be a strong belief in growth of the technology, but no clear direction of search in the different capture options. ∑ Companies are interested in CCS and are investing in research. Legitimation

There is a clear split between actors in legitimation of CCS. CCS is considered appropriate and desirable by most private sector actors, academia and by developed country governments, but much less so by civil society and many developing country governments (Hansson and Bryngelsson, 2009; Shackley et al., 2009).

Entrepreneurial ∑ Most entrepreneurial experimentation takes place experimentation within existing oil/gas and power companies. Some new companies have entered the market, but they face the incumbents which are often more resourceful and have easier access to e.g. underground storage reservoirs. ∑ Between capture technologies, there are investors



73

Table 3.3 Continued Function

Description in post-combustion as well as pre-combustion and oxyfuel combustion. ∑ There is no company (except for the gas industry) that controls the whole CCS chain, so for CCS in the power sector, power companies need to collaborate with companies active in the underground. The institutions for that are still largely absent.

Development of positive externalities

With a few exceptions of enhanced hydrocarbon recovery and conditional security of energy supply benefits, development of positive externalities for CCS is limited.

and the emerging system in the USA, the CO2 market prices are as yet too low to stimulate widespread diffusion of the technology. Legitimation is crucial, referring to acceptance of a technology by both policy makers and society at large. While understanding of the need for CCS technology among policymakers is on the rise, the larger public appears reluctant to accept the technology, and strategies for effectively involving the public in decision-making on CCS remain to be applied. Finally, entrepreneurial experimentation has increased steeply since CCS became an option, judged by the number of proposals for CCS pilots and demonstrations (CSLF, 2009). This points at a strong will from industrial parties to pioneer in this field, as well as a willingness to take (limited) risks, although lack of financial support, public support and even technical information have led to the cancellation of a number of planned CCS demonstrations (Hansson and Bryngelsson, 2009). The preliminary and rough analysis of the CCS Technological Innovation System does not reach the depth that country TIS studies, such as Alphen et al. (2009), have. However, it does confirm Hansson and Bryngelsson’s (2009) findings that there is an apparent discrepancy between the optimism of many CCS actors and the performance of the Innovation System functions. In the field of human capacity and belief in growth of the technology, CCS seems well developed. Possible problems arise with legitimation, direction of search and market formation, and to a lesser degree (and in some countries) entrepreneurial experimentation and resource mobilisation. We identify two areas where the TIS of CCS requires most work: legitimation with respect to public perception and participation, the topic of the next section, and in market incentives for CCS, the topic of Section 3.4.


74

3.3


Legitimation: results and gaps in social scientific research regarding public perception and participation

Bergek et al. (2008) define legitimation as ‘a matter of social acceptance and compliance with relevant institutions’. There is no need to look far for examples of how lack of legitimation can lead to technology failure, on both a local and a more global scale. An early experiment with ocean storage was cancelled in Hawaii after a number of local organisations formed a coalition and resisted the location of a small CO2 injection facility in the ocean (De Figueiredo, 2003). In Barendrecht, a town near Rotterdam in the Netherlands, local resistance to CO2 storage in two gas fields grew over the course of 2009. Although the national government has decided in favour of the project, local resistance is likely to at least delay the project substantially. On a global scale, broad resistance to nuclear energy has resulted in far less implementation of the technology than was initially projected. CCS, as a large-scale technology with local risks, which is associated with ‘dirty coal’ and which does not have clearly discernible co-benefits, risks lack of public legitimation. This section is limited to public opinion and will not discuss legitimation from a policymaker or industry viewpoint. As suggested in previous sections, policymakers, albeit slowly in some cases, are beginning to understand the relevance of CCS technology for the global climate. Aspects related to market formation are discussed in the next section. Industry is becoming increasingly appreciative as the viable business case that CCS may at some point offer. Although agenda-setting by industry and policy is certainly important, this section will provide a detailed discussion focusing on the more contentious issue of perception and involvement by the public at large. What is the current status of public opinion and of public involvement? Would the public agree that the choice of CCS for CO2 emissions is a legitimate one? Several issues arise when trying to answer these questions, and more than one of these issues is associated with the status of the scientific research on these topics. Another issue is how to define the status of public opinion. From the perspective of development of the CCS innovation system, public opinion could be defined as ‘good’ when the public is generally positive about CCS. However, from a societal perspective, one could argue that public opinion should be of high quality, being stable, well-informed and considered. This is related to the different possible goals of public participation, one of which can be democratic–idealistic, aiming at giving people the opportunity to develop well-informed opinions on which they can base their decisions. Another goal is purely instrumental, as a way to either gain complete agreement about the



75

need for CCS or at least to avoid possible protest against implementation of CCS technology, specifically around demonstration sites. With these perspectives in mind, we will discuss the issue of the current status of research on public opinion regarding CCS. This includes how knowledge on public opinion is generated and how it should be interpreted. Based on this, we will discuss which questions remain. We will then go into the effects of these gaps in current knowledge about public opinion and public involvement for the estimation of current legitimisation of CCS technology in the context of the TIS.

3.3.1 Current status of research on public opinion: issues encountered In recent years, empirical research regarding public attitudes towards CCS has been developed in several countries. Except for an early study in the Netherlands in 1998 that incorporated the possibility of CCS technology among other options for energy, the earliest studies focusing on public awareness or opinion of CCS technology emerged around 2004. Faced with the question of how the public might react to CCS technology, these studies tried to investigate if the main goal of CCS, CO2 emission reduction, was in fact an issue for the public, if the public was aware of the possibility of CCS, and how this possibility would be evaluated in light of further information. Awareness of CCS technology There have been two studies which tried to place CCS in a broader context by investigating if the public considers CO2 emissions as a problem to be solved. Giving people several societal issues to rank, Curry et al. (2007) reported that protecting the environment was the eleventh highest ranked priority for Americans, while Palmgren et al. (2004) found that respondents ranked ‘reducing climate change’ as the lowest social priority of the 15 choices offered to them. Moreover, most of the studies done so far have encountered a profound lack of awareness. For instance, Reiner et al. (2007) compared awareness of CCS in large adult samples in the USA, the UK, Sweden and Japan. They found low awareness in all four countries, ranging from 22 % of respondents confirming they had heard or read about CCS in Japan, to as little as 4 % of respondents confirming this in the USA. A more recent study in Japan (Itaoka et al., 2009) showed similar results, with 7–18 % of respondents stating that they knew of CCS to some extent. In a survey of 900 respondents in Australia, 29.9 % of respondents were able to give meaningful answers when asked what they understood about CCS (Ashworth et al., 2006). A less representative but more recent Australian study confirmed those average


76


levels (Ashworth et al., 2009). Sharp et al. (2007) also found low awareness of CCS in a survey among 1972 respondents in Canada. In a recent study in France, 6 % of a representative sample was able to define CCS (Ha Duong et al., 2009). Four surveys in the Netherlands conducted in 2004, 2005, 2007 and 2008 (Best-Waldhober et al., 2006, 2008; Lambrichs, 2008) showed similar results. Depending on the kind of CCS technology that the three samples, totalling 918 respondents, were asked about, between 51.2 % and 91.4 % of respondents stated themselves to be unaware of the technology. General public awareness did not increase over time which, specifically for the Dutch situation, is unexpected given that media analyses show a significant increase in media attention devoted to CCS in the Netherlands (Alphen et al., 2007). A scientific field in development: research issues This general lack of awareness leads to one of the major issues in this line of research: How to study public opinion if most of the public is unaware of the possibility of CCS technology? Most researchers faced with these questions have chosen to provide respondents with information before asking their opinion. However, as Malone et al. (2009) state, ‘because of the inherent difficulty of providing information in an unbiased way, surveys may be compromised at the outset if they seek to educate’. Providing respondents with elaborate, understandable, recent, accurate and balanced information on CCS technologies as well as their consequences and context is difficult and highly time- (and resource-)consuming. Most studies can only partially overcome these issues, and are thereby unable to rule out susceptibility to bias. Therefore the opinions found in these studies should be interpreted with care. Related to the question of how to study public opinion is the choice of what method to use. Most of the studies use one or a combination of three methods; a written or digital survey method, focus or discussion groups and experiments. In surveys, experimental surveys and experiments respondents are often given written information to read. In focus or discussion groups, participants are often informed by either a researcher, experts or handouts, or a combination thereof. The conclusions that can be drawn from these studies depend on the kind and quality of the information as well as the method used. Methods using some kind of discussion group often have the advantage of giving insight into the perceptions lay people have of CCS technology, which might be very different from the perceptions experts have. The downside is that these kinds of methods are expensive and time-consuming, which means that often only a few small discussion or focus groups are used, leading to conclusions that cannot be generalised to the population. Surveys with representative sample do not have this disadvantage. However, surveys



77

often use a restricted set of questions leading to a restricted set of possible answers. This restriction might lead to missing certain issues that experts or the researchers had not thought of, but that can be important to the public.

3.3.2 Main arguments and misconceptions Proof of this last argument is given by several studies. For instance, in the USA, observations of multi-state focus group interviews in local communities also showed that important factors for the opinion of CCS were past experience with government, existing low socioeconomic status, and/or desire for compensation. Specific benefits of CCS to the community (i.e. not national or global benefits) were more important than the concern about the risks of the technology itself (Bradbury et al., 2009). The work of Best-Waldhober and Daamen (Best-Waldhober et al., 2006, 2008, 2009; Daamen et al.,, 2007) shows that lay people base their opinion of CCS technologies only in part on the consequences of these technologies that experts deem important. Even though on average respondents’ opinions of CCS technologies are much more consistent with their evaluation of consequences after information on these consequences than without information, still part of their opinion is based on other arguments. A recent study in Switzerland (Wallquist et al., 2009) investigating the concepts which lay people had of CO2 and CCS showed that some people were worried that CO2 might cause cancer, or even that CO2 leaking from storage might cause DNA changes. These kinds of misconceptions are not likely to be anticipated by experts and are therefore less likely to be investigated in polls or surveys, or addressed in information about CCS technology. Studies that have tried to generate insight into the arguments the public deems important for decision-making about CCS technology show somewhat similar mixed results. Two Japanese studies, for instance, using survey questionnaires administered to large samples not only found positive effects resulting from information provision but also analysed what these effects where based on. Itaoka and colleagues (2005) found that the more information respondents obtained about CCS, the more likely they were to support storage options, except for the onshore option of geological storage. More elaborate analysis of these data (Itaoka et al., 2007) revealed that effectiveness of CCS (i.e. its effect on CO2 emission reduction) was the most influential factor in public acceptance of CCS. Tokushige et al. (2007) found that part of the information in their study, which was either a fact sheet with an overview of natural analogues or a fact sheet on field demonstrations of the geological storage of CO2, was effective in decreasing the risk perception and increasing public acceptance. The most important factor for public acceptance was the perception of benefit however, which was influenced by neither kind of information.


78


Recent Dutch studies also show a significant influence of perceived benefits of CCS (Best-Waldhober et al., 2008; 2009). These studies both combine a large sample with elaborate information from experts on aspects and consequences of CCS options. Although part of the sample of the second study became on average more positive after information, further analyses show that there is little to no relation between respondents’ opinion before information and after information. This means that a large portion of the respondents changed their opinion, becoming either more negative or more positive. Another factor influencing opinions of CCS that became apparent from this study was that, in the opinion of the respondents, CCS compares slightly unfavourably with other climate mitigation options such as efficiency, wind energy, nuclear energy or energy from biomass.

3.3.3 Knowledge gaps regarding public opinion and involvement The size of the CCS knowledge gap with the general public can be seen as an indicator for the current state of the innovation system function of public legitimation. Due to the recent start of research on this topic, the limited financial resources available and the politically sensitive nature of the topic, several gaps remain in the knowledge about public opinion and involvement in CCS technology. Gaps that we will discuss here are the development of public awareness and opinion over time, the lack of knowledge about lay arguments and concepts, the lack of knowledge of public information needs, effects of both location and local context, need for effective strategies for public involvement/participation, and lessons from case studies of (planned) demonstration projects for CCS technology. First, as Malone et al. (2009) point out, ‘surveys and polls measure public opinion at one point in time. Opinions are dynamic’. It would be informative to know how public awareness and public opinion evolve over time and which factors, such as media coverage, influence this process. Several studies acknowledge this, and in some countries measurements of awareness have been repeated more than once. Given that the first studies started only a few years ago, longitudinal studies are still at an early stage. Second, as argued before in this section, there is still much to learn about the arguments that the public deems important for making decisions about CCS technology, and the possible misconceptions people might have. Several studies, including some mentioned above, have investigated possible considerations of lay people regarding CCS technology, but the results are somewhat mixed and many questions remain. Are considerations from the general public the same as those from people living close to a possible storage site? Are these considerations influenced by culture, by context, by demographics? Is public understanding and acknowledgement of anthropogenic



79

global warming a driver or a prerequisite for a positive perception of CCS technology? Being able to answer these questions could provide insight into the information needs of the public. How to inform the public is more than a scientific question though. As stated earlier, the issue of informing the public can be viewed from an instrumental point of view or from an ethical point of view. We would propose that people should have the right and opportunity to participate in the debate and the decision-making about energy technology, specifically if local activities are an option. Knowing what kind of information is needed by the public for this is crucial, as is securing understandable, balanced and accurate information from multiple sources. However, even if one’s view is instrumental, the literature shows that what can be seen as the most ethical road can also be the most effective. Research by Mors (2009) and Terwel (2009) shows that effective communication about CCS relies on trust. People put more trust in environmental NGOs than in the industrial organisations involved or the government. Communication strategies by these stakeholders can be counterproductive because, if the source of the information is not trusted, the information is rejected, which may result in negative public attitudes towards CCS. Of course, information that is perfectly understood and trusted by the public does not guarantee a positive opinion of CCS technology. Balanced information will include both advantages and disadvantages of CCS technology, and the evaluation of these characteristics and consequences will differ between individuals. The investigation of effective public participation strategies and lessons from case studies are often linked together. This can be risky if the same organisation that is responsible for the design and/or implementation of the participation efforts is also the one evaluating the process and its effectiveness. A bigger issue for defining the status of legitimation regarding public involvement, however, is the early stage of the development of the technology and, consequently, the few demonstration projects currently planned. Although some studies have investigated participation strategies for CCS technology, these studies concerned very early stages of possibly planned demonstration projects (e.g., Raven et al., 2009). However, most of this kind of research is ongoing.1 Analyses that have been done are either not yet published or concern the early stages of demonstration projects. At the same time, attention devoted to CCS technology in the popular media seems to be on the rise, specifically articles about public protest against CO2 storage in several locations around the world1. Because these protests as well as the media attention are very recent, no scientific analyses have yet been published, making it hard to evaluate the current status of legitimation 1

Based on personal communication within project meetings and the CCS social scientists network.


80


regarding public participation and opinion. Media attention might just reflect the first reactions of a small percentage of people involved. Despite the inauguration of several studies, public legitimation, as a function in the CCS TIS, is still a great unknown. However, early studies as well as observations suggest that in specific project situations and depending on the socio-economic and environmental circumstances, public legitimation may be a cause of concern. Although it is clear that this issue should be more adequately addressed, the instrumental method of ‘convincing the public of CCS’s merits’ is unlikely to work. Rather, resources for real public engagement and, if appropriate, compensation may need to be mobilised.

3.4

Market formation and direction of search: an enabling regulatory framework for carbon dioxide (CO2) capture and storage (CCS) in the EU

Market formation and direction of search have various aspects. First, they comprise the institutions that ensure that the technology can be implemented safely. Second, the bridging market that CCS currently is should be nurtured towards becoming a mass market through incentive mechanisms. In the case of CCS, this means a successfully implemented demonstration phase, rapid roll-out and structural incentives for diffusion in the longer term. Following these threads, this section will subsequently review the enabling legal framework for CCS safety in the EU and internationally, CCS demonstrations, and regulations that are relevant for CCS market formation. As the countries of the EU are furthest ahead both on providing incentives and on the legal system of CCS, this section will focus on the EU, with extra attention devoted to Norway. In some places, developments in Australia and the USA will also be discussed.

3.4.1 Regulation of safety of CCS In recent years, significant progress has been made in regulating CCS. A number of international treaties have been amended to enable CCS, and frameworks for regulating the safety of CO2 capture and storage have been developed in various regions, including Australia and the USA, and adopted in the EU. The key objective of any of these regional frameworks is to guarantee safe storage of CO2, with a view to enabling the introduction of the technology. Consequently, the various regulatory frameworks resemble each other on essential aspects such as site selection and monitoring. All focus on the procedures that operators and competent authorities must follow to ensure operational safety, in line with recommendations and expertise obtained in



81

operations to date (see e.g. Groenenberg et al., 2008). Australia has in place a set of guidelines for monitoring CCS operations, which should support the regions in developing their own regulatory frameworks. In the USA, a draft proposal was tabled by the EPA. In the remainder of this section, we will focus on developments in international and EU regulations, including Norway, as those countries are furthest ahead in CCS market formation and direction of search, and thus provide the most insightful results. International frameworks The EU and Norway are signatories to three international treaties relevant to CCS: the London Protocol, the OSPAR Convention and the Espoo Convention. To some degree, therefore, one can speak of an international market formation effort. The London Protocol (LP) is part of the Convention on the Prevention of Marine Pollution by Dumping of Wastes and Other Matter, also referred to as the London Convention. It was amended in 2006 to allow for the sub-seabed storage of CO2 as of 10 February 2007. The amendments are relevant for offshore applications of CCS. They regulate the storage of CO2 streams from CO2 capture processes in subseabed geological formations, for permanent isolation, thereby creating a basis in international environmental law to regulate this practice and forming the institutions needed for CCS market formation. Storage is only allowed, however, if associated substances from the source material and from capture/ storage processes are incidental. The amendment of the LP triggered similar work on the OSPAR Convention, which is the North East Atlantic equivalent of the LP. The impacts of large-scale CO2 geological storage in sites that are either cross-boundary, or close to national boundaries, should be the subject of discussion among concerned Parties of the Espoo Convention of 1991. Transboundary CO2 transport by pipeline would obviously need to be subject to discussion among the national authorities involved as well. The EU Directive for the Geological Storage of CO2 In its Energy and Climate Package, agreed in December 2008, the EU tried to make a strong move towards providing institutional embedding for CCS. The safety of CO2 capture and storage was regulated in the Directive for the Geological Storage of CO2. This Storage Directive covers injection and storage of CO2 in geological formations for large operations. R&D pilots smaller than 100 kt CO2 per year are excluded, as are ocean storage and enhanced hydrocarbon recovery operations. The Directive focuses primarily on the safety of injection and storage for human health and the local environment. Risks for the global climate are covered by the EU ETS and applicable monitoring and reporting guidelines. The requirements in the Directive regarding safety


82


of injection and storage operations reveal the level of detail that is needed for the legal arrangements of onshore CO2 storage, and that remain to be overcome in other jurisdictions: ∑

Site selection and permit application. The Storage Directive assumes that the objective of geological storage is permanent containment of CO2. A good site selection is crucial, and a site may be selected only if there is no significant risk of leakage and if no significant environmental or health impacts are expected to occur. Site selection should comprise data collection, the construction of a static model representing the geological framework and dynamic flow modelling of the CO2 in the reservoir to support a risk assessment of leakage. This includes a description of the hazards, a quantification of the impacts of leakage for man and environment and an assessment of the risks. ∑ CO2 stream composition. Following the OSPAR CO2 Guidelines, the Storage Directive acknowledges that other substances may be present in the CO2 stream from the CCS processes, but it should consist overwhelmingly of CO2 and levels of the other substances should not affect the storage or transport integrity. No wastes may be added for disposal. ∑ Monitoring and inspection. Monitoring is essential to assess whether the injected CO2 behaves as expected and to detect any leakage that might occur. The results of monitoring are to be compared with projections from the dynamic flow modelling to assess whether site performance is satisfactory. A monitoring plan has to be submitted together with the permit application. Should any leakage occur, corrective measures must be taken. During injection of the CO2, reporting and inspections must take place at least once a year. ∑ Post-closure. After closure, the site remains the responsibility of the operator until legal obligations are transferred to the Member State’s competent authority. This may be done only when all available evidence indicates that the CO2 will be completely contained for the indefinite future. In addition, a minimum period of 20 years applies in principle. After the transfer of responsibility, monitoring may be reduced to a level which still allows for identification of leakages or significant irregularities, but should be intensified if leakages or significant irregularities are identified. Financial provisions are to be made by the operator prior to injection to ensure that all closure and post-closure obligations under the Storage Directive and the ETS directive can be met. The provision should at least be able to cover the costs of monitoring for a period of 30 years. EU Member States have two years to implement the Storage Directive in their national legislation. Whilst much of it may be mere transposition of



83

stipulations in the Directive, there are still a number of issues that each of the Member States may resolve differently. Issues include, for instance, the exact allowable composition of CO2 streams, or the scope of ‘legal obligations’ that will be transferred after site closure. As to the latter, some may understand this as a transfer of management and monitoring only, while others may see it as including a transfer of liability. This may have consequences for long-term liability and may be affecting the market formation in individual EU Member States. Furthermore, more guidance is needed for competent authorities to be sure that a site can be considered safe, that site performance is satisfactory and that legal obligations can be transferred, once a site has been closed. Although more research, demonstration and experience are necessary to find the answers to these questions, the adopted Storage Directive has enhanced confidence in the safety of CCS operations and allowed for progress on market formation.

3.4.2 CCS in a bridging market: financing demonstrations The central vehicle for CO2 emission reductions in the EU is the EU Emissions Trading Scheme (EU ETS), which is discussed in the next section. The EU ETS provides structural incentives for CO2 emission reductions and enables participants in the scheme to exploit available abatement options at the lowest cost. It does not, however, provide an incentive for investing in more costly and less mature reduction technologies, such as CCS (Groenenberg and Coninck, 2008). Although the ETS directive has been amended recently and has a time horizon up to and beyond 2020 now, emissions prices are still too low to sufficiently stimulate the option. CO2 forward prices for 2012 and beyond are on the order of 15–40 7/tCO2. Such prices are insufficient for inducing structural deployment of CCS given the technological uncertainties of components of CCS that are still in the demonstration phase. Therefore, funds were made available in the new entrants reserve of the EU ETS to provide an extra stimulus for companies to demonstrate CCS. 300 Mt worth of emission allowances will be allocated to demonstrations of CCS and innovative renewable energy demonstrations. Criteria for the allocation of the new entrants reserve funds remain to be drafted; the exact funding depends on the carbon price, and the share of renewable energy estimates for CCS demonstration funding from the new entrants reserve vary from 73 billion to 79 billion. In addition, additional funds for CCS as part of an EU economic stimulus package were agreed in April 2009 (Council of the EU, 2009), comprising another 71.05 billion allocated to 12 projects in the EU.


84


3.4.3 Towards mass markets: regulation for incentivising CCS The EU emissions trading scheme The EU considers the EU Emissions Trading Scheme (2003/87/EC) a crucial and cost-effective instrument for meeting its climate objective of reducing greenhouse gas emissions by at least 20 % in 2020 compared to 1990. It will be important also for the market formation of CCS, after most technological barriers have been addressed in a demonstration phase. The EU Energy and Climate Package, passed in December 2008, included provisions on ETS inclusion of CCS until 2013, the beginning of the third phase of the ETS. Before that, in the second phase of the EU ETS (2008– 2012), CO2 capture, transport and storage operations will be included, but not as separate ETS installations in their own right. Instead, CCS chains are included in the EU ETS in their entirety, requiring considerable coordination between CO2 sources, such as power companies, operators of pipelines and storage operators. Installations that operate a full CCS chain would not need to hand in allowances for the CO2 that is stored. From 2013 onwards, installations capturing, transporting or storing CO 2 should be covered by the trading scheme ‘in a harmonized manner’, in order to encourage and provide incentives for full-scale deployment of the option. After 2013, there will be no free allocation of emission allowances for installations in the power sector. Instead, allowances will be auctioned in this sector, and CO2 captured and transmitted for storage will not count as emitted under the EU ETS. This provides greater regulatory certainty and could enhance capture and storage of CO2 from power installations. Greenhouse gas allowances under the EU ETS require a monitoring scheme to be in place that is tailored to the various GHG emission sources in an ETS installation, and any emissions to the atmosphere would need to be offset under the EU ETS by returning a respective amount of emission certificates to the national competent authority. The EU ETS monitoring and reporting guidelines The inclusion of CCS in the EU Emissions Trading Scheme has created a need for monitoring and reporting guidelines for CCS operations under the scheme. The IPCC already gave an outline for accounting procedures for CCS in its 2006 Guidelines, which provide binding guidance for accounting in the national inventories, such as those to meet the obligations of countries under the Kyoto Protocol. These cover site characterisation, monitoring and the use of history matching to assess site performance. The Storage Directive does not address uncertainties in the monitoring program or detection limits and, as yet, it is uncertain to what extent the CCS



85

Annex to the 2007 Monitoring and Reporting Guidelines will contain more prescriptive wording (Wartmann et al., 2009). However, it will be difficult to report emissions from CCS operations at accuracy levels mentioned. The precision of in-situ methods, such as seismic methods, pressure measurements, electromagnetic monitoring and gravimetric techniques, is limited. Accuracies for these methods have not been systematically assessed, but it has been estimated that the precision will not likely exceed ±20 % (Benson, 2006). Flux measurements in most cases will be more exact. However, emissions need to be detected and located before monitoring equipment can be installed at the right locations, baseline emissions arising from e.g. the decomposition of soil organic matter would need to be measured too, and all in all substantial efforts may be needed to quantify a relatively small seepage. Norway In Norway, the introduction of a carbon tax in the early 1990s proved very effective in cutting back on carbon emissions and directly triggered the storage of CO2 under the North Sea, notably in the Sleipner operation. It indirectly also led to planned capture operations for gas-based power generation capacity. Still, the carbon tax, even at 407/tCO2, is too low to incentivise commercial CCS projects. Similarly, a domestic emissions trading scheme that could be linked to the EU ETS might not be strong enough to trigger more CCS projects. Norway currently uses 100 % hydropower for its electricity production. Although government support is being debated and although the intention is not to allow any emission from gas-based power plants, as yet in Norway no market has formed for CCS (Alphen et al., 2009).

3.4.4 Other incentives: capture readiness and emission performance standards Other options for CCS incentives were discussed as the Directive for Geological Storage of CO2 was developed and debated in the EU institutions. Apart from regulating safety and incentives for CCS, it includes a requirement on the capture-readiness of new plants. Member States will need to ensure that operators of all new combustion plants with a rated electrical output of 300 MW or more have assessed a number of conditions, namely the availability of a suitable storage site, and the technical and economic feasibility of both transport facilities and retrofit for CO2 capture. If these conditions are met, the Directive prescribes that suitable space on the installation site be set aside for the equipment necessary to capture and compress CO2. One of issues during discussions with regard to the Storage Directive was the inclusion of an emission performance standard for fossil fuel-based power generation. It was considered unacceptable, notably by Member States


86


with relatively high emissions and limited options to store CO2, such as in Central Europe. It shows that legitimation for CCS is not general among all countries. Another objection voiced was lack of arguments for duplication of policies: if an emissions trading scheme makes it possible to reach a low level of emissions at minimum cost, then any additional policy or regulation can only increase the cost for reaching that same emission level. Still, a mandatory requirement for emission performance standards for new electricity-generating large combustion plants may be considered again in 2015, during the evaluation of the Storage Directive.

3.4.5 The EU and CCS market formation: a mixed story In summary, the EU by its attempt to tackle both legislating and incentivising CCS in one Directive showed comprehension of the market formation function in the CCS technological innovation system. It seems to succeed in creating a legal safety regime for CCS, and other countries are currently considering adopting relevant parts of the EU regulation. On the incentives side, modelling studies show that the EU ETS is insufficient to provide incentives for CCS and that the demonstration funds committed might also not suffice. Although the EU takes market formation seriously, it has not succeeded fully in addressing this function.

3.5

Implementation outlook for carbon dioxide (CO2) capture and storage (CCS) technologies

Discussions of the outlook for CCS technology regularly stop at economic and technological arguments. For instance, it is often heard that CCS could grow to a major, ten to hundred billion dollar global market (IEA, 2008a) – a vision that positively affects the direction of search. Although the technological, geological and economic limits of the technology are also emphasised, those problems are presented as surmountable, resulting in a great deal of technological optimism around CCS. Rather than taking a technocratic view on the condition of CCS, this chapter used a socio-technical method, technological innovation systems, to provide an outlook for CCS technology and arrives at slightly different conclusions. This chapter argued, based on a global TIS analysis, that in the field of legitimation, market formation and, in some countries, also direction of search and entrepreneurial experimentation, the innovation system is currently insufficient to enable rollout and even demonstration of CCS. The TIS method points at a wide range of issues that all need to be addressed if CCS is to be advanced beyond the ‘bridge market’ stage. Of the conclusions, those around market formation are probably least contested. It is already widely recognised that CCS needs to be demonstrated at scale in order to address



87

market formation, direction of search and resource mobilisation issues (ZEP, 2007; G8, 2008; IEA, 2008b). Some government and industry actors are indeed pursuing demonstrations, and governments are making available funds and are formulating flanking policies. The EU is well underway with a legal framework and incentives for CCS, but these are broadly considered insufficient to demonstrate CCS in the short term, and diffuse it in the long term. Other countries largely lack structural incentives for CCS alltogether. The conclusions around legitimation are less broadly acknowledged. Policymakers, albeit slowly in some cases, increasingly understand the relevance of CCS technology for addressing climate change. Industrial parties are becoming more and more appreciative of the viable business case that CCS may offer at some point. However, public perception and involvement is still a contentious issue. Although many experts refer to public acceptance as one of the major challenges for CCS technology (Shackley et al., 2009), research and studies into public perception have been limited. As a result, several questions remain regarding the current status of public perception, opinion and involvement. And even firm conclusions from social scientists – such as that early public engagement should be a standard practice in CCS project planning – are often not broadly acknowledged by project developers, let alone implemented. Although market formation and legitimation stand out, other functions in the TIS on which CCS performs relatively well, such as private sector legitimation and knowledge development, knowledge diffusion, resource mobilisation and further guidance will also require constant and increasing attention from all actors involved in the system. Responsible actors for these functions are limited to a small number of directly involved industrial or governmental actors. It should also be noted that demonstration of CCS at scale is essential for all functions in the CCS TIS. CCS is often characterised as an industrialised-country mitigation option. Indeed, the lion’s share of current knowledge development is taking place in developed countries; particularly the EU, the USA and Australia (Tjernshaugen, 2008). Future models, however, project that the long-term demand for CCS is greater in developing countries than in industrialised countries (IEA, 2008a) and that deployment in emerging economies should start in the period 2020–2030. In those countries, the functions of the CCS TIS are less developed and CCS is likely to require more time to mature. In most developed countries, during the past couple of years significant development of human capacity, regulatory frameworks and political awareness has taken place. This points at a current need to identify structural components and develop functions for CCS in developing countries. As demonstration of CCS has collective action characteristics, it has been argued that a publicly funded, internationally coordinated demonstration programme would be able to make this happen (Coninck et al., 2009).


88


A last point might suggest an omission in the TIS methodology. As the TIS methodology concerns the advancing of a technology, it is not a methodology for evaluating whether a technology can be phased out again. CCS is a technology where an exit strategy may be as important as an entry strategy. Not only is the underground storage potential limited, CCS to some actors is only acceptable as a temporary solution to climate change on the road to a fully renewable-based energy system (Stangeland, 2007; Greenpeace, 2008). CCS has been called a ‘bridging technology’ by many players. This means that a strategy to rid the energy system of CCS after it has contributed to bridging the time required to transition to a renewable energy system is important. Research on methods to achieve this has been largely absent.

3.6


∑ IEA (2009) Technology Roadmap Carbon capture and storage. Paris, France, IEA. ∑ CATO: Netherlands research programme on CCS, including CCS innovation systems (www.co2-cato.nl). ∑ Global CCS Institute: www.globalccsinstitute.com.

3.7

References

Alphen K van, Ruijven J van, Kasa S, Hekkert M and Turkenburg W (2009) The performance of the Norwegian carbon dioxide, capture and storage innovation system. Energy Policy 37, 43–55. Alphen K van, Voorst tot Voorst Q van, Hekkert MP and Smits REHM (2007) Societal acceptance of carbon capture and storage technologies. Energy Policy 35(8), 4368– 4380. Ashworth P, Carr-Cornish S, Boughen N and Thambimuthu K (2009) Engaging the public on Carbon Dioxide Capture and Storage: Does a large group process work?. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4765–4773. Ashworth P, Pisarski A and Littleboy A (2006) Social and Economic Integration Report: Understanding and incorporating stakeholder perspectives to low emission technologies in Queensland. Dawson, Australia: CSIRO Centre for Low Emission Technology. Bakker S, Coninck HC de, Groenenberg H (2009) Progress on including CCS projects in the CDM: Insights on increased awareness, market potential and baseline methodologies. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4467–4474. Benson SM (2006) Monitoring carbon dioxide sequestration in deep geological formations for inventory verification and carbon credits. SPE 102833, Annual technical Conference and Exhibition, San Antonio TX, 24–27 September, Society of Petroleum Engineers. Bergek A, Jacobsson S, Carlsson B, Lindmark S and Rickne A (2008) Analyzing the



89

functional dynamics of technological innovation systems: A scheme of analysis. Research Policy 37(3), 407–429. Best-Waldhober M de, Daamen D and Faaij A (2006) Public perceptions and preferences regarding large scale implementation of six CO2 capture and storage technologies. Well-informed and well-considered opinions versus uninformed pseudo-opinions of the Dutch public. Leiden, the Netherlands: Leiden University. Best-Waldhober M de, Daamen D, Hendriks C, Visser E de, Ramirez A and Faaij A (2008) How the Dutch evaluate CCS options in comparison with other CO2 mitigation options. Results of a nationwide Information-Choice Questionnaire survey. Leiden, the Netherlands: Leiden University. Best-Waldhober M de, Daamen D and Faaij A (2009) Informed and uninformed public opinions on CO2 capture and storage technologies in the Netherlands. International Journal of Greenhouse Gas Control 3(3), 322–332. Bradbury J, Ray I, Peterson T, Wade S, Wong-Parodi G and Feldpausch A (2009) The role of social factors in shaping public perceptions of CCS: Results of multi-state focus group interviews in the U.S. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4665–4672. Coninck H de, Stephens JC and Metz B (2009) Global learning on carbon capture and storage: a call for strong international cooperation on CCS demonstration. Energy Policy 37, 2161–2165. Council of the EU (2009) Presidency compromise proposal for financing of the infrastructure projects put forward by the Commission as part of the EERP. 7848/1/09 REV 1: available at: http://www.consilium.europa.eu/uedocs/cms_data/docs/pressdata/en/ ec/106847.pdf (accessed January 2010). CSLF (2009) CCS integrated project database (on-line). Washington DC, Secretariat Carbon Sequestration Leadership Forum. Curry T, Ansolabehere S and Herzog HJ (2007) A Survey of Public Attitudes Towards Climate Change and Climate Change Mitigation Technologies in the United States: Analyses of 2006 results. Cambridge, MA: MIT, Laboratory for energy and the environment. Daamen D, Best-Waldhober M de, Damen K and Faaij A (2007) Pseudo-opinions on CCS technologies. In Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK: Elsevier, CD-ROM. De Figueiredo MA (2003) The Hawaii Carbon Dioxide Ocean Sequestration Field Experiment: A Case Study in Public Perceptions and Institutional Effectiveness (S.M. Thesis), Cambridge, MA: Massachusetts Institute of Technology. EC (2008) Commission Staff Working Document: Accompanying Document to the Proposal for a Directive of the European Parliament and of the Council Amending Directive 2003/87/EC so as to improve and extend the EU Greenhouse Gas Emission Allowance Trading System, (Jan 23, 2008), available at: http://ec.europa.eu/environment/climat/ emission/pdf/com_2008_16_ia_en.pdf (accessed December 2009). GCCSI (2009) Constitution Global CCS Institute Ltd, available at: http://globalccsinstitute. com/downloads/Global_CCS_Institute_Constitution.pdf. (accessed December 2009). Gibbins J and Chalmers H (2008) Preparing for global rollout: A ‘developed country first’ demonstration programme for rapid CCS deployment. Energy Policy 36(2), 501–507.


90


Greenpeace (2008) False Hope. Why carbon capture and storage won’t save the climate. Amsterdam, the Netherlands: Greenpeace, available at: http://www.greenpeace.org/ international/press/reports/false-hope (accessed December 2009). Groenenberg H and Coninck HC de (2008) Effective EU and Member State policies for stimulating CCS. International Journal for Greenhouse Gas Control 2, 653–664. Groenenberg H, Carpenter M, Flach T, Bernstone C, Zink-Jørgensen K, Schröder P, Beck B, Christensen NP, Chatwick A, Lescanne M and Aimard N (2008) Guidelines for licensing CO2 storage operations around the globe. The Leading Edge, April 2008, available at: http://www.mofa.go.jp/policy/economy/summit/2008/doc/doc080714__ en.html (accessed January 2009). G8 (2008) G8 Hokkaido Toyako Summit Leaders Declaration. Hokkaido Toyako, 8 July 2008. Ha-Duong M, Nadaï A and Campos AS (2009) A survey on the public perceptions of CCS in France. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4757–4765. Hansson A and M Bryngelsson (2009) Expert opinions on carbon dioxide capture and storage — A framing of uncertainties and possibilities. Energy Policy 37, 2273–2282 Hekkert MP, Suurs RAA, Negro SO, Smits REHM and Kuhlmann S (2007) Functions of innovation systems: a new approach for analyzing technological change. Technological Forecasting and Social Change 74, 413–432. IEA (2008a) World Energy Outlook. Paris, France: IEA. IEA (2008b) CO2 capture and storage: A key abatement option. Paris, France: IEA. IPCC (2005) Special Report on Carbon Dioxide Capture and Storage. Prepared by Working Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, Coninck HC de, Loos M, and Meyer LA (eds), Cambridge, UK and New York, NY: Cambridge University Press. IPCC (2007) Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Metz B, Davidson OR, Bosch PR, Dave R and Meyer LA (eds), Cambridge, UK and New York, NY: Cambridge University Press. Itaoka K, Okuda Y, Saito A and Akai M (2009) Influential information and factors for social acceptance of CCS: the 2nd round survey of public opinion in Japan. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4803–4810. Itaoka K, Saito A and Akai M (2005) Public acceptance of CO2 capture and storage technology: A survey of public opinion to explore influential factors. In: Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Cheltenham, UK: IEA GHG, Vol. 1, 1011–1019. Itaoka K, Saito A and Akai M (2007) A path analysis for public survey data on social acceptance of CO2 capture and storage technology. In Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK: Elsevier, CD-ROM. Lambrichs C (2008) Pseudo-opinions on carbon capture and storage technology: When and why are pseudo-opinions unstable? Masters Thesis. Leiden, the Netherlands: Leiden University. Lundvall B (1992) National Systems of Innovation: Towards a theory of innovation and interactive learning. London, UK: Pinter. © Woodhead Publishing Limited, 2010


91

Malone EL, Bradbury JA and Dooley JJ (2009) Keeping CCS stakeholder involvement in perspective. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4789–4794. Mors E ter (2009) Dealing with Information about Complex Issues: The role of source perceptions. PhD Dissertation, Leiden, the Netherlands: Leiden University. Murphy and Edwards (2003) Bridging the Valley of Death: Transitioning from Public to Private Finance, NREL/MP-720-34036, Golden, CO: NREL. Palmgren CR, Morgan G, Bruine de Bruin W and Keith DW (2004) Initial public perceptions of deep geological and oceanic disposal of carbon dioxide. Environmental Science & Technology 38(24) 6441–6450. Raven RPJM, Jovilet E, Mourik RM, Feenstra YCFJ (2009) ESTEEM: Managing societal acceptance in new energy projects: A toolbox method for project managers. Technological Forecasting and Social Change 76(7), 963–977. Reiner D, Curry T, De Figueredo M, Herzog H, Ansolabehere S, Itaoka K, Akai M, Johnsson F and Odenberger M (2007) An international comparison of public attitudes towards carbon capture and storage technologies. In Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK: Elsevier, CD-ROM. Shackley S, McLachlan C and Gough C (2005) The public perception of carbon capture and storage in the UK: results from focus groups and a survey. Climate Policy 4(4), 377–398. Shackley S, Reiner D, Upham P, de Coninck H, Sigurthorsson G and Anderson J (2009) The acceptability of CO2 capture and storage (CCS) in Europe: An assessment of the key determining factors Part 2. The social acceptability of CCS and the wider impacts and repercussions of its implementation. International Journal of Greenhouse Gas Control 3, 344–356. Sharp J, Jaccard M and Keith D (2007) Public attitudes toward geological disposal of carbon dioxide in Canada. In Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK: Elsevier, CD-ROM. Stangeland A (2007) Why CO2 capture and storage (CCS) is an important strategy to reduce global CO2 emissions. Bellona Position Paper, available at: http://www. bellona.org/filearchive/fil_Bellona_Paper_-_Why_CCS_-_1June07.pdf (accessed December 2009). Terwel B (2009) Origins and consequences of public trust: Towards and understanding of public acceptance of carbon dioxide capture and storage. Dissertation, Leiden, the Netherlands: Leiden University. Tjernshaugen A (2008) Political commitment to CO2 capture and storage: evidence from government RD&D budgets. Mitigation and Adaptation Strategies and Global Change 13, 1–21. Tokushige K, Akimoto K and Tomoda T (2007) Public perceptions on the acceptance of geological storage of carbon dioxide and information influencing the acceptance. International Journal of Greenhouse Gas Control 1, 101–112. Viebahn P, Nitsch J, Fischedick M, Esken A, Schuwer D, Supersberger N, Zuberbuhler U and Edenhofer O (2007) Comparison of carbon capture and storage with renewable energy technologies regarding structural, economic, and ecological aspects in Germany. International Journal of Greenhouse Gas Control 1(1), 121–133. Wallquist L, Visschers VHM and Siegrist M (2009) Lay concepts on CCS deployment


92


in Switzerland based on qualitative interviews. International Journal of Greenhouse Gas Control 3(5), 652–657. Wartmann S, Groenenberg H and Brockett S (2009) Monitoring and reporting rules for CCS operations in the EU Emissions Trading Scheme. In Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1, 4459–4466. 16–20 November 2008, Washington, DC. Zakkour P, King E, Cook G, Maruyama N and Rana S (2008) Carbon Dioxide Capture and Storage in the Clean Development Mechanism: Assessing market effects of inclusion. ERM report 2008/13, November, Cheltenham UK: IEA Greenhouse Gas R & D Programme. ZEP (Zero Energy Platform) (2007) The EU Flagship Programme: The key to making CO2 capture and storage (CCS) commercially viable by 2020, available at: http:// www.zero-emissionplatform.eu/website/docs/ETP%20ZEP/EU%20Flagship%20 Programme%20for%20CCS.pdf (accessed January 2010).


4

Energy supply planning for the introduction of carbon dioxide (CO2) capture technologies

A. E l k a m e l, H. M i r z ae s m aee l i, E. C r o i s e t and P. L. D o u g l a s, University of Waterloo, Canada Abstract: Planning for the future energy supply mix is a very challenging undertaking which requires consideration of various drivers and decision criteria. This chapter will address the multi-period energy planning problem with carbon dioxide (CO2) emission constraints and the option of CO2 capture and storage (CCS). The objective of the current chapter is to present a novel multi-period mixed-integer non-linear programming (MINLP) model that is able to realize the optimal mix of energy supply sources which will meet current and future electricity demand and CO2 emission targets, and lower the overall cost of electricity. The modeling framework discussed is based on an objective function that minimizes the net present value of the cost of electricity (COE) over a given time horizon. The formulation incorporates several time-dependent parameters such as forecasted energy demand, fuel price variability, construction lead time, conservation initiatives, and increase in fixed operational and maintenance costs over time. The model is applied to a case study in order to examine the economic, structural, and environmental effects that would result if an electricity sector was required to reduce its CO2 emissions to a specific limit. The case study examines a scenario in which the electricity sector must comply with CO2 emission limits similar to a Kyoto target. The presented model offers many potential benefits to the energy sector. In addition to providing an optimal solution for meeting future electricity demand, it can help meet emissions targets while minimizing the overall cost of electricity. Furthermore, although the case study was aimed at Ontario’s future energy supply mix, it could also be readily applied to other regions or even countries as a whole. Key words: energy supply, emissions targets, multi-period mixed-integer non-linear programming, carbon dioxide capture and storage, cost of electricity.

4.1

The emerging energy challenge and a case from Ontario, Canada

The ratification of the international Kyoto protocol, as extensively discussed in other chapters throughout this book, imposes additional pressure for many governments to reduce CO2 emissions. For example, under the Kyoto protocol, which was signed by the Canadian government in 1998 and ratified in 2002, Canada agreed to reduce greenhouse gas (GHG) emissions by 6 % below 1990 levels by 2008–2010. 93 © Woodhead Publishing Limited, 2010

94


Although the Kyoto directive is to reduce GHG emissions, there has been an increasing trend in emissions in Canada for the past several years. In 2003, Canada contributed about 740 Megatonnes (Mt) of CO2 equivalent of GHGs to the atmosphere, an increase of about 3 % over the recorded emissions from 2002 (Fig. 4.1). This increase in GHG emissions is significantly greater than the 1 % increase which occurred between 2001 and 2002. If no mitigating measures are taken, it is estimated that Canada’s GHG emissions will rise to 809 Mt by 2010 (Environment Canada, 2005). To provide a more specific example, consider Ontario’s operable generation capacity which is currently equal to approximately 30 662 MW from all sources (Ontario Power Authority, 2005). As shown in Fig. 4.2, 37 % of total capacity can be attributed to nuclear sources, 26 % is renewable, 21 % is coal-fired, and the remaining 16 % consists of gas and oil fueled sources.

GHG emission [Mt CO2 eq]

850

2003 emissions 740 Mt or 24 % above 1990

800 750 700

1990 Baseline 596 Mt

650 600

Kyoto target: 6 % below 1990 baseline

550 1990 1992 1994

1996

1998

2000 2002 Year

2004 2006

560 Mt 2008 2010 2012

4.1 Canada’s emission trend and Kyoto’s emission target (Environment Canada, 2005).

16 % 37 %

Nuclear Renewables

21 %

Coal Gas/oil 26 %

4.2 Ontario’s current installed generation capacity in terms of percentile (Ontario Power Authority, 2005).


Energy supply planning

95

Ontario’s electricity sector faces one of the most challenging times in its history. While the demand for electricity has increased due to economic growth and rising population, the electricity sector’s capacity has decreased over the past decade (Ontario Power Authority, 2005). As shown in Fig. 4.3, Ontario’s supply capacity is projected to decline over the next two decades with a rapid decline in supply emerging in the next several years. Two major factors contribute to this decline in supply capacity. Firstly, it can be seen that Ontario’s coal-fired capacity will be completely removed from the electricity supply mix by 2009. This is mainly due to Ontario’s current government legislating the closure of all coal-fired plants within the next few years. Secondly, nuclear capacity declines sharply due to the retirement of many of the existing nuclear units. The combination of these two factors will decrease Ontario’s installed capacity by approximately 17 316 MW in the next 20 years. In the meantime, Ontario’s demand increases steadily over the next two decades. If Ontario’s current consumption and demand continue, the required resources rise from 27 000 MW in 2006 to approximately 37 000 MW in 2025. The decline of supply and increase in demand results in a potential energy gap of 24 000 MW by 2025. A solution must be found to fill this energy gap and meet the long-term capacity needs. To further complicate Ontario’s future supply–demand shortfall problem, GHG emissions must be taken into consideration when evaluating potential solutions. CO2 is suspected to be the principal GHG responsible for global warming and climate change. With a growing concern with global warming and its effects on the environment, the industry is striving to reduce its CO2 emissions. Ontario Power Generation (OPG), which accounts for 70 % of electricity generation in Ontario, has had wavering CO2 emissions over the past few decades. CO2 emissions from OPG have varied from a high of 37 million tonnes in 2000 to a low of 15 million tonnes in 1994. The variability in CO2 emissions from OPG is mainly due to changes in electricity demand and technological improvements. Since 1995, OPG has entered into a voluntary commitment to reduce their GHG emissions to levels equivalent to the 1990 baseline. Though some progress in achieving this target has been made over the past few years, OPG has often had to resort to buying CO2 emission reduction credits to achieve their voluntary targets. As of this time, no long-term sustainable strategy has been established by OPG to address their ongoing CO2 emission challenge. Similar to the above case, many countries are facing a large energy gap due to increasing demand and a decline in installed generating capacity. Furthermore, as environmental regulations become more stringent, there will be an increasing need to reduce the amount of GHGs and other pollutants emitted to the environment. Consideration must be given to meeting the


96

40 000 — Required resources = peak demand + reserves

30 000

MW


25 000

Gap = 24 000 MW

Coal

20 000

15 000

10 000

Nuclear

Natural gas and oil

5000 Renewable 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026

4.3 Ontario’s demand growth and installed generating capacity portfolio over 2005–2026 (Ontario Power Authority, 2005).


35 000


97

rising energy demand in both an environmentally sound and cost-effective manner. In light of all the issues discussed, a sustainable energy mix in order to realise future challenges must be developed. There are various supply technologies available that could be used to help meet energy demand. These supply options differ based on a few factors, including economic, environmental, and operational characteristics. Some technologies offer lower capital and operating cost at high emission rates, while other supply options have higher associated costs but lower environmental impacts. Furthermore, there are several options available for pollutant mitigation, such as CO2 capture and storage (CCS). The underlying question then becomes what mix of supply technologies and pollutant mitigation options should be selected to meet Ontario’s energy demand and environmental limits at a minimal cost. This is the question that this chapter aims to answer and is its main motivation. Planning for a future energy supply mix is a very challenging undertaking which requires consideration of various drivers and decision criteria. From the literature review conducted, no prior work has been found addressing the problem of finding the optimal strategy for energy planning with CO2 emission constraints and the option to implement CCS. The objective of this chapter is to present a novel optimization model in order to realize the optimal mix of energy supply sources, with consideration for CO2 emissions. The aim is to develop a deterministic multi-period mixed-integer non-linear programming (MINLP) model that is able to realize the optimal mix of energy supply sources which will meet current and future electricity demand and CO2 emission targets, and lower the overall cost of electricity. The objective function will be based on minimizing a net present value of the cost of electricity (COE) over a given time horizon. The formulation incorporates several time-dependent parameters such as forecasted energy demand, fuel price variability, construction lead time, conservation initiatives, and increase in fixed operational and maintenance costs over time. Although this chapter focuses on Ontario’s future energy supply mix, the methodology and model discussed could also be readily applied to other regions or even countries as a whole. The remainder of this chapter is organized into four sections. Section 4.2 provides a detailed background on potential supply technologies and CO2 mitigating options. Moreover, this section gives an overview of current energy mix and projected future outlooks for the case of Ontario. The next sections look at future trends and energy conservation Section 4.5 presents the mathematical formulation for the deterministic multi-period MINLP model. Section 4.6 details the case study used to implement the mathematical model developed and contains data for the different parameters needed in the model and provides a comparative analysis of the case study. Finally, the chapter ends with concluding remarks and presents recommendations for future work.


98

4.2


Overview of supply technologies and carbon capture and storage

4.2.1 Supply technologies Nuclear power stations Nuclear power stations produce, contain, and control the energy obtained from splitting of uranium atoms. Nuclear reactors act like large steam engines. The energy released from electric power plants is used to heat water and produce steam, which then drives the turbine-generators to produce electricity. Just like fossil fuel plants use burning of coal, oil or gas as a heat source, nuclear power stations use heat given off from the splitting of U235 atoms for their heat source. The most common type of nuclear reactors are pressurized water reactors, comprising 59 % of reactor types used worldwide (Naini et al., 2005). These reactors can use either light water or heavy water to control the speed at which the atoms travel and hence increase the amount of energy released from fission of uranium atoms. An example of a pressurized heavy water reactor (PHWR) is CANDU (CANada Deuterium Uranium) reactor technology. A CANDU plant uses uranium as a fuel source. Uranium atoms are split in the reactor, giving off energy in the form of heat. This heat is then used to boil water in steam generators, producing high pressure steam which is used to turn the blades of a turbine. The turbines turn the electrical generators which produce electricity that is sent to the customers. Natural gas power stations Natural gas power stations use natural gas as a source of fuel. There are two types of turbines that can be used to provide power to natural gas power stations for electricity production: steam turbines or gas turbines. Steam turbine systems use high-temperature and high-pressure steam to transfer energy to rotating turbine blades, while gas turbines use gas expansion. The turbines are then used to turn electrical generators for production of electricity. There are three types of technologies that can be used in natural gas power stations: simple cycle gas turbine, natural gas combined cycle turbine, and cogeneration turbine. Each technology is discussed in more detail in the following sections. A fourth type, fuel cells, is also available but will not be discussed here. Simple cycle gas turbines compress air in an air compressor. This compressed air is used to burn natural gas in a combustion chamber. The resulting high-temperature combustion gas and air mixture expands in the turbine, driving an electrical generator to produce electricity. Natural gas combined cycle (NGCC) power plants produce electricity from



99

a combination of a gas cycle and a steam cycle. The gas cycle is identical to the one described in the simple cycle gas turbine section. In addition to the gas cycle, the waste heat of the exhaust gases leaving the gas turbine is used for steam generation in a heat exchanger. The steam generated from the heat exchanger is used to drive a steam generator and produce additional electricity. Cogeneration is similar to NGCC, and uses exhaust gases leaving the gas cycle as feed to a heat exchanger. However, whereas NGCC uses steam produced from the heat exchanger to drive a steam generator and produce additional electricity, cogeneration uses the thermal energy of the steam directly for purposes such as industrial processes or water heating. Hence, in cogeneration, the steam is not used to produce electricity. Coal power stations Coal power plants use coal as a fuel source for power generation. Technologies used in coal power plants are categorized into two groups: combustion and gasification. Pulverized coal power stations use combustion technologies while gasification technology is an integrated gasification combined cycle. For pulverized coal power stations, pulverized coal is fed to a steam boiler and steam turbine. Coal is first ground to a very fine powder for combustion. The pulverized coal is then combusted in a series of burners, generating hot gases that are used to produce steam in a boiler. The steam is used to turn a turbine which drives a generator and produces electricity. During the process of coal burning, ash is formed in the combustion chamber. The bottom ash consisting of large particles can be collected and removed. The rest of the coal ash remains in the combustion chamber and is known as fly ash. Some fly ash can be captured using various air pollution control technologies. For Integrated Gasification Combined Cycle (IGCC), coal is gasified by partial combustion to produce synthetic gas. This process uses a gasification agent consisting of air, oxygen, and steam. IGCC combines gas and steam turbines for electricity production. Coal slurry is reacted with oxygen (or air) and steam, and syngas is produced, consisting mainly of CO and hydrogen. The raw syngas is cooled and cleaned to remove particulates and sulphur impurities. The clean syngas is burned in a combustion turbine which drives a generator to produce electricity. The hot exhaust gases are recovered and used to produce steam. The resulting steam is used to drive a steam turbine, which turns a generator and produces additional electricity. Hydroelectric power stations Hydroelectric power stations generate electricity using the force of water that falls into turbines and rotates the shaft. By rotating the shaft of turbines,


100


the potential energy of the water is converted into kinetic energy. The shaft from the turbine is connected to a generator. The kinetic energy from the shaft turns the electrical generator and produces electricity. Water for use in hydroelectric power stations can be obtained by building a dam on a large river. Water is stored behind the dam in large reservoirs and can be released onto turbine propellers through a dam water intake. After passing through the turbine, the water is released back into the river. Wind power plants Wind power plants generate electricity by using wind to turn wind turbines. In principle, wind’s potential energy is converted to kinetic energy that rotates the blades of turbines, which in turn transfer this energy to an electrical generator. The electrical generators produce electricity. Comparison of the different technologies The environmental performance and cost of each of the above technologies are discussed in this subsection. Nuclear power plants do not emit any GHGs or ozone precursors during normal operation. However, there are radioactive emissions from nuclear power plant’s operation. These emissions have been found to be less than the radioactive emissions from coal-fired power plants (Naini et al., 2005). Capital costs and construction periods of nuclear power plants are generally higher than for coal or gas power stations. However, the fuel costs are considerably lower (Ontario Power Authority, 2005). Natural gas power stations costs depend on the size of the power plant and the selected turbine technology. Simple cycle gas turbines have lower capital costs than combined cycle and cogenerators, but generally have lower efficiency. One advantage of simple cycle gas turbines is that they have fast start-up times and can hence provide electricity for peak-load demand. However, since they do not have long operating times, simple cycle gas turbines are not efficient for base-load service. NGCCs have higher capital costs than simple cycle gas turbines, but lower operating costs. Also, NGCC is generally more efficient then simple cycle gas turbines. NGCCs can be used for base-load or peak-load. Cogeneration has higher capital costs than NGCC’s and simple cycle gas turbines, but higher efficiency (Ontario Power Authority, 2005). Coal power stations generally have lower fuel costs than other fossil fuels. IGCC power plants generally have higher capital costs than other competing technologies. Also, IGCC power plants have higher operating costs than pulverized coal power stations. Coal power plants emit pollutants, such as CO2, NOx, SO2, and particulates. In addition, combustion of coal results in the release of mercury, benzene, and formaldehydes. Radioactive elements



101

such as radon and uranium are also released from coal power stations. Whereas pulverized coal power stations produce fly ash, none is generated by IGCC plants. Use of hydroelectric power stations for electricity production is usually cheaper than use of other technologies because there are no fuel costs associated with hydroelectric plants. Also, the efficiency rate of electricity produced from hydro sources is about double compared to fossil fuel plants (Naini et al., 2005). However, hydroelectric plants depend on water availability which makes electricity production vulnerable to seasonal droughts and changes in weather. Hydroelectric power stations do not generate any GHGs or other atmospheric emissions. However, there are some negative environmental implications associated with hydroelectricity production. Notably, hydroelectricity generation has adverse impacts on agriculture and river ecological system since dams can lower water tables, alter water temperatures, and damage water wildlife. Similarly to hydroelectric power stations, wind power plants have no fuel costs associated with their operation since they use wind energy to produce electricity. However, wind power plants heavily depend on wind conditions. Moreover, wind is intermittent by nature and the electricity generated by wind turbines will vary depending on wind strength (Ontario Power Authority, 2005).

4.2.2 Carbon dioxide (CO2) capture and storage Carbon dioxide capture and storage (CCS) has received widespread interest as a potential method for controlling and reducing CO2 emissions from fossil fuel power plants (Rao & Rubin, 2002). The basic design of a CCS system includes four fundamental processes. The first process involves the separation and concentration of the CO2 present in the gas stream of fossil fuel power plants. Once the CO2 is separated and concentrated to a nearly pure form, it is compressed beyond its critical value in order to convert the concentrated CO2 gas into a liquid phase and allow for liquid phase transportation. The third stage of the process involves the transport of the concentrated liquid CO2 stream via a network of pipelines to a storage location. Finally, the last stage of the process is the sequestration of the CO2 into a medium such as a deep saline aquifer or a depleted oil and gas reservoir for long-term storage (Benson & Surles, 2006). A CCS system can be implemented on any new or existing power plant in order to reduce and control CO2 emissions. The cost associated with retrofitting an existing power plant with a CCS system generally tends to be higher than that of a new power plant with a CCS system already in place. This cost difference is largely due to the higher energy penalty that is incurred


102


by less efficient heat integration as well as potential site-specific difficulties that are inherent in most retrofit projects (Rao & Rubin, 2002). Although the cost of retrofitting an existing power plant with a CCS system may be higher, this may potentially be outweighed by the benefit of maintaining the operation of the power plant, while meeting CO2 emission targets, without having to build a new plant. Carbon capture technologies There exists a wide range of technologies that are currently available in order to separate and capture CO2 present in a gas stream. The carbon capture technologies available can be grouped in three general categories: postcombustion capture, pre-combustion capture and oxyfuel combustion. A post-combustion carbon capture process involves the removal of CO2 from the flue gas of power plants. The most common method for post-combustion carbon capture is a chemical absorption process that uses monoethanolamine (MEA) as a solvent. The process consists of running the flue gases through a low-pressure gas/liquid absorber where the CO2 is removed from the flue gas by partitioning with the amine solver. The amine is then heated to a specific temperature in order to release the pure CO2 and regenerate the solvent. Pre-combustion capture processes involve the removal of most of the carbon content in a fossil fuel before it is combusted. The process involves the pre-combustion reaction of the fossil fuel with steam and air, producing a syngas that comprises primarily CO and H2. The CO is then reacted with water to produce a mixture of CO2 and additional H2, which can then be separated and utilized for energy production and the CO2 stored respectively (Benson & Surles, 2006). Carbon dioxide sequestration refers to long-term and safe storage of carbon dioxide in a medium such as a deep saline aquifer or a depleted oil and gas reservoir. The process of carbon dioxide sequestration is normally performed after the carbon dioxide has been separated and captured from the gas stream of a power plant using a suitable process. For example, in Ontario, two large reservoirs for CO2 sequestration have been identified, one located in the southern part of Lake Huron and the other within Lake Erie. The approximate storage capacity for these two reservoirs has been estimated to be 289 and 442 million tonnes of CO2, respectively (Shafeen et al., 2004a). In order to achieve the estimated storage capacity of the reservoir, the injected CO2 must maintain a temperature and pressure condition beyond its critical value. This supercritical state of CO2 increases its density and allows large quantities of CO2 to be stored in a relatively small volume. The supercritical state of CO2 can only be maintained if it is stored at a minimum reservoir depth of 800 m (Shafeen et al., 2004b).



103

Ontario’s current energy mix Ontario’s current energy mix is composed of a variety of supply sources. The main technologies supplying electricity to Ontario are nuclear, hydroelectric, coal, natural gas and oil. The current installed capacity from all the supply sources totals approximately 30 662 MW (Ontario Power Authority, 2005). Table 4.1 presents the installed capacity of each supply technology in Ontario’s energy mix. Nuclear power plays a very important role in Ontario’s energy supply mix. Currently, nuclear power accounts for approximately 37 % of Ontario’s installed capacity and provides over 50 % of Ontario’s electrical energy needs. There are currently three CANDU nuclear power plants in Ontario: Pickering generating station, Darlington generating station and Bruce Power. Table 4.2 outlines the nuclear units available in Ontario and the expected operational lifespan of each unit. The end-of-service dates presented in Table 4.2 are uncertain estimates which may change based upon various factors such as refurbishment strategies and maintenance practices over the next few years. As shown in Table 4.2, most of the nuclear units were built in the 1970s and 1980s and are reaching the end of their expected service life. Consequently, most nuclear units will need to be retired or refurbished before 2018 (Winfield et al., 2004). Refurbishment of the existing nuclear units in Ontario would involve a wide range of work and require a great deal of economic investment. The most significant part of the refurbishment process, and incidentally the most expensive, is the replacement of the fuel channels of the reactors, a process referred to as large-scale fuel channel replacement (LSFCR). The LSFCR refurbishment process involves the restoration of the nuclear reactor core and requires the shut-down of the nuclear unit for a period of at least two years (Winfield et al., 2004). Ontario’s coal-fired power plants are a significant part of the current supply mix. Coal power plants account for approximately 21 % of Ontario’s installed capacity and provide for 19 % of Ontario’s electricity generation requirements. Ontario currently operates four coal-fired power plants and Table 4.1 Ontario’s current installed capacity based on supply sources (Ontario Power Authority, 2005) Technology

Existing capacity (MW)

Nuclear Hydroelectric Coal Gas/oil Other Total

11 397 7756 6434 4976 99 30 622


104


Table 4.2 Operational and out-of-service nuclear units in Ontario. The data presented in this table include the gross capacity, first commercial operation, and the estimated end-of-service date for each nuclear unit in Ontario Unit Status Pickering Nuclear Plant Pickering A Unit 1 Operational – was returned to service in 2005 Unit 2 Out of service Unit 3 Out of service Unit 4 Operational – was returned to service in 2003 Pickering B Unit 5 Operational Unit 6 Operational Unit 7 Operational Unit 8 Operational

Gross capacity (MW)

First End of commercial service operation dates

515

07/1971

n/a

515 515 515

12/1971 06/1972 06/1973

n/a n/a 2016

516 516 516 516

05/1983 02/1984 01/1985 01/1986

2008 2009 2010 2011

750

09/1977

n/a

750 750 750

01/1977 01/1978 01/1979

n/a 2012 2016

Bruce Nuclear Plant Bruce A Unit 1 Unit 2 Unit 3 Unit 4

Refurbished: expected start date 2009 Refurbished: expected start date 2010 Operational Operational

Bruce B

Operational Operational Operational Operational

785 820 785 785

03/1985 09/1984 04/1986 05/1987

2010 2009 2011 2012

Darlington Nuclear Plant Darlington Unit 1 Operational Unit 2 Operational Unit 3 Operational Unit 4 Operational

881 881 881 881

11/1992 10/1990 02/1993 02/1993

2017 2015 2018 2018

Unit Unit Unit Unit

5 6 7 8

one dual-fuelled oil and natural gas power plant. The coal power plants are Lambton, Nanticoke, Atitokan and Thunder Bay. The oil and natural gas power plant is the Lennox generating station. Table 4.3 presents the existing coal-fired power plants in Ontario. In Ontario, hydroelectric power accounts for approximately 26 % of the installed capacity available to the province, and provides for 23 % of the electricity generation. There are currently 108 hydroelectric stations within Ontario, but only 58 stations are directly connected to the electricity grid (Ontario Ministry of Energy, 2005). The largest hydroelectric stations in Ontario are the Niagara Plant Group which operate on the Niagara River and at DeCew Falls in St Catharines. These stations have a combined capacity of 2278 MW (Ontario Power Generation, 2006).



105

Table 4.3 Existing coal-fired power plants in Ontario Fuel

No. of units

Capacity (MW)

% of fossil fuel capacity

Dates in service

Nanticoke Lambton Thunder Bay Atikokan Lennox

8 4 2 1 4

3938 1975 310 215 2140

46 23 4 3 25

1973/1978 1969/1970 1981/1982 1985 1976/1977

Coal Coal Coal Coal Oil/gas

MW

L

h

8760 hours

4.4 Typical load–duration curve.

Currently, natural gas accounts for approximately 7 % of the supply mix in Ontario. There are presently 60 natural gas power plants of various capacities in Ontario, but only 19 of these stations are connected to Ontario’s electricity grid. The total installed capacity of all the natural gas-fired generating stations is approximately 2100 MW (Ontario Ministry of Energy, 2007).

4.3

Future trends

Electricity demand forecast A load duration curve is often used to help plan for electrical utilities, and a typical curve is presented in Fig. 4.4 (Murphy, et al., 1982). From Fig. 4.4, h is the number of hours in a year during which the demand is greater than or equal to a given load L (MW). The area under the curve represents the amount of energy, given in megawatt-hours, for a given period of time. For large-scale applications, such as large nuclear units and gas turbines, the load duration curve can be simplified using linear approximation. A typical two-step linear approximation is given in Fig. 4.5. From Fig. 4.5, a specific generating unit is assumed to operate in base-mode and/or peak-


106


P1 Peak

MW

P2 Base

B1

8760 hours

4.5 Linear approximation of load–duration curve.

mode. This two-step linear approximation is used in this chapter in order to simplify the problem. Various electricity demand forecasts for Ontario have been published. In 2005, Independent Electricity System Operator (IESO) forecast the energy and peak-load demand for Ontario for the ten-year period from 2006–2015. The results show that the energy demand is predicted to grow by 0.9 % annually over the forecast period. Total energy demand is expected to increase from 157 TWh to 170 TWh by 2015. IESO predicts an increase in the normal weather peak from 24 200 MW in 2006 to 25 700 MW in 2015, while the normal weather summer peak is expected to increase from 24 000 MW to 26 900 MW over the same time period. Furthermore, the forecast shows an average annual increase of 0.7 % for the winter peak and an average annual growth rate of 1.3 % for the summer peak (IESO, 2006). Navigant Consulting Ltd used IESO’s 2005 forecast to extrapolate electricity demand to 2025. In this forecast, annual hourly data was extracted from IESO’s forecast for the period 2006–2015. For the remaining analysis period 2016–2025, the 2015 typical week profile was extrapolated and fit to the annual energy and peak demand forecast (Navigant Consulting, 2005). The Ontario peak demand and energy consumption forecasts from Navigant Consulting Ltd are shown in Table 4.4. Chui et al. (2006) used a stochastic model to forecast Ontario’s electricity demand from 2006–2020. In this model, employment forecasts from the Ontario Ministry of Finance and various weather scenarios were used to predict electricity demand. This forecast contains a lower, median and upper bound. The lower bound uses a low employment growth rate and mild weather conditions, while the median bound uses median employment growth rate and median weather scenarios. Finally, the upper bound uses high employment growth rate and extreme weather scenarios. The forecast annual energy


2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Annual energy (TWh) Peak demand (MW)

156.8

158.3

160.3

161.2

162.6

164.2

166.0

167.0

168.4

169.7

24 205

24 374

24 627

25 045

25 228

25 534

25 840

26 461

26 461

26 874

2016

2017

2018

2019

2020

2021

2022

2023

2024

2025

Annual energy (TWh) Peak demand (MW)

171.2

172.7

174.3

175.8

177.4

178.9

180.5

182.1

183.7

185.3

27 211

27 552

27 898

28 248

28 602

28 961

29 692

29 692

30 064

30 441



Table 4.4 Forecasted peak demand (MW) and energy demand (TWh) from Navigant Consulting Ltd

107

108


demand, annual peak-load demand and annual base-load demand from Chui et al. (2006) are shown in Figs 4.6, 4.7 and 4.8, respectively. As can be seen from Fig. 4.6, Ontario’s annual energy demand grows by a range of 0.7 % to 0.97 % between 2006 and 2020. Figure 4.7 shows an annual increase in peak-load demand in the range of 1.21 % to 1.82 % for the same time period. Finally, from Fig. 4.8, it can be seen that the annual base-load demand grows in the range of 0.71 % to 0.99 %. In this chapter, we will use the above electricity forecast for the case study that will be discussed in Section 4.4. Fuel prices forecast Future fuel prices in North America will be affected by numerous factors, such as demand growth, productive capacity and the type of supply sources. For instance, the prices may vary depending on the availability of conventional and non-conventional supply sources and the industry’s cost and ability to bring them to market. There are numerous natural gas price forecasts reported in the literature. Sproule forecast natural gas prices based on Henry Hub daily closing prices. Based on this forecast, an upward trend was observed from 1997–2007, and prices were expected to fall in 2008 as the new set of liquefied natural gas (LNG) terminals come on-line (Naini et al., 2005). The Energy Information Administration’s (EIA) Annual Energy Outlook (AEO) was released in 2005 and included forecasts for the Lower 48 US Supplier’s average wellhead price. Unlike the Sproule forecast, the EIA’s AEO2005 forecast does not include the transportation costs of delivering natural gas to consumers and is hence based on lower prices than Henry Hub. The AEO2005 forecasts natural gas prices rising until 2008, and falling after the new LNG terminals come on-stream (Naini et al., 2005). This chapter uses the National Energy Board’s (NEB) natural gas price forecasts. The NEB forecasts natural gas prices delivered to industrial consumers in Ontario. It is based on two scenarios: a supply–push (SP) case and a techno–vert (TV) case. The SP scenario is based on an assumption that technology advances gradually and that there is limited action on the environment in Canada. One of the major premises of the SP case is the security and development of conventional North American gas sources using proven technologies. The TV scenario is based on the assumption that technology advances occur more rapidly and that Canadians take broad action on the environment. The heightened concern for the environment is assumed to result in an increasing demand for cleaner fuels and advances in technology. The outcome is rapid technological advances resulting in development of non-conventional gas sources (Naini et al., 2005). The NEB’s forecast is presented in Fig. 4.9.


200 000 195 000

Annual energy demand [GWh]

185 000 180 000 175 000 170 000 165 000 160 000 155 000

Median

Low

High

150 000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025

4.6 Ontario’s forecasted annual energy demand (GWh) for low, median, and upper bound (Chui et al., 2006).



190 000

109

110

38 000

Peak-load demand [MW]


36 000 34 000 32 000 30 000 28 000 26 000 24 000 Median

Low

High

22 000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025

4.7 Ontario’s forecasted annual peak-load demand (MW) for low, median, and upper bound (Chui et al., 2006).


40 000

15 000 14 500

Base-load demand [MW]

13 500 13 000 12 500 12 000 11 500 Median

Low

High

11 000 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025

4.8 Ontario’s forecasted annual base-load demand (MW) for low, median, and upper bound (Chui et al., 2006).



14 000

111

112

6 5.5


5 4.5

3.5 1986C$/GJ


4

3 2.5 2 1.5 1 0.5 Supply plus

Techno-vert

0 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026

4.9 Natural gas price forecast from NEB (1986 C$/GJ) showing both Supply-Plus and Techno-Vert scenarios. Numerical data for the annual natural gas forecast is presented. Costs are expressed in terms of 1986 Canadian dollars.


113

From Fig. 4.9, the natural gas prices are forecast to decrease after 2010 for both SP and TV scenarios. The TV scenario predicts a slight increase in prices between 2006 and 2010. Similarly to Sproule and AEO’s outlooks, the gas price decrease may be a result of the assumption that LNG terminals come-on stream after 2010. For coal prices, Sproule and EIA’s AEO2005 forecasts predict prices, measured in terms of US export price of coal, to decline from 2007–2025. AEO2005 forecast indicates coal prices will not drop below 2003US$ 35/ short ton for the next two decades (Naini et al., 2005). This chapter uses coal price forecasts from the NEB. The NEB coal price forecast is shown in Fig. 4.10. From Fig. 4.10, NEB’s coal prices are measured as delivered prices to industrial consumers in 1986 Canadian dollars per GJ coal. NEB projects coal prices to decline by 1 % until 2015, after which time the coal prices are expected to remain constant. It is assumed that there are no significant resource constraints on coal production. Also, continuing efficiency improvements such as mergers in the transportation industry are assumed.

4.4

Energy conservation strategy

Historically, conservation has occurred naturally with advances in technology. For instance, home appliances have been replaced by more efficient ones and building materials have become more energy-conserving with technological advances. Such energy efficiency improvements are known as ‘technology improvements’ and are typically reflected in demand forecasts. Conservation and demand management (CDM) is the use of a wide range of activities in an effort to reduce consumer demand and use of electricity. CDM usually results in higher levels of conservation than technology improvements due to more direct intervention in the market through incentives, standards or other mechanisms. The activities undertaken to reduce the use of electricity can be classified into three distinct categories: conservation efforts that result in less than normal use of electricity; energy efficiency activities that result in less electricity utilized for the same level of service; and load management activities to reduce demand during peak times. Though technology improvements are typically included in demand forecasts, it is often difficult to determine the extent of technology improvement present in such forecasts. For example, the IESO’s 10-year Outlook is heavily influenced by past trends and behaviours, and it is thus difficult to quantify the contribution made by technology improvements. Furthermore, energy savings due to CDM activities can be substantial but are even more difficult to quantify without detailed information on the programs, tools and standards. In order to assess electricity conservation potential in Ontario, ICF


114

1.22


1.2


1986C$/GJ

1.18

1.16

1.14

1.12

1.1

1.08 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026

4.10 Coal price forecast from NEB. Numerical data for the annual coal forecast is presented. Costs are expressed in terms of 1986 Canadian dollars.


115

Consulting has developed a methodology based on a combination of two approaches. The first approach is known as ‘experience-based’, and is based on a review of the effects of energy-conservation programs in various US jurisdictions as well as other efficiency potential analysis. The second approach, known as ‘accounting approach’, is based on an assessment of potential efficiency improvement contributions by various sectors, sub-sectors and end users (Ontario Power Authority, 2005). ICF used a combination of these two complementary approaches to help utilize the strengths of each. Though each approach has its advantages, they also have some inherent weaknesses. For instance, the experience-based approach uses US data that may not be applicable to Ontario. Using the two approaches described above, ICF considered four scenarios reflecting increasing levels of aggressiveness in energy conservation efforts. Energy Efficiency (EE) 25 refers to information-based programs with financial incentives of 25 % of incremental cost of new equipment installation. EE50 are common programs in which financial incentives equal 50 % of incremental cost of new equipment installation. EE100 are programs that involve intensive technical assistance and have financial incentives of 100 % of incremental cost. Finally, EE100 Plus Standards programs take into account a broad range of aggressive generic standards and also involve financial incentives of 100 % of incremental cost (ICF Consulting, 2005).

4.5

Planning model

From our literature review to date, we have found several authors who have used multi-period optimization methods for planning purposes. Iyer et al. (1998) have developed a multi-period mix-integer linear programming (MILP) model for the planning and scheduling of offshore oil field facilities. This mathematical model employs a general objective function that optimizes a selected economic indicator. Maravelias and Grossmann (2001) proposed a complex multi-period optimization model to address the challenge of planning for the production of a new product in highly regulated industries, such as pharmaceuticals and agrochemicals. The model uses a multi-period MILP model that maximizes the expected net present value of a multiperiod project. The model, although comprehensive, does not account for the lead time required for construction of new plants. Mo et al. (1991) developed a stochastic dynamic model for handling the uncertainties in generation expansion problems. The model makes it possible to identify the connection between investment decisions, time, construction periods and uncertainty. Hashim et al. (2005) developed a single-period deterministic MINLP optimization model aimed to predict a fleet-wide system configuration which simultaneously satisfies electricity demand and CO2 emission constraints at


116


minimum cost. The mathematical model developed was linearized using exact linearization techniques in order to overcome the inherited problems with solving non-linear models. Although the model developed by Hashim et al. (2005) is very comprehensive and complex, its single period mathematical structure does not allow the incorporation of multi-period factors such as construction lead time and fuel price fluctuations over time. In order to improve the optimization model and make it more realistic, the model developed by Hashim et al. (2005) must be extended to a multi-period domain. From the journal review conducted, no publication was found addressing the problem of finding the optimal strategy for energy planning with CO2 emission constraints and the option to implement carbon capture and storage. This chapter involves the development of a novel deterministic multi-period MINLP optimization model in order to realize the optimal mix of energy supply sources, while meeting CO2 emissions targets. The details of the formulation are given in the next sub-sections.

4.5.1 Model formulation The formulation developed here is a MILP model that is able to realize the optimal mix of energy supply sources which will meet current and future electricity demand and CO2 emission targets, and minimize the overall cost of electricity. The model presented is an extension and improvement to an earlier version that we have presented recently (Sirikitputtisak et al., 2009). The present model is initially a MINLP model that is then linearized using exact linearization methods. The linearization of the non-linear model is done with the aim of avoiding inherited computational difficulties encountered with large convex non-linear models. This linearization is able to lower the computation expense while retaining the consistency of the solution. For the illustrative case study that is presented in Section 4.6, the developed model was programmed and implemented in the GAMS (General Algebraic Modeling System) optimization package and solved using the CPLEX 10 solver. The indices, sets, variables and parameters used in the planning model are given in Table 4.5. Objective function The objective function of the planning model is to minimize the total discounted present value of the costs associated with meeting electricity demand while satisfying a CO2 reduction target over a specified planning horizon. The components associated with the objective function include: fixed and variable operating and maintenance cost, fuel cost, retrofit cost, capital cost for new power plants, carbon capture and storage cost and cost of purchasing carbon credits.



117

Table 4.5 Indices, sets, variables, and parameters used in the multi-period planning model Indices t i j l k

Time period (years) Boiler Fuel type (coal or natural gas) Load block (peak or base-load) Carbon capture technology

Sets F NF new new – cap

Fossil fueled power plants Non-fossil fueled New power plants New power plants with carbon capture

Parameters Fijt Vijt Cij Plt Ujt Gij Rit Sit (CCost)t T (CO2)ij Ekmax eikt b i Q i Dtl Btl r CLimitt

Fixed operating cost of boiler i using fuel j during period t ($/MW) Variable operating cost of boiler i using fuel j during period t ($/MWh) Capacity of boiler i using fuel j (MW) Duration of load block l during period t (hrs) Fuel cost for fuel j during period t ($/GJ) Heat rate of boiler i using fuel j (GJ/MWh) Cost associated with fuel-switching coal-fired boiler i during period t Capital cost of power plant i during period t Cost of carbon credits during period t ($/tonne of CO2) Time horizon (years) CO2 emission from boiler i using fuel j (tonne of CO2/MWh) Maximum supplemental energy required for kth capture technology Percent of CO2 captured from boiler i using carbon capture technology k during period t (%) Construction lead time for power station i (years) Cost of carbon capture and storage for boiler i ($/tonne of CO2) Electricity demand during period t for load l (MWh) Conservation and demand management during period t and load block l (MWh) Factor for transmission and distribution losses Specified CO2 limit during period t

Binary variables nit = 1 if power plant i is built during period t = 0 otherwise yit = 1 if power plant i is operational during period t = 0 otherwise xijt = 1 if coal-fired boiler i is operational while using fuel j during period t = 0 otherwise zijkt = 1 if the carbon capture technology k is used on boiler i, which uses fuel j, during period t. hit = 1 if coal-fired boiler i undergoes fuel-switching during period t = 0 otherwise Continuous variables Eijlt Power allocation from boiler i using fuel j for load block l during period t (MW) (Cre)t Carbon credits purchased during period t (tonne of CO2)


118


The objective function for the deterministic multi-period MINLP model is as follows: min f (i, j ,k ,l ,t ) = S SS FijtF CijF xijt + S S FitNF CiNF yitNF iŒF j t NF t   iŒ  Fixed O&M cost of existing power plants

F + S SSS VijtF Eijlt Plt + S SSVitNF EiltNF Plt iŒF j l t iŒNF l t   Variable O&M cost of existing power plants

F + S SSS U jt GijF Eijlt Plt + S S Rit hit iŒF j l t iŒF t    

Retrofit cost for fuel switching

Fuel cost for fossil fuel plants

+

S S

Sitnew Cinew nit new t i ŒP  

+

Capital cost for new power plant

+

S SS

Vitnew Eiltnew Plt new l t iŒ Œ P  Variable O&M cost of new power plant

S S

Fitnew Cinew yitnew new t i ŒP  Fixed O&M cost of new power plant

+

S SS

Uit Ginew Eiltnew Plt new l t i ŒP  Fuel cost for new power plant

F + S (Cre)t (CCost )t + S SSSS Qi (CO 2 )ij e ikt E ijlt zijkt Plt t i Œ F j k l t       Cost of purchasing CO2 emission credits

+

S

CO2 capture and storage cost for existing power plants

SS

Qi (CO 2 )i e ikt Eiltnew Plt new-cap l t i ŒP   CO2 capture and storage cost for new power plants

The construction of new power plants involves the use of postulated power plants that have a pre-assigned capacity and operational parameter. Energy production from these new hypothetical power plants can only occur if the optimizer has previously decided to build the new power plant. Several constraints, which are discussed in the next section, have been formulated in order to prevent the generation of electricity from new power plants that have not been constructed. It is important to note that no binary variable is associated with the cost of CCS for new power plants. For new power stations, the option to have a carbon capture system in place is dependent on which power station is chosen. For every hypothetical new power station there is an equivalent power station, with a similar capacity and operational parameters, that has an integrated CCS system. The optimizer considers the two corresponding power plants and will decide whether to build the power plant with a CCS or the one without CCS.



119

The non-linear term in the objective function comes from the equation that considers the CCS for existing power plants. This non-linearity is due to the cross-product of the binary variable zijkt (decision whether to put the kth carbon capture technology on the ith boiler using the jth fuel during time period t) and the continuous variable EFijtl (power allocation from ith fossil fuel boiler using the jth fuel type during period t and lth demand). Linearization of this term can be achieved by an exact linearization method. In order to achieve linearity, the following equation must be reformulated:

F S SSSS Qi (CO2 )ij e jkt Eijtl zijkt Plt

iŒF j k l t

[4.1]

The reformulation of this equation involves the introduction of a new continuous variable and several auxiliary constraints. The newly defined continuous variable aijtl is introduced into the equation and will replace the nonlinear expression.

F a ijktl = Eijtl zijkt

"i,"j,"t ,"k,"l

[4.2]

By substituting Equation 4.2 into Equation 4.1 the following equation is achieved;

S SSSS Qi (CO2 )ij e ikta ijktl Plt

iŒF j k l t

[4.3]

In order to ensure that this reformulation will yield the same results as its non-linear counterpart, additional constraints must be defined. The constraints proposed are as follows:

0 ≤ a ijtl ≤ CijF F Eijtl – CijF

max 

max 

[4.4]

"i,"j,"t ,"l

(1 – zijkt ) ≤ a ijtl ≤ CijF

max  

zijkt

"i,"j,"t ,"k,"l [4.5]

Model constraints The objective function that is discussed above is subject to the following constraints. Annual electricity demand The annual electricity generated from the entire fleet minus the supplemental energy required for potential carbon capture processes (Eikt) must be greater or equal to the annual electricity demand:


120


iŒF j

F S S Eijtl Plt + S EiltNF Plt + S Eilnew t Plt – SC S Eikt ≥ Dtl new iŒNF

iŒF

iŒP

k

"t,"l

[4.6]

Taking into account potential energy savings due to CDM strategies, Equation 4.6 becomes:

F S S Eijtl Plt + S EitlNF Plt + S Eitnew l P – SC S E ≥ D – B new

iŒF j

iŒNF

"t,"l

[4.7] tl

Œ

iŒP

lt

l

i F

k

ikt

tl

where, Btl is the forecasted annual energy savings (MWh) due to CDM strategies. The energy constraint in Equation 4.7 is enhanced further by considering the potential electricity losses incurred during the stages of transmission and distribution. Although the electricity losses in the transmission and distribution system are non-linear with transmitted power (Scherer, 1978), an approximation could be achieved by factorizing the power received with the dispatched power. Taking into account transmission losses, Equation 4.7 becomes; ˆ Ê F (1 – r )Á S S Eijtl Plt + S EitlNF Plt + S E new P – S S E ˜ ¯ Ë iŒF j iŒNF Œ C iŒP new

≥ D – B l "t ,"l

[4.8] itl

lt

ikt

k P i F losses. where r represents a factor for transmission and distribution tl

t

Capacity constraint for existing power stations In terms of the capacity allocation, the net power capacity (MW) of any power station cannot be exceeded. The maximum capacity constraints for existing fossil fuel and non-fossil fuel power plants are expressed in equations 4.9 and 4.10 respectively:

F S Eijlt ≤ CijF

max 

i

S EiltNF ≤ CiNF i

xijt

max 

yitNF

"iŒF,"t ,"j

[4.9]

"iŒNF,"t

[4.10]

Construction lead time and capacity constraint for new power stations The multi-period nature of the planning model requires the consideration of construction lead time for new power stations, which differs depending on the type of generating technology considered. For new power stations,



121

no power can be supplied to the grid unless the construction of the new power plant has been completed. To achieve this, Equation 4.11 has been formulated to ensure that during the construction phase of a new power plant, no electricity generating capacity is available. Furthermore, the constraint in Equation 4.11 also functions as a capacity constraint in which the net power capacity limit of a new power plant cannot be exceeded.

EØ(it≠)≠new £ CØi≠max(1 – nØit)

"iŒPnew, "t,"t¢ = 1, …, [t + (bi – 1)]

[4.11]

The binary variable nit determines whether power plant i should start construction during year t. Since the start of a construction project occurs only once for a given power plant i, the value of nit must be less or equal 1 for the sum of all time period t (Equation 4.12). The parameter bi represents the construction lead time for power station i:

S nit ≤ 1 "iŒP new [4.12] t A relationship between the binary variables nit and yitnew can be attained by formulating Equation 4.13. This equation ensures that if construction of a new power plant i occurs during year t, the plant is operational for all time periods t + bi:

(T – t ) –

T

S yitnew + nit ≤ bi

t =(t +bi )

"iŒ

new

, " = 1, . . ., ( – bi )

i P

t

T

[4.13]

An alternative approach that can be utilized in order to incorporate construction lead time into the model involves the use of a three indices matrix which restricts the maximum power output of a given power station. Each row in the matrix corresponds to a specific year of construction and each column refers to a ‘regular’ year. The non-zero values in the matrix specify the maximum capacity of the power station. A sample matrix for a hypothetical power plant P1 is illustrated in Fig. 4.11. This matrix can be used in conjunction t1

t2

t3

t4

t5

P1 t1

0

0

20

20

20

P1 t2

0

0

0

20

20

P1 t3

0

0

0

0

20

P1 t4

0

0

0

0

0

P1 t5

0

0

0

0

0

4.11 Sample matrix used in the construction lead time constraint. Each row in the matrix corresponds to a year during which construction would have commenced. The non-zero values in the matrix specify the maximum capacity of power plant P1.


122


with the binary variable yitc, which specifies the year in which construction should commence, in order to constrain the model from generating power from a power plant that has not yet been constructed. Equations 4.14 and 4.15 may be used as alternative to the mathematical constraints formulated in Equations 4.11 and 4.12.

Eitnew ≤ S yitc K itct

"i Œ P new ,"t

tc

[4.14]

where Kitct represents a three indices matrix:

S yitc ≤ 1 tc

"iŒP new

[4.15]

Capacity constraint on capture process The operation of any capture process requires the use of energy, either from the plant itself or from the grid. Equation 16 is formulated in order to ensure that the energy required for the kth carbon capture process is zero when no capture process is assigned to the ith coal-fired boiler. The parameter Ekmax represents the maximum supplementary energy required for the kth carbon capture process:

Eikt ≤ zikt Ekmax

"i Œ F c ,"k,"t

[4.16]

Fuel-selection and power plant shutdown Given that the model considers the option of fuel-switching existing coal-fired boilers with a less carbon intensive fuel, such as natural gas, a constraint must be formulated in order to restrict the use of two different fuel types on the same boiler. To achieve this goal, Equation 4.17 has been formulated. The binary variable xijt represents the fuel selection (coal or natural gas) for the ith fossil fuel boiler during time period t and could have a value of zero if the ith boiler is shut down:

S xijt ≤ 1 j

"iŒF,"t

[4.17]

The binary variables xijt and hit (decision whether to fuel-switch coal power plant i during time t) can be related by formulating the mathematical relation presented in Equation 4.18. T

(T – t + 1) – S xijt + hit ≥ 1 t =t

"t ,"i Œ ,

g

[4.18]

Since fuel-switching of a coal power boiler i can occur only once during the F "jbe Œ nincluded: time horizon T, the constraint in Equation 4.19 Œ must © Woodhead Publishing Limited, 2010


S hit ≤ 1

"iŒF

t

123

[4.19]

Selection of CO2 capture process In terms of CO2 capture process selection for a given boiler, a capture process can only be retrofitted if the boiler is operational. Equation 4.20 ensures that if an existing coal-fired boiler is shut down, no CO2 capture process can be put online:

S zikt ≤ S xijt k

j

"iŒF c ,"t

[4.20]

Furthermore, only one type of carbon capture technology can be used for a given boiler i during a time period t. The constraint formulated in Equation 4.21 can be used to prevent the use of two carbon capture technologies on the same boiler:

S zikt ≤ 1 k

"iŒF c ,"t

[4.21]

Carbon dioxide emission constraint The annual CO2 emissions produced as a result of electricity generation are limited by the constraint formulated in Equation 4.22. This constraint specifies that the annual CO2 emissions emitted by all existing and newly constructed boilers must be less than or equal to the specified annual CO2 target. It is assumed that the only power plants that generate CO2 emissions are those which use fossil fuel. Power stations that utilize non-fossil fuel, such as nuclear power plants, are assumed to have no CO2 emissions and therefore are not included in Equation 4.22. The CO2 constraint presented in Equation 4.22 also considers the potential of CO2 reduction by means of carbon credits. The CO2 emitted by the entire fleet for a particular year may be reduced by the purchase of CO2 credits for that year. È

(

)˘˚

S S ÍÊÁ S CO 2ijF EijtF Plt ˆ˜ 1 – S e ikt zikt ˙ ¯ Ëj k

i ŒF c l

+

Î

S S CO 2inew Eitlnew Plt – Cret ≤ Climitt

i ŒP new l

"t

[4.22]

Although the constraint discussed above only pertains to CO2 emissions, similar constraints can be formulated for other emissions, such as SO2 and NOx, by substituting the corresponding emission coefficients and specified


124


annual limits. Incorporating these constraints would allow the model to consider multiple pollutants and allow the emissions of these pollutants to be controlled. The drawback of including additional emission constraints within the model is that it increases the size of the model significantly and, from a computational point of view, would make the model difficult to solve. Therefore, the model presented in this chapter will only constrain the annual CO2 emissions and will not limit the emissions of any other pollutants.

4.6

Illustrative case study

The sections that follow outline a case study that was implemented using the model developed in the previous section. This case was selected in order to examine the economic and structural impact on Ontario’s electricity sector when forced to comply with a given CO2 emission limit. The case is based on a 14-year time horizon, starting in 2006 and ending in 2020. The case presents a future scenario in which CO2 emissions from the entire fleet must be 6 % below 1990 levels after 2011. To achieve this, annual CO2 emissions from the entire fleet must be less than 20 Mt per year after 2011. The emission limit specified is based on the Kyoto target of 6 % below 1990 levels. In order to address future electricity demand, several supply sources are considered in the case study. The technologies that are considered include nuclear, natural gas, coal, hydroelectric, pulverized coal combustion (PC), IGCC and NGCC power plants. Although additional power plant technologies exist, the scope of the case study discussed in this chapter only considers the above mentioned technologies as possible supply candidates. As discussed in the previous section, the optimization model takes into account several distinctive characteristics of each supply technology, such as economic, environmental and operational specifications, and determines the optimal mix of supply sources needed to satisfy each case study. The economic, environmental and operational parameters for each supply technology are presented in Section 4.6.1. Even though the time horizon for the case study is up to 2020 and investment decisions are being taken that would require consideration of the value of plants well beyond 2020. The full economic life for investment was considered through amortizing the costs within the modeling activity. The different costs were amortized with a 30-year lifetime and 15 % annual interest rate. The results of the case study are presented in Section 4.6.2. The model can be further improved by pursuing the following recommendations which we leave for future work. (i) The model can be reformulated from a deterministic model into a stochastic model. Reformulating the model into a stochastic multiperiod framework would allow handling of probabilistic parameters.



125

In reality, parameters such as electricity demand and fuel price fluctuations are random in nature and do not follow a deterministic path. However, the reformulation of the model into a stochastic framework may significantly increase the complexity of the model and inevitably complicate the computational time of the solution. (ii) The developed model currently does not take into account the geological location of the new power plants being built. In future work, the model can be modified in order to incorporate the geographical location of the new power plants. The location of the new stations may directly affect both transmission losses and local distribution strategies. (iii) The model could be improved by formulating an additional mathematical function that would allow the optimizer to design and map a complete pipeline network for the CCS system. In order to achieve this, the geological map of a region can be divided into a zoning matrix. The path of the pipeline network would be determined by several factors such as the cost of building a pipeline through that zone and the particular characteristics of the area. (iv) Currently the formulated model is designed as a single objective function model which attempts to minimize the cost of electricity while meeting electricity demand and a specified annual CO2 limit. The model can be reformulated into a multi-objective function that minimizes the total cost of electricity and CO2 emission of the entire fleet simultaneously. (v) The fixed and variable O&M costs of the power stations considered in this chapter were assumed to remain constant over time. In reality, the O&M costs of power stations increase over time due to ageing of the unit. In order to improve the results of the model, it is recommended that reliable time-dependent O&M costs be found and used. (vi) The model can be expanded by considering several additional pollutants such as NOx, SO2 and particulate matter (PM). Specifying emission limits of additional pollutants may increase the size of the model significantly and, hence, increase the overall computational time. (vii) A consideration may be given to modifying the model to include the option of importing and exporting electricity from neighboring regions. (viii) The model does not explicitly consider the impact of part load operation on plant performance.

4.6.1 Data for the case study This section provides several of the required input data parameters necessary to implement the model developed in Section 4.5.


126


Existing power plants Fossil-fueled power plants Ontario currently operates four coal-fired power plants with a combined capacity of approximately 6285 MW. The four coal power plants currently in operation are Lambton, Nanticoke, Atitokan, and Thunder Bay. In addition to the four coal-fired power plants, there is a dual-fueled oil and natural gas power plant referred to as the Lennox generating station. The economic and operational parameters for existing coal and NG and oil power plants in Ontario are presented in Table 4.6. This table also presents the cost associated with fuel-switching an existing coal-power plant to natural gas. The variable and fixed operating and maintenance (O&M) cost for the coal power plants were obtained from Ontario Ministry of Energy (2005). All other parameters were attained from Hashim (2006). Natural gas power plants In Ontario, there are 60 natural gas power plants in operation, but only 20 of these power plants are connected to Ontario’s electricity grid. The case study discussed in this chapter only takes into account the natural gas power stations which are connected to the grid. The operational and economic parameters for the existing natural gas power plants are presented in Table 4.7. The data outlined in Table 4.7 were obtained from Ontario Power Authority (2005). Nuclear power plants As mentioned earlier, most of the existing nuclear units in Ontario will reach the end of their service life by 2018. Consequently, Ontario’s nuclear units will need to be decommissioned, refurbished or replaced within the time horizon of the case study presented in this chapter. The end-of-service dates for all 20 nuclear units in Ontario were presented in Table 4.2. The refurbishment of a nuclear unit involves a significant amount of capital investment. The estimated refurbishment costs for Ontario’s nuclear power units are presented in Table 4.8. The case study will use the midpoint estimate when considering refurbishment cost of nuclear units (e.g., $3.5 billion for Pickering B). Furthermore, it is assumed that the lead time for the refurbishment of a single nuclear unit will be approximately two years. During the refurbishment process, the unit being refurbished will be shut down and consequently no electricity can be produced from that unit. The estimates for refurbishment costs were attained from Winfield et al. (2004). The decommissioning of nuclear units is a very complex and cost-intensive process. The work involved in a nuclear decommissioning project include


Power Gross plant capacity (MW)

Non-fuel variable O&M cost ($/MWh) coal ng

Fixed O&M Capacity Retrofit Heat rate CO2 emissions Cost of cost factor cost CCS ($/MW) (%) ($/MW) (GJ/MWh) (tCO2/MWh) ($/t CO2) coal ng coal ng coal ng

Elec. req. for CCS (MWh/t CO2) coal ng

Lambton 1948 Nanticoke 3820 Atikokan 211 Lennox 2100 Thunder Bay 306

2.45 2.25 5.11 0

0 0 0 0

36 804 32 715 74 631 n/a

0.5631 55.83579 0.558 55.38001 0.6138 212.7123 0.651 n/a

0.317 0.317 0.317 0.356

0.356 0.356 0.356 0.356

5.11

0

74 631 20 994 0.75

0.6138 216.2164

0.317

0.356

15 970 15 970 20 994 15 970

0.75 0.75 0.75 0.75

23 676.79 23 676.79 23 676.79 n/a

9.84 9.88 9.82 7.82

23 676.79 11.7

6.77 6.77 6.77 6.77

0.9278 0.93 1.023 0.651

6.77 1.023



Table 4.6 Economic and operational parameters of existing coal-power plants and cost associated with fuel-switching to natural gas. All costs are expressed in terms of 2005 Canadian dollars

127

128


the dismantling of the plant structure, decontamination of equipment, site remediation and long-term storage of nuclear waste. The estimated cost of decommissioning all 20 nuclear units in Ontario is 7.4 billion (Winfield et al., 2004). The case study assumes that all the nuclear power units in Ontario will be refurbished before the end-of-service year outlined in Table 4.2. The capacity profile for all nuclear units is shown in Table 4.9. The operational and economic parameters for Ontario’s existing nuclear units are presented in Table 4.10. The data outlined in Table 4.10 were obtained from Ontario Power Authority (2005). Hydroelectric Hydroelectric power plays a very important role in Ontario’s current energy mix. Approximately 26 % of Ontario’s installed capacity is composed of hydroelectric power. There are currently 108 hydroelectric stations within Ontario, but only 58 stations are directly connected to the electricity grid (Ontario Ministry of Energy, 2007) Ontario’s existing hydroelectric stations provide electricity for both baseload and peak-load demand. The total hydroelectric capacity available to serve base-load demand in Ontario is approximately 3424 MW. The hydroelectric capacity to meet intermediate and peak-load demand is approximately 3299 MW (Atomic Energy of Canada Limited, 2005). The hydroelectric stations that are designated for base-load electricity production are the Beck and Table 4.7 Operational and economic parameters for existing natural gas power plants. All costs are expressed in terms of 2005 Canadian dollars (Ontario Power Authority, 2005) Single cycle 3.42 Combined cycle 2.64 Cogeneration 2.74

5310 16 020 29 880

0.85 0.85 0.85

0.408 0.290 0.290

Table 4.8 Estimated refurbishment cost for nuclear units in Ontario. All costs are expressed in terms of 2005 Canadian dollars (Winfield et al., 2004 Station

Cost

Bruce 3 & 4 Bruce 1 & 2 Bruce B (5–8) Pickering A (1–4) Pickering B (5–8) Darlington (1–4) Total

$720 million $1.5–$2.5 million $3–$4 billion $3–$4 billion $3–$4 billion $3–$4 billion $14.2–$19.2 billion


Table 4.9 Capacity (MW) profile for all 20 nuclear units in Ontario from 2006–2020. The shaded area in the table represent the periods in which the unit was shut-down in order to undergo refurbishment

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

Pickering 1 515 Pickering 2 0 Pickering 3 0 Pickering 4 515 PICKERING A 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 515 1030

515 0 0 0 515

515 0 0 0 515

515 0 0 515 1030

515 0 0 515 1030

Pickering 5 Pickering 6 Pickering 7 Pickering 8 PICKERING B

516 516 516 516 2064

516 516 516 516 2064

516 516 516 516 2064

0 516 516 516 1548

0 0 516 516 1032

516 0 0 516 1032

516 516 0 0 1032

516 516 516 0 1548

516 516 516 516 2064

516 516 516 516 2064

516 516 516 516 2064

516 516 516 516 2064

516 516 516 516 2064

516 516 516 516 2064

516 516 516 516 2064

Bruce 1 Bruce 2 Bruce 3 Bruce 4 BRUCE A

0 0 750 750 1500

0 0 750 750 1500

0 0 750 750 1500

750 0 750 750 2250

750 750 750 750 3000

750 750 750 750 3000

750 750 750 750 3000

750 750 0 750 2250

750 750 0 750 2250

750 750 750 750 3000

750 750 750 750 3000

750 750 750 0 2250

750 750 750 0 2250

750 750 750 750 3000

750 750 750 750 3000

Bruce 5 Bruce 6 Bruce 8 Bruce 7 BRUCE B

785 820 785 785 3175

785 820 785 785 3175

785 820 785 785 3175

785 820 785 785 3175

785 0 785 785 2355

0 0 785 785 1570

0 820 0 785 1605

0 820 0 0 820

785 820 785 0 2390

785 820 785 785 3175

785 820 785 785 3175

785 820 785 785 3175

785 820 785 785 3175

785 820 785 785 3175

785 820 785 785 3175

Darlington 1 881 Darlington 2 881 Darlington 3 881 Darlington 4 881 DARLINGTON 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 881 881 881 3524

881 0 881 881 2643

881 0 881 881 2643

0 881 881 881 2643

0 881 0 0 881

881 881 0 0 1762

129

2007



2006

130

Developments and innovation in CCS technology Table 4.10 Operational and economic parameters for existing nuclear units. All costs are expressed in terms of 2005 Canadian dollars (Ontario Power Authority, 2005) Station

Variable O&M Cost ($/MWhd O&M

Bruce A Bruce B Darlington Pickering A Pickering B

1.42 1.42 1.42 1.42 1.42

105 105 105 105 105

720 720 720 720 720

0.9 0.9 0.9 0.9 0.9

Decew hydro stations in Niagara, and the R.H. Saunders hydro station near Cornwal (Ontario Ministry of Environment, 2006). The total capacity for the existing hydroelectric stations is 40 830 (MW) as obtained from Ontario Power Authority (2005). New power plants In order to meet future electricity demand, new power plants will need to be built. The case study will examine the use of the following supply sources to meet future demand: ∑ nuclear ∑ pulverized coal combustion (PC), ∑ integrated gasification combined cycle (IGCC) ∑ natural gas combined cycle (NGCC) ∑ long-term out-of-province hydroelectric imports Although additional supply sources exist, the scope of the case study presented in this chapter only considers the above-mentioned technologies as possible candidates. The outlined supply sources all have distinctive characteristics and may differ greatly based on environmental, economic and operational parameters. Some technologies offer lower capital and operating cost at high emission rates, while other supply options have higher associated costs but lower environmental impacts. In terms of capital cost, there are economies of scales that favor construction of large power stations over smaller ones. The capital and operational cost for building one large unit is often lower than if two smaller units, with the same total capacity, were built. The economic and operational parameters for the PC, IGCC and NGCC power units used in the case study are presented in Tables 4.11, 4.12, and 4.13, respectively. The data outlined in the tables were obtained from the Integrated Environmental Control Model (IECM) developed by Carnegie Mellon University and the Department of Engineering and Public Policy. The IECM is a computer modeling tool that performs a complete performance,


Table 4.11 Economic and operational parameters for pulverized coal (PC) power plants. This table presents data for single PC units and PC units that have been retrofitted with a MEA CO2 capture and storage (CCS) system. All costs are expressed in terms of 2005 Canadian dollars PC

PC

Gross capacity (MW) 457.7 526.5 Non-fuel variable O&M cost ($/MWh) 2.866172247 2.855230313 Fixed O&M cost ($/MW) 57 290.89899 52 839.93878 Capital cost ($/MW) 1 776 943 1 724 854 Capacity factor (%) 0.75 0.75 Heat rate (BTU/MWh) 9.59801646 9.59485123 CO2 emissions (tonne CO2/MWh) 0.875075515 0.874791439 NO2 emissions (tonne NO2/MWh) 3.100E-05 3.0946E-05 NO emissions (tonne NO/MWh) 0.0004 0.000383 SO2 emissions (tonne SO2/MWh) 0.001085 0.001084 Cost of CCS ($/tonne CO2) N/A N/A Construction lead time (years) 5 5 Project cash flow

PC with CCS

PC with CCS

PC with CCS

337.4 459.2 20.28099538 19.62101636 96 454.497 83296.71947 3 074 431 2 900 407 0.75 0.75 13.0196244 13.0090736 0.118789039 0.118692638 3.1485E-05 3.1471E-05 0.000520273 0.000519972 4.27E-07 4.27E-07 74.28 69.27 5 5 Year 0: 3.1 % (down payment) Year 1: 16.1 % Year 2: 30.8 % Year 3: 34.1 % Year 4: 15.9 %

491.7 19.46717892 80 439.44886 2 850 685 0.75 13.0090736 0.11867017 3.1457E-05 0.000519735 4.27E-07 62.88 5



131

132

IGCC

IGCC


Gross capacity (MW) 274.8 552.4 Non-fuel variable O&M cost ($/MWh) 1.24 1.23713470 Fixed O&M cost ($/MW) 97 748.609 72 521.9550 Capital cost ($/MW) 2 377 150 2 217 331 Capacity factor (%) 0.85 0.85 Heat rate (BTU/MWh) 11.173243 11.1099388 0.9860847 0.98092224 CO2 emissions (tonne CO2/MWh) NO2 emissions (tonne NO2/MWh) 4.3048E-06 4.2814E-06 NO emissions (tonne NO/MWh) 5.33E-05 5.31E-05 SO2 emissions (tonne SO2/MWh) 8.92E-05 8.88E-05 Cost of CCS ($/tonne CO2) N/A N/A Construction lead time (years) 5 5 Project cash flow

IGCC

IGCC with CCS

IGCC with CCS

IGCC with CCS

830.3

231.4

465.8

700.5

11.1619224 112 030.15 3 327 773 0.85 13.2728424 0.08939411 5.1942E-06 6.43677E-05 0.0001 15.41 5

10.4532119 74 494.85 3 327 773 0.85 13.2306394 0.08915146 5.1802E-06 6.41958E-05 0.000106 15.41 5

1.24402075 12.9655949 63 476.4947 145 191.72 2 140 382 3 562 173 0.85 0.85 11.0888373 13.3572484 0.97896847 0.08997359 4.2732E-06 5.2259E-06 5.29E-05 6.47653E-05 8.86E-05 0.000107 N/A 19.79 5 5 Year 0: 3.1 % (down payment) Year 1: 16.1 % Year 2: 30.8 % Year 3: 34.1 % Year 4: 15.9 %


Table 4.12 Economic and operational parameters for integrated gasification combined cycle (IGCC) power units. This table presents data for single IGCC units and IGCC units that have been retrofitted with a MEA CO2 capture and storage (CCS) system. All costs are expressed in terms of 2005 Canadian dollars

Table 4.13 Economic and operational parameters for natural gas combined cycle (NGCC) power units. This table presents data for single NGCC units and NGCC units that have been retrofitted with a MEA CO2 capture and storage (CCS) system. All costs are expressed in terms of 2005 Canadian dollars NGCC

NGCC

Gross capacity (MW) 253.3 506.5 Non-fuel variable O&M cost ($/MWh) 0 0 Fixed O&M cost ($/MW) 20 994.10 15 970.730 Capital cost ($/MW) 752 685 748 542 Capacity factor (%) 0.85 0.85 Heat rate (BTU/MWh) 7.177674 7.1776746 CO2 emissions (tonne CO2/MWh) 0.367458 0.3673516 NO2 emissions (tonne NO2/MWh) 4.6237E-06 4.6264E-06 NO emissions (tonne NO/MWh) 5.73E-05 5.73E-05 0 0 SO2 emissions (tonne SO2/MWh) Cost of CCS ($/tonne CO2) N/A N/A Construction lead time (years) 3 3 Project cash flow

NGCC

NGCC with CCS

759.8 216.1 0 8.57967927 14 284.6016 39 322.5348 746 411 1 319 981 0.85 0.85 7.17767461 8.41105719 0.36738713 0.04307133 4.6255E-06 4.0662E-06 5.73229E-05 6.71678E-05 0 0 N/A 71.53 3 3 Year 0: 0 % (down payment) Year 1: 50 % Year 2: 50 %

NGCC with CCS

NGCC with CCS

432.3 6.68023991 29 082.5243 1 220 539 0.85 8.41105719 0.04304038 4.0648E-06 6.71732E-05 0 46.98 3

648.4 5.93628639 27 003.3712 1 240 664 0.85 8.41105719 0.04305070 4.0658E-06 6.71714E-05 0 46.98 3



133

134


emissions and cost assessment of various fossil-fueled power plants of different capacity and operational specifications. The estimates for project cash flow during construction were obtained from Ayres et al. (2004). The costs of CCS presented in the tables were obtained from Hashim (2006). For the power stations that have an integrated CCS system, the cost associated with CCS is incorporated in the cost and operational parameters. The economic and operational parameters for nuclear power plants used in the case study are presented in Table 4.14. The data in the table were obtained from Ayres et al. (2004). The case study presented examines two types of nuclear technologies. The first nuclear unit is a twin ACR-700 nuclear reactor with a net capacity of 1406 MW. The second reactor is a twin CANDU 6 nuclear unit with a net capacity of 1346 MW. The cost and operational parameters of these two nuclear units are significantly different. The capital and fixed operating costs of the twin CANDU 6 units are generally higher than those of the ACR-700 reactors. A slight advantage that the CANDU 6 reactor possesses is the fact that it has a significantly lower variable operating cost than its predecessor, the ACR-700. In addition to the power plant technologies discussed above, the case study examines the potential long-term electricity supply from out-of-province hydroelectric imports. More specifically, the case study considers the potential use of the Ontario–Manitoba Interconnection (OMI) project as a long-term electricity Table 4.14 Economic and operational parameters for nuclear power units. All costs are expressed in terms of 2005 Canadian dollars

Twin ACR-700

Gross capacity (MW) Net capacity (MW) Variable O&M cost including fuel ($/MWh) Fixed O&M cost ($/MW) Capital cost ($/MW) Capacity factor (%) CO2 emissions (tonne CO2/MWh) NO2 emissions (tonne NO2/MWh) NO emissions (tonne NO/MWh) SO2 emissions (tonne SO2/MWh) Construction lead time (years) Project cash flow

Twin CANDU 6

1506 1406 4

1456 1346 2.3

11 160.00 13 270.00 2 414 170 3 057 050 0.9 0.9 0 0 0 0 0 0 0 0 8 8 Year 0: 3.1 % (down payment) Year 1: 8.0 % Year 2: 21.0 % Year 3: 27.1 % Year 4: 19.6 % Year 5: 12.0 % Year 6: 7.2 % Year 7: 5.1 %



135

supply source. The economic cost for the OMI project is outlined in Table 4.15 (Ontario Power Authority, 2005).

4.6.2 Case study results The case study, as discussed previously, presents a scenario in which Ontario’s electricity sector must comply with annual CO2 emissions 6 % below 1990 levels. This regulation comes into effect after 2011. There are no CO2 emission limits enforced between 2006 and 2011. After 2011, Ontario’s fleet must comply with an annual CO2 emission limit of 20 Mt (6 % below 1990 level). Moreover, this case study assumes that a carbon credit system or market is established in which individual power stations may purchase carbon credits at a cost. It is assumed that the technology for CCS in Ontario is available and can be implemented if needed. This case study also assumes that the phase-out of the coal power plants is not enforced by the policymakers. The existing nuclear power units will be refurbished based upon their estimated end-of-service dates. Furthermore, the assumption is made that all new or existing nuclear power plants are only used to meet base-load demand and are not utilized to satisfy peak demand. Wind power will be used for the purpose of meeting peak demand and not base-load demand. The results are presented in the following sub-sections. Fleet structure: new construction, fuel-switching, and CCS retrofit Table 4.16 illustrates the construction of new power stations. It includes the year in which the construction of new power plants started (represented by an ‘X’), as well as the years during which the unit is under construction (represented by the shaded area). It can be seen that three new NGCC power plants without CCS and two NGCC power plants with CCS system need to be built between 2006 and 2020. The net capacity of all new NGCC units without CCS and NGCC with CCS totals 4305.5 MW and 1080.7 MW, respectively. In addition to Table 4.15 Economic and operational parameters for long-term out-of-province hydroelectric imports. All costs are expressed in terms of 2005 Canadian dollars (Ontario Power Authority, 2005)

Gross capacity (MW)

Long-term 1250 out-of-province purchase

Variable O&M cost ($/MWh)

Fixed O&M cost ($/MW)

Capital cost ($/MW)

CO2 emissions (t CO2/MWh)

0

42 350

4 550 000

0


136

NGCC21 NGCC32 NGCC33 NGCC21 NGCC31 ACR-700

Net capacity (MW)

2006

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

New power plant without CCS 506.5 X 1519.6 X 2279.4 X New power plant with CCS 432.3 X 648.4 X Nuclear power plants 1406 X



Table 4.16 Construction of new power stations. The ‘X’ represents the year in which construction of the new generating unit started and the shaded area represents the years during which the unit is under construction. The years after the shaded areas assume the unit to be fully operational

Table 4.17 Existing coal power plants that have been fuel-switched to natural gas. The ‘X’ represents the year in which fuel-switching was implemented and the shaded area represents the years in which that specific coal power plant is using natural gas as its fuel source

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

Table 4.18 Existing coal power plants that have been retrofitted with a CCS system. The ‘X’ represents the year in which the coal power plant was retrofitted with a CCS system and the shaded area represents the years in which the plant is operating with a CCS system in place

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

Lambton X Nanticoke Atikokan Lennox Thunder Bay

2020



Lambton Nanticoke X Atikokan X Lennox Thunder Bay X

137

138


the three new NGCC power plants, one ACR-700 nuclear power plant is built in 2006. The net capacity of this nuclear power plant is 1406 MW. The ACR-700 power plant will be under construction for seven years and will be available for service after 2012. The model chose to keep all existing coal, natural gas, wind and hydroelectric power stations operational throughout the study period (2006–2020). Fuel-switching was implemented at Nanticoke, Atikokan and Thunder Bay during 2012, 2017 and 2017, respectively. Table 4.17 illustrates the coal power plants that have been fuel-switched to NG. The ‘X’ in the table represents the year in which fuel-switching should be implemented and the shaded area represents the years in which that particular power plant is operated using NG as its fuel. As shown in Table 4.18, a CCS system is to be retrofitted onto Lambton coal power plant in 2018. The ‘X’ represents the year in which CCS retrofit should be implemented and the shaded area represents the years in which that particular power plant is operating with CCS system in place. Power allocation and electricity production Total power (MW) allocated from each supply technology for the case study is presented in Fig. 4.12. The percent of power allocation based on generating technology for 2006, 2010, 2015 and 2020 is given in Fig. 4.13. From Fig. 4.12, it can be seen that the power allocated from renewable sources stays constant at a maximum level of 7902 MW. As discussed earlier, the maximum capacity is reached because no new renewable supply sources are considered in the case study. Existing NG and oil allocation reaches a maximum of 9151 MW in 2018. The increase in the maximum capacity of existing NG and oil plants is due to the fact that Nanticoke, Atikokan and Thunder Bay coal power stations undergo fuel-switching during 2012, 2017 and 2017 respectively. Power allocated from coal power plants ranges from 0 MW to 6269 MW, as shown in Fig. 4.12. Coal power is highly utilized throughout the timeframe in which no CO2 limits are imposed on Ontario’s electricity sector (2006–2011). In 2012, 2013 and 2017 there is no power production from any of the coal power plants. During these years, other electricity-generating technologies must be utilized in order to fill the energy gap created. Electricity-generating technologies used during these years include NGCC+CCS and existing NG and oil. As shown in Fig. 4.12, no long-term hydroelectric imports are realized. Furthermore, power allocated from nuclear plants ranges from 8820 MW in 2010 to 10 172 MW in 2016. The construction of the new ACR-700 nuclear power plant is completed at the end of 2012. After 2012, the newly constructed nuclear units are available for power supply to the grid.


40 000

30 000 25 000 20 000 15 000 10 000 5000

Renewable

0

2006 7902 Existing NG & oil 3614 Coal 6285 Hydro imports 0 Nuclear 8849 NGCC + CCS 0 IGCC + CCS 0 PC + CCS 0 NGCC 0 IGCC 0 PC 0

2007 7902 4098 6285 0 8849 0 0 0 0 0 0

2008 7902 4598 6285 0 8849 0 0 0 0 0 0

2009 7902 4770 6285 0 9144 0 0 0 51 0 0

2010 7902 4814 6285 0 8820 0 0 0 772 0 0

2011 7902 4814 6285 0 8227 0 0 0 1820 0 0

2012 7902 7975 0 0 8253 1081 0 0 4306 0 0

2013 7902 8081 0 0 8630 1081 0 0 4306 0 0

2014 7902 6164 1948 0 9941 236 0 0 4306 0 0

2015 2016 7902 7902 6289 6416 1948 1948 0 0 10 057 10172 471 719 0 0 0 0 4306 4306 0 0 0 0

2017 7902 8983 0 0 9695 1081 0 0 4306 0 0

2018 7902 9151 477 0 9695 1081 0 0 4306 0 0

2019 2020 7902 7902 8985 9151 1769 1140 0 0 9446 10 122 1081 1081 0 0 0 0 4306 4306 0 0 0 0



Power allocation (MW)

35 000

4.12 Total power allocated (MW) from each supply technology.

139

140

Developments and innovation in CCS technology 2010 NGCC 3 %

2006

Renewable 30 %

Renewable 27 %

Nuclear 33 % Coal 24 %

Nuclear 31 % Coal 22 %

Existing NG & oil 17 %

Existing NG & oil 13 % 2015

NGCC 14 %

Renewable 26 % Existing NG & oil 20 %

Coal 6 %

Nuclear 32 %

2020 NGCC + CCS 2 %

NGCC 13 % NGCC + CCS 3 %

Renewable 24 % Existing NG & oil 27 %

Nuclear 30 %

Coal 3 %

4.13 Total power allocation in terms of percentage for the years 2006, 2010, 2015, and 2020.

No IGCC+CCS, PC+CCS, IGCC and PC were constructed during the time horizon considered, and hence no power was allocated from these supply sources. From Fig. 4.12, power allocated from NGCC ranges from 51 MW in 2014 to 4306 MW in 2012. The power allocated by NGCC+CCS ranges from a minimum of 236 MW to maximum of 1081 MW. From Fig. 4.13, it can be seen that in 2006, 13 % of power is allocated from existing NG and oil, 24 % from coal, 30 % from renewable sources and 33 % from nuclear plants. As energy demand rises over the years, the percent of power allocated from nuclear sources remains relatively constant. This constant percentage from the total supply mix is maintained due to the construction of the new ACR-700 nuclear power plant which is scheduled to start producing electricity by 2013. While the percent power allocation from nuclear power plants remains constant, the percent power allocated by coal power plants decreases over time. In 2006, the percent of power allocation from coal power plants is 24 %. By 2020, the percent of power allocated from coal power plants amounts to only 3 % of the total mix. Figures 4.14 and 4.15 show the power (MW) allocated from each supply source to meet base-load and peak-load demands respectively. As shown in Fig. 4.14, base-load demand is predominantly satisfied with the use of renewable




141

16 000

Renewable base

14 000

Existing natural gas and oil base

12 000

Coal base Hydro imports base

10 000

New and existing nuclear base

8000

NGCC + CCS base

6000

IGCC + CCS base

4000

PC + CCS base

2000

NGCC base IGCC base PC base

2020

2016 2017 2018 2019

2013 2014 2015

2010 2011 2012

2007 2008 2009

2006

0

4.14 Power allocated to meet base-load demand (MW). 25 000 Renewable base Existing natural gas and oil base


20 000

Coal peak Hydro imports base

15 000

New and existing nuclear base NGCC + CCS base

10 000

IGCC + CCS base PC + CCS base NGCC base

5000

IGCC base PC base 2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

0

4.15 Power allocated to meet peak-load demand (MW).

and nuclear power. The reason these two technologies are primarily used to meet base-load demand is because they are cheap and clean technologies that may be used to generate electricity on a continuous basis. Coal power plants are utilized in order to help meet base-load demand during the time period in which no CO2 emission constraints are imposed on the electricity sector (2006–2011). After 2011, the year after which CO2 emissions are imposed, the use of coal technology is no longer utilized to meet base-load demand. The use of coal technology to meet base-load


142


demand is no longer chosen after 2011 because of the high CO2 emissions that coal power plants generate. NGCC, NGCC+CCS and existing NG and oil power plants are used to meet base-load demand after 2011. The utilization of these generating technologies in order to meet base-load demand is minimal due to the high fuel cost associated with the continuous operation of these plants. As illustrated in Fig. 4.16, peak-load demand is satisfied by various supply sources, including NGCC, NGCC+CCS, renewable, coal and existing NG and oil. Conversely to base-load demand, NGCC is highly utilized since it is operated only during periods of peak demand and is hence cost-effective. The model did not allocate any power from nuclear sources due to the assumption that nuclear units can only be used to meet base-load demand. The utilization of coal power in order to meet peak-load demand decreases significantly after 2011. The decrease in power allocation from coal power plants is due to the CO2 emission restrictions imposed on the electricity sector after this year. In order to reduce CO2 emissions to target levels, the model chose to reduce the use of coal power plants, and instead utilize less carbon-intensive fueled plants such as NGCC. During the time period between 2006 and 2020, the forecast peak-load demand increases steadily and new power plants must be brought online in order to satisfy this demand. The total electricity production (TWh) from each supply technology is presented in Fig. 4.16. The percent of electricity production from each supply source for 2006, 2010, 2015 and 2020 is given in Fig. 4.17. As shown in Figs 4.16 and 4.17, a significant amount of electricity production is generated from nuclear power. The electricity produced from nuclear power plants ranges from 72 TWh to 89 TWh. By 2015, nuclear power produces about 51 % of the electricity needed to meet Ontario’s demand. Electricity generated from NGCC starts in 2009, after the first NGCC power plant has been constructed. The electricity production from NGCC rages from 0.2 TWh in 2009 to 13 TWh in 2014. Electricity produced from the two NGCC+CSS is connected to the grid in 2012 and ranges from 1 TWh to 8 TWh. The electricity production from coal power plants decreases significantly after 2011. This decrease in electricity production from coal power plants is compensated by increasing the electricity production output of other supply technologies. The underlining reason why the model decided to decrease electricity production from coal power plants is the CO2 emission targets set after 2011. Electricity production from existing NG and oil, nuclear, NGCC and NGCC+CSS generally increases over time. The increase in electricity production from these supply technologies is due to a decline in electricity generation from coal power plants and an increase in electricity demand.


200 180

140 120 100 80 60 40 20

0 Renewable Existing NG & oil

Coal

Hydro imports

2007 45 14 22 0 78 0 0 0 0 0 0

2008 45 15 23 0 78 0 0 0 0 0 0

2009 45 16 22 0 80 0 0 0 0.2 0 0

2010 45 16 24 0 77 0 0 0 3 0 0

2011 44 16 28 0 72 0 0 0 6 0 0

2012 44 27 0 0 72 8 0 0 15 0 0

2013 44 29 0 0 76 7 0 0 14 0 0

4.16 Total electricity production (TWh) from all supply sources.

2014 44 19 6 0 87 1 0 0 13 0 0

2015 44 19 6 0 88 1 0 0 13 0 0

2016 44 19 6 0 89 2 0 0 13 0 0

2017 43 29 0 0 85 5 0 0 13 0 0

2018 43 28 1 0 85 6 0 0 13 0 0

2019 43 27 5 0 83 8 0 0 13 0 0

2020 43 28 3 0 89 4 0 0 12 0 0

143

Nuclear NGCC + CCS IGCC + CCS PC + CCS NGCC IGCC PC

2006 46 13 22 0 78 0 0 0 0 0 0



Electricity production (TWh)

160

144

Developments and innovation in CCS technology 2006

Renewable 29 %


Coal 14 %

2010 NGCC 1 %

2020 NGCC 8 %

Renewable 25 % Nuclear 51 %

Nuclear 47 %

Coal 15 %


2015

Existing NG & oil 11 % Coal 4 %

Renewable 27 %

Nuclear 49 %

NGCC + CCS 1 %

NGCC 7 % NGCC + CCS 2 %

Renewable 24 %

Existing NG & oil 16 % Coal 2 %

Nuclear 49 %

4.17 The percentage of electricity production from each supply source for years 2006, 2010, 2015, and 2020.


140 Renewable base

120

Existing natural gas and oil base Coal base

100

Hydro imports base 80

New and existing nuclear base NGCC + CCS base

60

IGCC + CCS base

40

PC + CCS base NGCC base

20

IGCC base PC base 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

2007 2008

2006

0

4.18 Electricity production generated to meet base-load demand (TWh).

Figures 4.18 and 4.19 illustrate the electricity production generated to meet base-load and peak-load, respectively. Electricity generated to meet base-load demand is predominantly produced from nuclear power plants. As shown in Fig. 4.18, the electricity generated from nuclear plants accounts




60

145

Renewable peak Existing natural gas and oil peak

50

Coal peak 40

Hydro imports peak New and existing nuclear peak

30

NGCC + CCS peak IGCC + CCS peak

20

PC + CCS peak NGCC peak

10

IGCC peak PC peak 2018 2019 2020

2017

2015 2016

2014

2011 2012 2013

2009 2010

2007 2008

2006

0

4.19 Electricity production generated to meet peak-load demand (TWh).

for more than half of Ontario’s base-load electricity demand. The remaining electricity demand is satisfied by renewable, coal, NGCC, NGCC+CCS and existing NG and oil supply technologies. After 2011, coal power plants are no longer used in order to meet Ontario’s base-load electricity production demand. Energy production for peak-load electricity demand is generated from various supply sources. Renewable, existing NG and oil and coal generate most of the electricity to meet peak-load demand from 2006–2011. After 2011, coal power plants play a less significant role in energy production for peak-load demand and other supply technologies, such as NGCC, become large contributors to electricity generation. Economic analysis The annual expenditure, presented in 2006 Canadian dollars, of the entire electricity sector is shown in Fig. 4.20. The annual expenses consist of: variable O&M for new and existing power stations, fixed O&M for new and existing power stations, capital cost associated with fuel-switching, cost of refurbishment of existing nuclear units, cost of CO2 credits, fuel costs and capital cost for construction of new power stations. The major factors that contribute to the cost of generating electricity are fuel costs, refurbishment costs for existing nuclear units and fixed O&M costs for existing power stations. The cost of fuel is the biggest contributor to the total annual cost of generating electricity. Fuel cost for the entire fleet rises steadily from 2006–2012. The increase in fuel prices is mainly


146

Captial for new power Capital for fuel switching

3

2006 $CAN billion


Fixed O&M of existing 3 Fixed O&M of new 2

Fuel Variable O&M of existing

2

Variable O&M of new 1

Capital and O&M of CCS CO2 credits

1

Nuclear refurbishment

0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

4.20 Annual expenditure of entire electricity sector. All costs are expressed in terms of 2006 Canadian dollars.


4


147

due to a rise in electricity demand, variability in natural gas prices and the utilization of new supply technologies which use natural gas as fuel. The cost of fuel drops during 2013–2015, but continues to rise steadily after 2015. The highest expenditure occurs in 2012, when $2.8 billion dollars are spent on fuel costs. The cost of nuclear refurbishment is particularly high from 2010–2014. During this time period, nine nuclear units are scheduled to be refurbished. The maximum expenditure for refurbishment of existing nuclear units occurs during 2011–2013. The fixed O&M cost for existing power stations remains relatively steady during the entire time horizon studied. The maximum expenditure for fixed O&M costs occurs during year 2015, at a cost of $1.59 billion. The capital expenditure for building new power plants is significantly high from 2006–2012. The high capital expenditure experienced during this time period is due to the construction of six new power plants (three NGCC, two NGCC+CCS and one nuclear). The construction of these new units requires a considerable amount of cash-flow during 2006–2012. The lowest contributor to the annual expenditure is the variable O&M cost for new power plants and the cost associated with CCS. The variable O&M cost associated with new power stations is not considered until 2009, since no new power plants have been built until this time. After 2009, a new NGCC-21 power station is brought online and the fixed O&M cost associated with operating this power plant is accounted for. The variable O&M cost for new power stations increases after 2009 as new power stations are built, and reaches a maximum of $108.1 million in 2020. The cost associated with CCS is considered in 2012, when the two new NGCC+CCS power plants are scheduled to start operation. The expenditure for CCS is not significantly high during 2011–2017, due to the low amount of CO2 captured and sequestered from the new NGCC power plants. The cost of CCS increases considerably in 2018, when the Lambton coal power plant is retrofitted with a CCS system. The CCS annual expenditure reaches a maximum of 161.7 million in 2020. The cost associated with carbon credits is zero since no carbon credits were purchased in any year. The breakdown of the total expenditure by sector for the entire study period (2006–2020) is presented in Fig. 4.21. The highest contributors to total expenditure are cost of fuel, fixed O&M costs for existing generating stations and nuclear refurbishment costs with a total price of $30.34, $21.33 and $11.97 billion, respectively. This is in line with the year-to-year results shown in Fig. 4.21. The costs associated with fuel-switching, capital and O&M costs of CCS retrofit and variable cost associated with new power plants account for the lowest parts of the total expenditure, with a total cost of $10, $54 and $54 million, respectively. The total expenditure for the entire study period is $79.10 billion.


148

Developments and innovation in CCS technology Capital and O&M of CCS, 0.54 Variable O&M of new, 0.54

Variable O&M of existing, 5.07

Captial for Nuclear new power, refurbishment, 8.17 11.97

Capital for fuel switching, 0.10

Fixed O&M of existing, 21.33

Fuel, 30.34

Total expenditure = $79.09674 billion

Fixed O&M of new, 1.03

4.21 Total expenditure for entire study period (2006–2020). All costs are expressed in terms of 2006 Canadian dollars ($billion).

Cost of electricity (cents/kWh)

4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

4.22 Annual cost of electricity (COE) for the entire study period (2006–2020). All costs are expressed in terms of 2006 Canadian dollars.

Figure 4.22 illustrates the annual COE. The values were obtained by dividing total annual expenditure with the annual electricity production. The average COE for the study period is 3.129 cents/kWh. The COE varies significantly throughout the span of the study period, ranging from a minimum of 2.252 cents/kWh in 2006, to a maximum of 4.12 cents/kWh in 2012. The variability associated with the COE in any particular year is dependent on all the factors that are considered in the total expenditure for that year. For instance, the high COE observed in 2012 is due to a large amount of money being spent on fuel, construction of new power plants and refurbishing nuclear



149

units, relative to how much electricity is generated. Similarly, the low COE experienced in 2006 is due to the low capital expenditure spent relative to the electricity generated. Carbon dioxide emissions Annual CO2 emissions from the entire fleet are presented in Fig. 4.23. The total CO2 emissions over the study period amount to 359 Mt. Note that an annual CO2 emissions limit of 20 Mt after 2011 was imposed. As can be seen, the annual CO2 emissions from the entire fleet are relatively high during the years in which no CO2 emission limits are imposed (2006–2011), and constant at 20 Mt after 2011 when an annual CO2 limit is imposed on the entire fleet. The CO2 emissions from the fleet increase from 2006–2011, reaching a peak of 36 Mt in 2011. Nanticoke coal-fired generating station is the single largest source of CO2 emissions during the years in which no emissions limits are imposed. In 2011, Nanticoke alone is responsible for 50.09 % of the CO2 emissions generated from the entire fleet. After 2011, the annual CO2 emissions from the entire fleet remain constant at 20 Mt due to the annual CO2 emissions imposed on the fleet.

4.7

Conclusions

This chapter presented a deterministic multi-period mixed-integer nonlinear programming (MINLP) model that is able to realize the optimal mix of energy supply sources that meet current and future electricity demand and CO2 emission targets, and lower the overall cost of electricity. Detailed data were acquired on various supply options that were used as parameters for the model. The cost and feasibility of using CO2 capture and storage in Ontario were examined. The model was applied to a case in which Ontario’s electricity sector must comply with annual CO2 emissions of 20 Mt (6 % below 1990 level) after 2011. The relative impacts studied were based on economic, structural and environmental affects. It should be noted that, although the case study was aimed at Ontario’s future energy supply mix, the model presented in this chapter has been formulated in a way that allows its application to other regions or countries. The model can also be reformulated from a deterministic model into a stochastic model. Reformulating the model into a stochastic multi-period framework would allow handling of probabilistic parameters. In reality, parameters such as electricity demand and fuel price fluctuations are random in nature and do not follow a deterministic path. However, the reformulation of the model into a stochastic framework may significantly increase the complexity of the model and inhevitably complicate the computational time of the solution. The developed model currently does


150

40


35

30

CO2 emission (Mt co2)


25

20

15

10

5

0

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2020

Lambton CO2 emission

Nanticoke CO2 emission

Atikokan CO2 emission

Existing NG emission

Lennox CO2 emission

Thunder Bay CO2 emission

New power plants CO2 emission

CO2 emission credit used

4.23 Annual CO2 emissions from entire fleet.


151

not take into account the geographical location of the new power plants being built. In future work, the model can be modified in order to incorporate the geographical location of the new power plants. The location of the new stations may directly affect both transmission losses and local distribution strategies. The model can also be expanded by considering several additional pollutants such as NOx, SO2 and particulate matter (PM).

4.8

References

AECL (2005) Maintaining Flexibility: Ontario’s Electricity Supply Gap and Implications for the Supply Mix. Toronto, ON: Atomic Energy of Canada Limited. Ayres M, MacRae M and Stogran M (2004) Levelised Unit Electricity Cost Comparison of Alternate Technologies for Baseload Generation in Ontario. Calgary, AB: Canadian Energy Research Institute. Benson S M and Surles T (2006) Carbon dioxide capture and storage: an overview with emphasis on capture and storage in deep geological formations. IEEE, 94, 1795–1805. CERI (2005) Electricity Generation Technologies: Performance and Cost Characteristics. Toronto, ON: CERI. Chui F, Elkamel A, Croiset E and Douglas P (2006) Long term electricity demand forecasting for power system planning using economic, demographic, and climatic variables. IEEE transactions on power systems. Environment Canada (2005) Canada’s Greenhouse Gas Inventory, 1990–2003, available at: http://www.ec.gc.ca/pdb/ghg/inventory_report/2003_report/sum_e.cfm (accessed December 2009). Hashim H (2006) An Optimal Fleet-Wide CO2 Mitigation Strategy for a Network of Power Plants. Waterloo, ON: University of Waterloo. Hashim H, Douglas P L, Elkamel A and Croiset E (2005) An optimization model for energy planning with CO2 emission considerations. Industrial & Engineering Chemistry Research, 44, 879–890. ICF Consulting (2005) Electricity Demand in Ontario – Assessing the Conservation and Demand Management (CDM) Potential. Toronto, ON: OPA. IESO (2006) Ontario Demand Forecast: 10-Year Outlook. Toronto, ON: Independent Electricity System Operator. Iyer R R, Grossmann I E, Vasantharajan S and Cullick A S (1998) Optimal planning and scheduling of offshore oil field infrastucture investment and operations. Industrial and Engineering Chemistry Research, 37, 1380–1397. Maravelias C T and Grossmann I E (2001) Simultaneous planning for new product development and batch manufacturing facilities. Industrial and Engineering Chemistry Research, 40, 6147–6164. Mo B, Hegge J and Wangenstee I (1991) Stochastic generation expansion planning by means of stochastic dynamic programming. IEEE Transaction of Power System, 6, 662–668. Murphy F H, Sen S and Soyster A L (1982) Electric utility capacity expansion planning with uncertain load forecasts. AIIE Transaction, 14, 52–29. Naini A, Walden T, Pinno K, Stogran M and Mutysheva D (2005) Electricity Generation Technologies: Performance and Cost Characteristics. Toronto, ON: Canadian Energy Research Institute.


152


Navigant Consulting (2005) Overview of the Portfolio Screening Model. Toronto, ON: OPA. Ontario Ministry of Energy (2005) Replacing Ontario’s Coal-Fired Electricity Generation. Toronto, ON: OME. Ontario Ministry of Energy (2007) Electricity. Retrieved March 15, 2007, from http:// www.energy.gov.on.ca/ Ontario Ministry of the Environment (2006, February) Ontario Government Ensures a Competitive Ontario Through Price Stability. Retrieved January 12, 2007, from Ontario Ministry of Environment: http://www.energy.gov.on.ca/index.cfm?fuseaction=english. news&back=yes&news_id=120&backgrounder_id=92 Ontario Power Authority (2005) Supply Mix Advice Report. Toronto, ON: OPA. Ontario Power Generation (2006) OPG. Retrieved January 14, 2007, from www.opg. com Ontario Power Generation (2007) Ontario Power Generation – Emissions In Perspective. Toronto, ON: OPG. (2007, March 15) Retrieved 2007, from Ontario Power Generation: http://www.opg.com/news/releases/ Rao A B and Rubin E S (2002) A technical, economic, and environmental assessment of amine-based CO2 capture technology for power plant greenhouse gas control. Environmental Science Technology, 36, 4467–4475. Scherer C (1978) Estimating electric power system marginal costs. The Economic Journal, 88(349), 152–154. Shafeen A, Croiset E, Douglas P L and Chatzis I (2004a) CO2 sequestration in Ontario, Canada. Part I: Storage. Energy Conversion Management, 45, 2645–2659. Shafeen A, Croiset E, Douglas P L and Chatzis I (2004b) CO2 sequestration in Ontario, Canada. Part II: Cost estimation. Energy Conversion Management, 45, 2645–2659. Sirikitputtisak T, Mirzaesmaeeli H, Douglas P L, Croiset E, Elkamel A and Gupta M (2009) A multi-period optimization model for energy planning with CO2 emission considerations, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 4339–4364. Winfield M S, Horne M, McClenaghan T and Peters R (2004) Power for the Future: Towards a Sustainable Electricity System for Ontario. Toronto, ON: Canadian Environmental Law Association.


5

Advanced absorption processes and technology for carbon dioxide (CO2) capture in power plants U. D e s i d e r i, Università degli Studi di Perugia, Italy Abstract: This chapter describes the chemical absorption technologies for carbon dioxide CO2, capture, covering the current state of the art and current commercial technologies, the chemical and physical principles on which they are based, recent advancements and innovations and, finally, the advantages and drawbacks, including the costs and the problems that might be expected in future applications. Key words: chemical absorption, carbon dioxide capture, amine-based CO 2 absorption.

5.1

Introduction

Every combustion process of fossil fuels, and several different industrial processes, produces flue gases, which are a mixture of carbon dioxide (CO2), water vapour and other gases. In order to avoid CO2 emissions to the atmosphere from these types of sources, CO2 has to be separated from the other vapours and gases, and absorption technologies for CO2 capture use liquid solvents for this purpose. Absorption processes with chemical solvents are currently the most used technology for post-combustion CO2 capture (Chakma, 1997; Desideri and Corbelli, 1998; Herzog, 1999; Desideri and Proietti, 2002; Singh et al., 2003; CO2NET, 2004; IEA, 2004; Wilson et al., 2004; Yokoyama, 2004; Bailey and Feron, 2005; IPCC, 2005). This is because they are the most efficient systems and have the lowest costs compared to other post-combustion capture processes, and they have reached the commercial stage for CO2 separation from natural gas and for CO2 production as a technical gas from coal combustion and gasification (Singh et al., 2003; Iijima and Takashima, 2004; Rao et al., 2006; Rubin et al., 2007; Abadie and Chamorro, 2008; Romeo et al., 2008). However, this technology has not yet reached the commercial stage for a full-scale power plant of 400–500 MW size. Chemical absorption processes are applicable to gas streams with high or low pressure, but which have a low CO2 partial pressure, and they use the reversible nature of the chemical reactions, which are affected by temperature and pressure changes. The heat of absorption of CO2 is generally between 50 and 80 kJ/mole CO2 and, in order to reuse the solvent, chemical absorption systems include 155 © Woodhead Publishing Limited, 2010

156


a regeneration stage where the CO2 is desorbed from the solvent at high temperature (100–140 °C) and at moderate pressures (approx. 1 bar). The regeneration requires thermal energy to heat up the solvent. Power plants and other industrial processes that produce flue gases with CO2 are generally at atmospheric pressure with low CO2 partial pressure, and the flue gases contain oxygen, nitrogen and other impurities. Therefore chemical solvents have to be carefully selected to have a low desorption energy requirement and a high capacity to absorb CO2. In addition, they have to have stable chemical composition to maintain solvent absorption capacity and to limit solvent make-up and disposal. The most widely used solvents are water solutions of amines and alkaline salt solutions. Table 5.1 reports the main characteristics of the commonest chemical solvents for CO2 absorption. Monoethanolamine (MEA) is the stateof-the-art solvent, but many studies have proposed other amines, mixtures of amines and other solvents to reduce energy consumption for regeneration and to reduce plant size.

5.2

Absorption processes

MEA has been the solvent most commonly used in chemical plants producing high-purity CO2 from the flue gases of power plants, industrial boilers and furnaces up to 1200 tonnes per day. Most plants using MEA are based on two technologies developed during the 1970s and 1980s: Kerr–McGee/ABB Lummus Crest (Barchas and Davies, 1992; Chapel et al., 1999; Herzog, 1999; CO2NET, 2004; Bailey and Feron, 2005; IPCC, 2005) and FluorDaniel/ Dow Chemical (Sander and Mariz, 1992; Chapel et al., 1999; Herzog, 1999; CO2NET, 2004; Bailey and Feron, 2005; IPCC, 2005). The third and more recent commercial technology currently available was developed in the 1990s by the Kansai Electric Power Company with Mitsubishi Heavy Industries (Mimura et al., 1995; Iijima and Takashima, 2004; CO2NET, 2004; Yokoyama, 2004; Bailey and Feron, 2005; IPCC, 2005). The most important benefits of this technology are the low heat requirements for regeneration, low amine losses and low amine degradation due to the use of inhibitors and additives. The Kerr–McGee/ABB Lummus amine technology can operate with boilers or cogeneration systems that fire fuels ranging from natural gas to high-sulphur coal and coke. The solvent is an aqueous solution of MEA with 15–20 % concentration by weight. The process tolerates oxygen in the flue gas as well as a limited amount of sulphur dioxide. The first Kerr–McGee unit was started in 1978 in Trona (California) with a production of 800 t/day of CO2 recovered from flue gases from boilers fired with natural gas, coal and coke. Since 1991 three more units have been licensed using this technology:


Solvent

Monoethanolamine (MEA)

Diglycolamine (DGA)

Diethanolamine (DEA)

Diisopropanolamine Methyldiethanolamine (DIPA) (MDEA)

Concentration (% mass)

< 30

< 60

< 40

< 40

< 50

Solvent loading (mole/mole)

0.3

0.35

0.30–0.70

0.45

0.45

Heat of regeneration (MJ/kg CO2)

2.0

2.0

1.5

1.5

1.3

Chemical formula

C2H7NO

C4H11NO2

C4H11NO2

C6H15NO2

C5H13NO2

Advanced absorption processes and technology


Table 5.1 Characteristics of chemical solvents for CO2 absorption (Bailey and Feron, 2005)

157

158


1. Applied Energy Systems, Poteau, Oklahoma. A 300 MW coal-fired cogeneration plant incorporates a 200 t/day food-grade liquid CO 2 unit as the steam host. 2. Soda Ash Botswana, Pty Ltd, Sue Pan, Botswana. This soda ash facility, incorporates a 300 t/day CO2 unit, 3. AES Corporation has built a second food-grade CO2 plant in Warrior Run, Maryland with a 150 short tons/day liquid CO2 plant, using flue gas from coal-fired circulating fluidized bed (CFB) boilers. In late 1990, Kerr–McGee and Lummus Global concluded a joint licensing agreement whereby Lummus gained worldwide exclusive marketing rights to Kerr–McGee’s CO2 recovery technology and became responsible for marketing and basic engineering. Kerr–McGee maintains a continuing role in technology transfer, process improvement, quality control of new designs, operator training and licensing. The Fluor Daniel Econamine process was acquired by Fluor Daniel, Inc. from Dow Chemical in 1989; it is also a MEA-based process with a 30 %wt aqueous solution, developed for enhanced oil recovery purposes. It can process flue gases with oxygen, NOx and SOx after SO2 removal and uses an inhibitor to avoid carbon steel corrosion due to the higher amine concentration. However, it cannot be used with reducing gases containing CO, H2 and H2S and less than 1 vol% oxygen. The Fluor Daniel technology has been used in more than 20 plants worldwide – and in Lubbock, Texas, with a capacity of 1200 t/day of CO2 is currently the largest plant of this kind. The process can recover 85–95 % of the CO2 present in the flue gases and can generate 99.95 % pure CO2 for use in food industry, urea production and enhanced oil recovery. Most of the commercial plants use natural gas as fuel, but a few use coal and heavy fuels. Some pilot plants have been installed as demonstration units in power plants in Canada and Japan. The Mitsubishi–Kansai technology offers as its most important benefits low heat requirements for regeneration, low amine losses and low amine degradation due to the use of inhibitors and additives. It based on sterically hindered amines giving three solvents, KS-1, KS-2 and KS-3, which are proprietary. KS-1 has a lower flowrate than in the Econamine process, due to the higher CO2 loading, lower regeneration temperature (110 °C) and lower heat of regeneration. KS-1 is also less corrosive for carbon steel. The first commercial plant was built in Malaysia in 1999 and produces 200 t/day of CO2. Table 5.2 shows a list of CO2 recovery commercial plants installed worldwide and Table 5.3 gives a comparison of their performance. Figure 5.1 shows the historical development of CO2 recovery technologies (Yokoyama, 2004).



Licenser Technology Flue gas source

Capacity [t/day]

Location and date

Use

Dow Chemical KerrMcGee ABB/Lummus Crest KerrMcGee ABB/Lummus Crest KerrMcGee ABB/Lummus Crest Fluor Daniel MHI/KEPCO Fluor Daniel Dow Chemical Fluor Daniel Dow Chemical KerrMcGee ABB/Lummus Crest

1100

Lubbock, Texas, 1982

EOR

Gas/Spec FT-1

Natural gas

CO2 recovery

Coal, coke and natural gas

800

Trona, California, 1978

Soda

CO2 recovery

Coal

300

Sua Pan, Botswana, 1991

Soda

CO2 recovery Econamine KS-1 Econamine Gas/Spec FT-1 Econamine Gas/Spec FT-1

Coal Natural gas Steam reforming and flue gas Natural gas, oil, coal NH3 reformer Natural gas Natural gas

200 320 160 165 150 90 2 ¥ 60

Poteau, Oklahoma, 1991 Bellingham, Massachusetts, 1991 Kedah Darul Aman, Malaysia, 1999 Chiba, Japan, 1994 Uttar Pradesh, India, 1988 Rio de Janeiro, Brasil, 1997 Altona, Australia, 1985

Food Food Food Food Urea Food Food

CO2 recovery

Coal

Warrior run, Maryland, 2000

Food

6


Table 5.2 CO2 recovery commercial plants (Herzog, 1999; CO2NET, 2004)

159

160


Table 5.3 Comparison of amine-based technology (CO2NET, 2004)

KerrMcGee/ABB Lummus Crest

Fluor Daniel Econamine

Steam for solvent (GJ/tCO2) 5–6.5 4.2 Solvent flow rate (m3/tCO2) 25 17 Electricity for fans and pumps 100–300 40–110 (kWh/tCO2) 165 Cooling water (m3/tCO2) 75–150 Solvent consumption (kg/tCO2) 0.45 1.5–2.0

1990

1995

2000

MHI/KEPCO KS-1 3.2 11 11 150 0.35

2003

* DOW Co.: Developed and commercialized energy saving technology for MEA absorbent in 1980s (the technology is now owned by Fluor Daniel Inc.) * Kerr Mcgee Co.: Commercialized technology as coal flue gas-compatible technology using MEA absorbent in 1980s (the technology is now owned by ABB Lummus Crest Inc.) Commercialization  * MHI developed and improved the chemical absorption method jointly with the Kansai Electric Power Co., Inc. Various companies in Japan endeavor to * RITE started research and develop CO2 recovery technologies. development of chemical absorption method and * Tokyo Electric Power Co., Inc./Hitachi: membrane separation method. Chemical absorption method * European countries, the USA., * Tokyo Electric Power Co., Inc./MHI: PTSA and Canada started research * Tohoku Electric Power Co., Inc./MHI: and development of CO2 PSA recovery methods. * Hokuriku Electric Power Co., Inc./CCEC: * CCP: Compared various Fluidized-bed PSA systems with each other. * Electric Power Development Co./IHI: O2 * Regina University: Chemical combustion method absorption method * Central Research Institute of Electric * University of Texas: Power Industry: Chemical absorption Chemical absorption method method * Kvaerner: Research and development of Membrane/Amine method

5.1 Historical development of CO2 recovery technologies (Yokoyama, 2004).

The MHI/KEPCO and Econamine processes are being developed for power plant application, by scaling up the commercial systems already in operation and by resolving a number of issues concerning flue gas purification prior to carbon recovery and energy requirements. These are major issues because in the currently operational commercial plants the product of the process is CO2 itself and therefore the cost of CO2 recovery is recouped by the use of the gas in food, EOR and other chemical industries. Recovering CO2 from the flue gases of a power plant, however, can be justified only for environmental reasons and the cost of capture represents a pure cost to be added to the cost of electricity. Therefore all the energy penalties for solvent



161

regeneration and all the capital and operational costs represent a significant drawback in the application of this technology. The following requirements are of prime importance in the development of chemical absorption technologies for power plants: ∑ Regeneration energy: the lowest possible to reduce energy consumption as steam is extracted from the power plant. ∑ Chemical stability at the operating temperature and pressure and low degradation with time. ∑ Low volatility to minimize evaporation losses. ∑ Oxygen should be tolerated to enable the use of carbon steel instead of more expensive materials. ∑ SOx should be eliminated because they form corrosive salts with MEA that cannot be recovered. Flue gas desulphurization should be enhanced in this type of plant. ∑ Fly ash and soot should also be removed because they may cause foaming in the absorber, scaling, plugging, erosion and solvent degradation. ∑ NOx have not caused problems in Econamine units, but they have done in other systems. They may produce salts with amines. ∑ Flue gas temperature should be reduced to 50 °C, but this can be very expensive in large power plants due to capital costs of heat exchangers and significant use of cooling water.

5.3

Description of the technology

The chemical process to capture and separate CO2 from other gases is completed in two steps: absorption of the CO2-rich gas in a lean aqueous solution of the solvent in a component called an absorber; and the separation of the CO2 from the enriched solvent in a component called a stripper. The absorption process is enhanced at low temperatures and high pressures, the stripping process at high temperatures and low pressures. Therefore the system requires a cooling source and a heat source and a compressor to increase the flue gas pressure before it enters the absorber (Yagi et al., 1992; Chakma, 1995, 1997; Desideri and Paolucci 1999; Simbeck, 2001; CO2NET, 2004; IEA, 2004; Bailey and Feron, 2005; IPCC, 2005; Mofarahi et al., 2008). Figure 5.2 shows the scheme of a CO2 recovery system based on chemical absorption and stripping. The flue gas (RICHGAS) enters the absorption column (ABSORBER) after being cooled to temperatures well below 100 °C and possibly close to 40 °C. The pressure of the flue gases can be close to but higher than the ambient pressure. Common values of flue gas temperatures are 180–200 °C from coal-and oil-fired power plants and down to 90 °C in natural gas-fired combined gas steam cycles. Flue gas pressure is always atmospheric and a fan is therefore necessary to raise the pressure.


162

Developments and innovation in CCS technology MKP-MEA

MIXER

MKP-H2O

MEA+H2O 25 leangas

B18

CO2

absorber 24 IN-STRIP richgas Bottom 26

B17

B23

B19

Cooler

B22 30

29

stripper

23

B15 21

22

Leansol

HEAT-EX

B1

B11

B14 27 31

20

B21

richpump

19

28 18

leanpump B12

17

5.2 Scheme of an absorption stripper system for CO2 capture (Desideri and Paolucci, 1999).

Flue gas in the absorber reacts with the aqueous amine solution producing a CO2-lean gas (LEANGAS) and carbon-rich amine solution exiting at the BOTTOM of the ABSORBER. The purified gas is vented to the atmosphere and the rich solution is pumped by RICHPUMP to a heat exchanger (HEATEX) where it recovers heat from the lean solution (LEANSOL) coming from the stripper. The rich solution is thus preheated in the lean/rich heat exchanger (HEATEX) and is introduced (IN-STRIP) in the STRIPPER where it is further heated by using a heat source raising the temperature to 120–140 °C using steam purposely generated or extracted from the power plant. At the higher temperature and lower pressure found in the stripper, the rich amine solution is regenerated by producing CO2 and the lean solution (LEANSOL) that is recirculated by the LEANPUMP to the absorber. The recovered CO2 in the stripper is collected and brought to the sequestration system. By using this arrangement, it is possible to capture more than 90 % of the CO2, which is contained in the flue gases. Large amounts of cooling water are necessary to cool the flue gases in the COOLER, and the lean solution after the heat exchanger before entering the absorber. The main energy consumption is due to the heat necessary for the stripping



163

process. Since the rich solution in the stripper has to be heated at 120–140 °C, the stripper is normally equipped with a steam reboiler that uses steam generated separately for the process or extracted from the steam turbine in a steam power plant. In gas–steam combined cycles, steam can also be extracted from the lower pressure levels in the heat recovery steam generator. Since the heat duty of the stripper can be between 3 and 5 GJ/tCO2, a significant fraction of the steam evolving in the turbine of a steam power plant is necessary and the penalties in produced energy and in efficiency can reach 20 % and approximately 10 percentage points, respectively. Both the absorber and the stripper are vessels with packed columns (Aroonwilas et al., 2001, 2003) that can be built with different fillings and with different heights depending on the CO2 concentration in the flue gas, the solvent concentration in the aqueous solution and the pressure and temperature of the vessels and the desired percentage of CO2 recovery (Desideri and Paolucci 1999; Mofarahi et al., 2008). To design or to predict the performance of the absorber and the stripper it is necessary to know the physical, thermal and transport properties of the gases and the liquids flowing in the system, the vapour–liquid equilibrium data and the rate and equilibrium data of the chemical reactions (Versteeg and Swaaji, 1988a,b; Versteeg et al., 1989, 1990; Kucka et al., 2003). The basis of all the calculations is the knowledge of the chemical reactions that happen in the absorption and stripping processes. The chemical absorption of CO2 with aqueous solutions of amines such as MEA is based on the following reactions (Aboudheir et al., 2003): ∑ Ionization of water: ∑

Dissociation of dissolved CO2 through carbonic acid:

∑

HCO–3 + H2O ´ CO32– + H3O+

Zwitterion formation from MEA and CO2 reaction:

∑

CO2 + 2H2O ´ HCO–3 + H3O+

Dissociation of bicarbonate:

∑

2H2O ´ OH– + H3O+

CO2 + RNH2 ´ RNH+2 COO–

Carbamate formation by deprotonation of the zwitterion:

RNH+2 COO– + RNH2 ´ RNH+3 + RNHCOO– RNH+2 COO– + H2O ´ H3O+ + RNHCOO– RNH+2 COO– + OH– ´ H2O + RNHCOO– ∑

Carbamate reversion to bicarbonate:


164


RNHCOO– + H2O ´ RNH2 + HCO–3 ∑

Dissociation of protonated MEA:

RNH+3 + H2O ´ H3O+ + RNH2 ∑

Bicarbonate formation:

CO2 + OH– ´ HCO–3

For high concentration of bicarbonates and carbonates in the aqueous solution, two more reactions have to be considered: RNH+2 COO– + HCO–3 ´ H2CO3 + RNHCOO– RNH+2 COO– + CO32– ´ HCO–3 + RNHCOO– Extensive data concerning CO2 absorption in MEA aqueous solutions can be found in Aboudheir et al. (2003). The design of the absorption system involves the selection of some parameters that need an optimization in terms of a cost-benefits analysis (Chakma et al., 1995). A high partial pressure of CO2 in the absorber is beneficial to the absorption process, but raising the pressure of flue gases at the plant exit requires power for the blower, and these two competing factors have to be considered in the selection of the absorption pressure. Since the flue gas has a large concentration of N2, and CO2 is approximately 10 %, a unit increase of the total pressure only increases the CO2 pressure by one tenth. Therefore the energy for flue gas compression should be the minimum possible and the pressure in the absorber should be the lowest possible, by using a low-pressure drop tray design in the packed column. The solvent flow rate is another important factor in the design of both the absorber and the stripper. The higher the flow rate, the lower the number of trays in the absorber and the stripper. The higher the flow rate, the higher is the solvent cost and the greater the diameter of the absorber and the stripper. Therefore, the optimum flow rate can be determined by the balance of these two competing factors: cost of solvent and number of trays in the absorber and stripper. Increasing the MEA concentration is beneficial because it can increase the CO2 absorption in the absorber with a lower flow rate and a lower heat duty in the stripper. Optimal MEA concentrations in water are in the range of 20–30 % by weight, because higher concentrations could give rise to problems in the absorber and stripper material due to corrosion. Therefore, more expensive materials may be required and solvent degradation may entail a larger make up and consequently higher costs. The selection of the amine type is also a very important factor for the performance and cost of the capture system. MEA is the more reactive amine, and a 30 % amine solution allows the number of trays in the columns, and



165

so the solvent flow rate, to be minimised, thus reducing the overall costs. Using secondary and tertiary amines is generally more expensive in terms of capital and operational costs compared to MEA (Chakma et al., 1995). It must be noted that the energy required for regeneration is proportional to the sum of the heat of reaction and the latent heat of vaporization of the solvent (Chakma, 1997). By looking at the different solvents that can be used in CO2 capture systems (Table 5.4), it is possible to see that tertiary amines require less energy for regeneration. However, they have a very low absorption rate and thus a lower mass transfer rate and, ultimately, a larger equipment size for absorption and stripping will be required. For this reason, blending secondary and tertiary amines with primary ones or adding activators to MEA can reduce the heat of regeneration without significantly reducing the reaction rate. In order to increase CO2 loading in the amine aqueous solution and reduce the regeneration heat, sterically hindered amines, such as those used by Mitsubishi Heavy Industries in their CO2 recovery plants, have been tested and studied (Mimura et al., 1995, 1997). They have some very interesting properties capable of improving the performance of the system. Figure 5.3 shows that sterically hindered amines such as KS-1 have a higher CO2 absorption capacity than MEA (Mimura et al., 1995). This is mainly due to the fact that the chemical reactions do not produce the carbamate ion; this allows the CO2 loading to be increased because 1 mol of CO2 can react with 1 mol of amine rather than 0.5 mol. Figure 5.4 shows that the hindered amines have a heat of regeneration 10–15 % lower than is the case with MEA. Looking at equilibrium conditions (Fig. 5.4), it is possible to note that KS-1 absorbs more CO2 at a lower temperature and has a lower CO2 loading at a higher temperature than MEA, meaning that regeneration is easier for KS-1 (Maceiras et al., 2008). Even though there have been several studies concerning the utilization of mixtures of primary, secondary and tertiary amines and there have been studies concerning the utilization of activators mixed with amines, current commercial plants for CO2 recovery are all using either MEA or KS-1.

Table 5.4 Heat of reaction, latent heat of vaporization and reaction rates constants of different solvents (Chakma, 1997) Solvent Concentration CO2 loading Heat of Latent heat of Reaction rate (M) (mol CO2/mol) reaction vaporization @298 K (kJ/mol CO2) (kJ/kg) (mol/L s) MEA DEA TEA MDEA

5 (30 % wt) 3.5 (36 % wt) 3.35 (50 % wt) 4.28 (50 % wt)

0.4 0.4 0.5 0.5

72 65 62 53.2

826 670 535 550


7600 1500 16.8 9.2

166

Developments and innovation in CCS technology Unhindered amine Moderately and severely hindered amine

Equilibrium CO2 loading (mol CO2/mol amine)

1.0 GROUPA KS-1 DIPA

0.5

DEA

DGA MEA

MDEA 0 10.5

11.5 12.5 13.5 pH of solution (before absorption)

5.3 pH of solution and equilibrium CO2 loading (Mimura et al., 1995). 100

Partial pressure of CO2 (psia)

50.0

KS-1 120 °C

10.0 5.0

MEA 40 °C

1.0 0.5 MEA 120 °C 0.1

KS-1 40 °C

0.05 0.01 0.0

0.2 0.4 0.6 Mol CO2/Mol amine

0.8

5.4 CO2–H2O–amine phase equilibrium (Mimura et al., 1995).

5.4

Advancements in the technologies

Many new ideas have been proposed and tested with the aim of improving the performance of CO2 recovery systems, either with mixtures of known amines, or with activated amines, or with novel absorption substances. Mandal et al. (2001) have proposed mixing small amounts of MEA, which is a fast-reacting primary amine, with large concentrations of 2-amino-2methyl1-propanol (AMP), a sterically hindered amine, as a better solvent for CO2 capture than mixtures of methyldiethanolamine (MDEA) and MEA. Two



167

years later, Mandal et al. (2003) also proposed a blend of diethanolamine (DEA) and AMP, in similar proportions as MEA and AMP, as a potential solvent with even better characteristics than the one proposed earlier. Van Loo et al. (2007) compared different blends of MDEA with other amines, considered as accelerators of the process, by calculating the number of trays in the absorber required to obtain the same performance. They showed that the addition of MEA produces the largest reduction in the number of trays, whereas AMP has practically no effect. Figure 5.5 shows the number of trays for different activators of MDEA. They also showed that adding only 0.01 mol of MEA to an aqueous MDEA solution enabled a reduction in the number of trays from 40 to 29, and adding 0.025 % mol reduced the number to 25. Schubert et al. (2001) proposed immobilized activators on a solid surface wetted by an aqueous MDEA solution and compared this to a blend of MDEA and DEA in aqueous solution. They showed that the performance was similar and that the absorber could be built without packing but only with a reactive surface. The improvement is also due to a larger area of the vessel where the reaction can take place. Erga et al. (1995) proposed instead a MDEA aqueous solution aimed at having a very low energy of regeneration for flue gases with very low CO2 concentrations (3–6 % vol) as commonly found in power plants. To improve the performance, they considered and tested a blend with 50 % MDEA and 5 % piperazine that produced a reduction of 20 % in the reboiler duty of the stripper. 45 40 35

N_trays

30 25 20 15 10 5 0 no-acc.

AMP

DIPA

DEA

DGA

MMEA

MEA

5.5 Number of trays required for CO2 absorption in aqueous MDEA (40 % wt) solution with various activators (2.5 % wt) (DIPA = diisopropanolamine) (van Loo et al., 2007).


168


When the overall performance of the capture and the power plant system is taken into account, different results can be obtained and a study of the entire power plant–capture system has to be performed. Mofarahi et al. (2008) have shown that a very high concentration (over 60 %) diglycolamine (DGA) aqueous solution provides better economic results than the conventional 30 % MEA solution. Alie et al. (2005) have shown that the reboiler duty per mole of CO2 separated in the stripper increases when CO2 concentration in the flue gases is lower, thus implying that the operational cost of capture for a natural gasfired power plant would be higher than for a cement plant. A recent work by Ma’mun et al. (2007) has studied the absorption rate and CO2 loading of a number of new solvents in comparison with MEA and MDEA. The following amines and absorbents were tested: monoethanolamine (MEA), 2-(butylamino)ethanol (BEA), N-methyldiethanolamine (MDEA), 2-(methylamino)ethanol (MMEA), 2-(ethylamino)ethanol (EMEA), 2-(2aminoethyl-amino)ethanol (AEEA), piperazine (PZ), potassium salt of taurine (PT). All absorbents were used in aqueous solutions with 30 % wt and a temperature of 40 °C. Some blends were also tested. Figures 5.6 and 5.7 show the CO2 absorption rate vs CO2 loading of individual solvents and solutions of blends, respectively. Figure 5.8 shows the equilibrium partial pressure of CO2 with two solvents in both absorption and regeneration conditions. The results are quite interesting for AEEA, which shows a similar behaviour to MEA but has a lower vapour pressure and a higher cyclic capacity and maintains a high absorption power at high loading. r CO2, absorption rate ¥ 105 (mol L–1s–1)

50

40

5.0M 2.6M 2.9M 2.0M

MEA MDEA AEEA PT

2.5M 3.3M 4.0M 2.5M

BEA EMEA MMEA PZ

30

20

10

0 0.00

0.20 0.40 0.60 0.80 a, CO2 loading (mol CO2/mol amine)

1.00

5.6 Absorption rate of CO2 in amine-based solvents at 40 °C (Ma’mun et al., 2007).



r CO2, absorption rate ¥ 105 (mol L–1s–1)

50

2.6M 2.6M 2.6M 2.6M 5.0M

40

169

MDEA MDEA + 0.62M MEA MDEA + 0.62M AEEA MDEA + 0.62M PZ MEA

30

20

10

0 0.00

0.10

0.20 0.30 0.40 0.50 a, CO2 loading (mol CO2/mol amine)

0.60

5.7 Absorption rate of CO2 in amine blends at 40 °C (Ma’mun et al., 2007).

p CO2, CO2 partial pressure (kPa)

1000 100 10

2.9M 2.9M 5.0M 5.0M 5.0M

AEEA 40 °C (this work) AEEA 120 °C (this work) MEA 40 °C (Jou et al., 1995) MEA 120 °C (Jou et al., 1995) MEA 120 °C (this work)

1 0.1 0.01 0.001 0.001

0.01 0.1 1 a, CO2 loading (mol CO2/mol amine)

10

5.8 Equilibrium partial pressure of CO2 in 5.0 MMEA and 2.9 AEEA solutions (Ma’mun et al., 2007).

A good absorption capacity has also been demonstrated by ammonia (Diao et al., 2004), which has a loading capacity and absorption rate similar to MEA. However, the product of the absorption process is ammonium bicarbonate, which is a salt that cannot be easily regenerated, even though it may have some uses in agriculture. This solution could be interesting for small-scale CO2 recovery applications where the final storage of captured CO2 underground might not be feasible for economic reasons.


170


Other examples of current developments in new absorption compounds are: – Aqueous potassium carbonate with piperazine (Culliname and Rochelle, 2003). – Non-aqueous solvents which have the benefit of reduced energy for stripping (Leites, 1998). – Amino acid salt solutions which have a lower vapour pressure and a high stability with oxygen (Erga et al., 1995; Feron and ten Asbroek 2005). – Di-amines which have more than one amine group and are able to absorb more CO2 molecules, thus reducing the solvent flow rate (Aresta and Dibenedetto, 2003). – Ionic liquids, which have no vapour pressure and so no thermal energy requirement for the production of stripping steam (Baltus et al., 2005). However, if carbon capture technologies are to be used in power plants then the energy consumption associated with regeneration in the stripper will need to be reduced. This is certain to stimulate research aimed at finding new solvents in order to mitigate the cost penalty of introducing such systems in fossil-fired power plants.

5.5

Advantages and disadvantages

Capture of CO2 from the flue gases of power plants by means of chemical absorption systems has been studied not only from a technical but also from an economic point of view. The major technical issues are the scalability of current commercial plants and the flue gas clean-up required to avoid solvent degradation with time or other technical problems such as foaming, plugging and scaling. Despite these problems chemical absorption systems are the most proven for CO2 recovery and could be considered closer to applicability in power plants. However, the difference between the current use of solvent-based absorption systems for CO2 recovery, where CO2 is the main commercial product, and their use in power plants where CO2 is not a product but a gas to be disposed of for final storage, is that in the latter case there is no revenue to compensate for capture costs. This requires costs to be lowered as much as possible because they have a direct influence on electricity costs. Several studies have been published where CO2 capture cost was estimated, but there is a large degree of uncertainty and they are difficult to compare because they have been calculated in different time frames and economic conditions. However, it is interesting to compare previous results in order to understand the trends and expectations for the future. Table 5.5 shows



171

Table 5.5 Comparison of estimated capture costs Source

Year

Application

Cost of CO2 capture

Chakma et al. Herzog Herzog Simbeck Singh et al. Rubin et al. Rubin et al. Mofarahi et al. Romeo et al.

1995 1999 1999 2001 2003 2007 2007 2008 2008

Coal-fired Coal-fired NGCC Coal-fired Coal-fired + NGGT Coal-fired NGCC NGGT Coal-fired + NGGT

$45–122/t CO2 $34–40/t CO2 $47–61/t CO2 $33/t CO2 $55/t CO2 $29–51/t CO2 $37–74/t CO2 $47/t CO2 725-61/t CO2

some published results of CO2 capture costs since 1995. It is important to note that natural gas-fired gas–steam combined cycles (NGCC) and natural gas-fired gas turbines (NGGT) show the highest recovery costs due to the lower CO2 concentration in flue gases. This means that larger capture systems are required with larger flow rates and larger plant equipment (absorber and stripper), and capital costs for the CO 2 capture plant are therefore higher. Rao et al. (2006) have presented a comprehensive study based on expert opinions about amine-based CO2 capture technologies. The final estimates of the best and most optimistic judgements were an average cost reduction from a baseline of $47/t CO2 of 18 % (with a range from 8–30 %) and of 36 % (with a range from 21–38 %), respectively. Therefore the worst expectations of the cost of captured CO2 would be $43/t CO2 and the best would be $29/t CO2. According to working group 1 of the European technology platform for zero emission fossil fuel power plants (ZEP WG1, 2006), that is a group including the major companies, research institutions and stakeholders involved in the field of carbon capture and storage, the post-combustion CO2 capture costs for hard coal, natural gas and lignite should be 722.4, 749.7 and 718.2 /t CO2 respectively. Figure 5.9 (ZEP WG1, 2006) shows the range of cost of CO2 capture with different technologies. The commercial introduction of CO2 capture technology in power plants will still require important steps in research and development. Some of these are reported in Table 5.6 (ZEP WG1, 2006) on page 173. In all the cases and scenarios, the cost of a tonne of CO2 captured should not exceed the cost of CO2 allowance traded in the EU ETS platform or similar systems to exchange rights to emit carbon dioxide, or the eventual fine that those who have to comply with emission restrictions will have to pay in case of non-compliance with the regulations. Abadie and Chamorro (2008) have presented the current situation in comparison with the average price of carbon allowances in the phase 0 of the Kyoto protocol implementation in the EU.


172


100

Note: CO2 avoidance cost without transport and storage cost

Eur/t CO2

80

Pre-combustion Post-combustion Oxyfuel

60 Power plant and CCS technology improvement potential

40 20 0

Hard coal

Lignite

Natural gas

5.9 Expected cost for large-scale plants in operation by 2020.

5.6

Applications and future trends

ZEP WG1 (2008) includes a list of demonstration projects in the EU for different technologies and applications. The list contains 34 projects that are currently ongoing or have been announced (Table 5.7 on page 175). The programme is wide and encompasses all the technologies for CO2 capture. Most projects will feature post-combustion capture technologies, but many are not designed for full-scale capture, but only for partial capture, to make it possible to study different solvents, operational parameters and improvements in the technology. When these demonstration projects are completed and in operation, it will be possible to compare the different technologies and their operational problems much more effectively.

5.7

Conclusions

It is expected that CO2 capture using chemical absorption systems will be implemented as post-combustion CO2 recovery in power plants within the next few years, because it is one of the most mature and cost-effective technologies to separate CO2 from flue gases, even though the cost of chemical capture and storage is still much higher than the current cost of CO2 in the carbon emission trading system. The existing commercial applications for CO2 production have been in operation for decades and the major problems involved in the installation of such systems in power plants are known and well defined. There is still a strong need for research and development to reduce plant costs and the energy penalty for regeneration and to improve the lifetime of the absorbents and the plant.


Table 5.6 Summary of generic R&D areas that are of common interest for different capture technologies (ZEP WG1, 2006) Commonly used designations

Description

Postcombustion

Coal-fired boiler/NGCC

CO2 removal from flue gas after power production

Current and near future technologies

Amine scrubbing

∑ Solutions based on existing technology and chemical solvents. ∑ Energy demanding solvent regeneration and separation of lowconcentration CO2 in flue gas

Air separation/ Fuel makeup

Gas clean up Gas separation

Boiler–steam cycle/gas turbine

Environmental Whole system issues issues

Important, CO2 removal particularly from flue gas for solid fuel

In principle any

Solvent emissions, handling of solid wastes Solvent emissions, handling of solid wastes

Combustion Improving Chemical with oxygen clean up by solvents – enriched new schemes amines (MEA) air would improve postcombustion capture process; Integration with fuel drying/ preheating process

In principle any; but aim for highest efficiency (USC)

Steam cycle integration

Full-scale demonstration in 2015, coal and natural gas



System

173

Enabling/ emerging and new technologies

∑ Membranes and membrane contactors ∑ Advanced solvents and processes ∑ Adsorbents ∑ Antisublimation ∑ Carbonation/ calcination

Safe and efficient CO2 separation from flue gas enabled by: ∑ Flue gas recirculation ∑ Use of nextgeneration solvents for removing CO2 from flue gas ∑ Efficient contactors ∑ Antisublimation process

Air separation/ Fuel makeup

Gas clean up Gas separation

Boiler–steam cycle/gas turbine

Environmental Whole system issues issues

High2nd temperature generation gas clean up CO2 solvents ∑ Membranes incl contactors ∑ Frosting process ∑ Adsorbents and processes

Gas turbine development to allow increased amount of CO2 (recycle), explore back-pressure trade-off for process integration

Need for solvents not generating emissions and waste streams; processes not requiring chemicals

Fully integrated approach, reduced investment cost by simplifications/ standards, switching logic (antisublimation) and phase change problems

x

x

x

x

x

x

x

Natural gas-based power cycle technology

Sulphur removal at high temperature

Coal-based power cycle technology

x

3rd generation hightemperature CO2 solvents Solid sorbents


Description


Commonly used designations

174

Table 5.6 Continued System

Table 5.7 List of announced CCS projects in the EU (ZEP WG1, 2008) Project name

Partners/ Country Location Industry participants

New/ Plant size CO2 Start of retrofit [MW] [Mt/yr] operation

Pre- combustion Post- combustion Post- combustion Post- combustion Post- combustion Oxyfuel or postcombustion Oxyfuel

Lignite

N

650

3.43

Lignite, biomass Lignite

R

105

0.5

2015

R

660

3.48

2015

Hard coal R

600

3.58

2015

Hard coal R

470/310

1.8

2013

Hard coal R

560/400

3.35

2015

Gas

30

R

2015

2015

175

Post- Hard coal, R combustion petcoke Oxyfuel & Lignite N & R 250 (Oxy), 1.79 post- 250 (post) combustion Post- Hard coal N 500 (100 0.6 combustion captured) Post- Hard coal, R combustion petcoke Post- Hardcoal N 1600 8 combustion

2010



MARITSA Bulgaria Maritsa Power HODONIN CEZ CEZ Czech Hodonin, Power Republic SE LEDVICE CEZ CEZ Czech Ledvice, N Power Republic KALUNDBORG DONG DONG Energy Denmark Kalundborg Power AALBORG V.FALL Vattenfall Denmark Aalborg Power MERI PORI FORTUM Fortum, TVO Finland Meri Pori Power LACQ TOTAL Total, France Lacq plant Power ALSTOM, and Rousse Air Liquide field FLORANGE ARC.MIT ArcelorMittal France France Steel JANSCHWALDE V.FALL Vattenfall Germany Jänschwalde, Power Brandenburg WILHELMSHAVEN E.ON E.On CE Germany Wilhelmshaven Power EISENHUTTENSTADT ArcelorMittal Germany Eisenhüttenstadt Steel GREIFSWALD DONG DONG Energy Germany Greifswald, Power Mecklenburg

Capture Fuel technology type


Capture Fuel technology type



HUERTH RWE RWE Germany Huerth, North Power Pre- Lignite N 450 2.8 Rhine-Westfalia combustion ENEL CCS1 ENEL Italy Power Post- Hard coal R 242 1.5 combustion ENEL CCS2 ENEL Italy Power Oxyfuel Hard coal N 320 2.1 SALINE JONICHE SEI SEI (Rätia Italy Saline Joniche Power Post- Hard coal N 1320 3.94 Energie & (RC) combustion Partners) BARENDRECHT SHELL Shell Nether- Barendrecht, Chemicals, H2 Heavy 0.4 lands Pernis Refinery production oil EEMSHAVEN RWE RWE Power, Nether- Eemshaven Power Post- Hard coal R 40 0.2 BASF, Linde lands combustion ROTTERDAM E.ON E.On Benelux Nether- Maasvlakte, Power Post- Hard coal N 1070 (100 5.6 lands Rotterdam combustion captured) ROTTERDAM ENECO ENECO, Nether- Pistoolhaven, Power Post- Gas N 845 International lands Rotterdam combustion Power EEMSHAVEN NUON Nuon Nether- Eemshaven Power Pre- Hard coal, N 1200 4.14 lands combustion biomass ROTTERDAM CGEN CGEN NV Nether- Europoort Power Pre- Hard coal, N 450 2.5 lands Rotterdam combustion biomass ROTTERDAM ESSENT Essent Nether- Rotterdam Power Pre- Hard coal, N 1000 4 lands combustion biomass MONGSTAD STATOIL StatoilHydro, Norway Bergen Power, Post- Gas N 280 EE + 1.5 Gasnova refinery combustion 350 heat HAMMERFEST H.ENERGI Hammerfest, Norway Hammerfest Power Post- Gas N 100 Sargas, combustion Siemens

2014 2014 2016

2011 2015

2011

2013 2016 2016 2014


Project name

176

Table 5.7 Continued

2012 2014

2015 2015 2014 2015 2015

2016 2014



2015 2016

177

HUSNES TINFOS Tinfos, Sor- Norway Husnes Various Post- Hard coal N 400 2.5 Norge, combustion Eramet, Sargas KARSTO AKER Aker, Fluor, Norway Karsto Oil/gas Post- Gas R 420 1.2 Mitsubishi combustion MONGSTAD BKK BKK Norway Mongstad Power Post- Gas N 450 1.2 combustion or pre combustion HAUGESUND Haugaland Norway Haugesund Power Hard coal N 400/800 Kraft SIEKIERKI V.FALL Vattenfall Poland Warsaw Power Post- Hard coal N 480 2.87 combustion KEDZIERZYN PKE PKE/ZAK Poland Kedzierzyn Power, Pre- Hard coal N 250 EE + 3.4 Kozle/Slaskie Chemical combustion 500 heat BELCHATOW BOT PGE, ICPC, Poland Belchatow Power Post- Lignite N 858 5.1 CMI, PGI combustion (1/3 CCS) COMPOSTILLA ENDESA Endesa Spain Compostilla, Power Oxyfuel Various N 500 Leon CFB coals, biomass UNION FENOSA Union Fenosa Spain Power Post- Hard coal N 800 combustion (200 CCS) KINGSNORTH E.ON E.ON UK UK Kingsnorth, Power Post- Hard coal N 300 2 SE England combustion SCUNTHORPE CORUS CORUS UK Scunthorpe Steel Post- Hard coal, R combustion petcoke COCKENZIE SCOT.PWR Scottish UK Scotland Power Post- Hard coal N Power combustion FERRYBRIDGE S&S Scottish & UK Ferrybridge, Power Post- Hard coal R 500 Southern West Yorkshire combustion TILBURY RWE RWE nPower UK Tilbury, Thames Power Post- Hard coal N 1600 9.56 Estuary combustion



KILLINGHOLME E.ON E.ON UK UK Humberside, Power Lincolnshire HATFIELD P.FUEL PWR Powerfuel UK Hatfield, South Power Power Ltd Yorkshire TEESSIDE PROG.EN Centrica, UK Teesside Power Progressive Energy, Coastal Energy DRYM PROG.EN Progressive UK Onllwyn, Power Energy, BGS, South wales CO2STORE

Capture Fuel technology type Pre- combustion Pre- combustion Pre- combustion


Hard coal N

350

2.5

2016

Hard coal N

900

4.75

2012

Hard coal, N petcoke

800

4.22

2013

Pre- Hard coal N combustion

450

2.4


Project name

178

Table 5.7 Continued


179

This can be done by focusing on the following issues that are also listed in Table 5.6: – development of new absorbents or new mixtures of existing ones, and development of activators and inhibitors of corrosion; – study of the optimal integration with the power plant in terms of heat transferred from the cycle to the CO2 recovery system and vice versa; – study of packed columns to improve the efficiency of heat and mass transfer during absorption and stripping; – development of catalysts to enhance the absorption and stripping processes; – development of efficient flue gas clean-up systems to reach the desired purity for the capture system.

t en ym

it

lo

un

ep

n

rd

tio ra Re

ad

y

fo

st on m De

Co

n an cep d tue la bo l in ra ve to sti ry ga Pi lo te tio tp st s n la nt

Most of all, a technology transfer to build the first full-scale and demonstration units is necessary if all the issues shown in Fig. 5.10 are to be resolved (ZEP WG1, 2006).

Overall status Full process integration and optimization for power Component status Boiler and power process Extended desulphurization DeNOx process CO2 capture process Capture process optimization incl. new solvents and scale-up CO2 processing

5.10 Illustration of the maturity of the post-combustion technology. In the diagram, it is indicated that almost all major components are commercially available, but at another scale and not integrated and optimized for this purpose. The process also demands a very clean flue gas, which is not common in ordinary power plants.


180

5.8


References

Abadie L M and Chamorro J M (2008) ‘European CO2 prices and carbon capture investments’, Energy Economics, 30, 2992–3015. Aboudheir A, Tontiwachwuthikul P, Chakma A and Idem R (2003) ‘Kinetics of the reactive absorption of carbon dioxide in high CO2-loaded, concentrated aqueous monoethanolamine solutions’, Chemical Engineering Science, 58, 5195–5210. Alie C, Backham L, Croiset E and Douglas P L (2005) ‘Simulation of CO2 capture using MEA scrubbing: a flowsheet decomposition method’, Energy Conversion and Management, 46, 475–487. Aresta M and Dibenedetto A (2003) ‘Carbon dioxide absorption with aqueous potassium carbonate promoted by piperazine’, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Elsevier (Pergamon), Oxford, UK, Vol. II, 1599–1602. Aroonwilas A, Tontiwachwuthikul P and Chakma A (2001) ‘Effects of operating and design parameters on CO2 absorption in columns with structured packings’, Separation and Purification Technology, 24, 403–411. Aroonwilas A, Chakma A, Tontiwachwuthikul P and Veawab A (2003) ‘Mathematical modelling of mass-transfer and hydrodynamics in CO2 absorbers packed with structured packings’, Chemical Engineering Science, 58, 4037–4053. Bailey D W and Feron P H M (2005) ‘Post-combustion decarbonisation processes’, Oil & Gas Science and Technology – Rev. IFP, 60(3), 461–474. Baltus R E, Counce R M, Culbertson B H, Luo H, De Paoli D W, Dai S and Duckworth D C (2005) ‘Examination of the potential of ionic liquids for gas seprations’, Separation Science and Technology, 40, 525–541. Barchas R and Davis R (1992) ‘The KERR/McGEE/ABB LUMMUS CREST for the recovery of CO2 from stack gases’, Energy Conversion and Management, 33, 333–340. Chakma A (1995) ‘An energy efficient mixed solvent for the separation of CO2’, Energy Conversion and Management, 36(6–9), 427–430. Chakma A (1997) ‘CO2 capture processes – opportunities for improved energy efficiencies’, Energy Conversion and Management, 38, Suppl., S51–S56. Chakma A, Mehrotra A K and Nielsen B (1995) ‘Comparison of chemical solvents for mitigating CO2 emissions from coal-fired power plants’, Heat Recovery Systems & CHP, 15(2), 231–240. Chapel D G, Mariz C L and Ernest J (1999) ‘Recovery of CO2 from flue gases: commercial trends’, Canadian Society of Chemical Engineers Annual Meeting, 4–6 October, Saskatoon, SK, Canada. CO2NET (2004) Capturing and storing carbon dioxide: technical lessons learned, available at: http://www.co2net.com/infocentre/reports/technical_lessons_learned_final_report. pdf (accessed January 2010). Culliname J T and Rochelle G T (2003) ‘Carbon dioxide absorption with aqueous potassium carbonate promoted by piperazine’, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Elsevier (Pergamon), Oxford, UK, Vol. II, 1603–1606. Desideri U and Corbelli R (1998) ‘CO2 capture in a small size cogeneration plants: technical and economical considerations’, Energy Conversion and Management, 39(9) 857–867. Desideri U and Paolucci A (1999) ‘Performance modelling of a carbon dioxide removal system for power plants’, Energy Conversion and Management, 40, 1899–1915.



181

Desideri U and Proietti S (2002) ‘CO2 capture and removal system for a gas–steam combined cycle’, 2002 Proceedings of the ASME Advanced Energy Systems Division, American Society of Mechanical Engineers, New York, NY, 275–282. Diao Y F, Zheng X Y, He B S, Chen C H and Xu X C (2004) ‘Experimental study on capturing CO2 greenhouse gas by ammonia scrubbing’, Energy Conversion and Management, 45, 2283–2296. Erga O, Juliussen O and Lidal H (1995) ‘Carbon dioxide recovery by means of aqueous amines’, Energy Conversion and Management, 36(6–9), 387–392. Feron P H M and ten Asbroek N A M (2005) ‘New solvents based on amino-acid salts for CO2 capture from flue gases’, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 2, 1153–1158. Herzog H (1999) An introduction to CO2 separation and capture technologies, Massachusetts Institute of Technology Energy Laboratory, Cambridge, MA. IEA (2004) Improvement in power generation with post-combustion capture of CO2, Report number PH4/33, IEA Greenhouse Gas R&D, Programme, Cheltenham, UK. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, de Coninck H C, Loos M and Meyer L A (eds), Cambridge University Press, Cambridge, UK. Iijima M and Takashina T (2004) ‘A view of oil resources and the mitigation of CO2 emissions’, Mitsubishi Heavy Industries Technical Review, 41(4), 1–7. Jou F-Y, Mather A E and Otto F D (1995) ‘The solubility of CO2 in a 30 mass percent monoethanolamine solution,’ Can J. Chem. Eng., 73, 140–147. Kucka L, Muller I, Kenig E Y and Gorak A (2003) ‘On the modelling and simulation of sour gas absorption by aqueous amine solutions’, Chemical Engineering Science, 58, 3571–3578. Leites I L (1998) ‘The thermodynamics of CO2 solubility in mixtures monoethanolamine with organic solvents and water and commercial experience of energy saving gas purification’, Energy Conversion and Management, 39, 1665–1674. Maceiras R, Alvarez E and Angeles Cancela M (2008) ‘Effect of temperature on carbon dioxide absorption in monoethanolamine solutions’, Chemical Engineering Journal, 138, 295–300. Ma’mun S, Svendsen H F, Hoff K A and Juliussen O (2007) ‘Selection of new absorbents for carbon dioxide capture’, Energy Conversion and Management, 48, 251–258. Mandal B P, Guha M, Biswas A K and Bandyopadhyay S S (2001) ‘Removal of carbon dioxide by absorption in mixed amines: modelling of absorption in aqueous MDEA/ AMP and AMP/MEA solutions’, Chemical Engineering Science, 56, 6217–6224. Mandal B P, Biswas A K and Bandyopadhyay S S (2003) ‘Absorption of carbon dioxide into aqueous blends of 2-amino-2-methyl-1-propanol and diethanolamine’, Chemical Engineering Science, 58, 4137–4144. Mimura T, Shimojo S, Suda T, Iijima M and Mitsuoka S (1995) ‘Research and development on energy saving technology for flue gas carbon dioxide recovery and steam systems in power plant’, Energy Conversion and Management, 36(6–9), 397–400. Mimura T, Simayoshi H, Suda T, Iijima M and Mituoka S (1997) ‘Development of energy saving technology for flue gas carbon dioxide recovery in power plants by chemical absorption method and steam system’, Energy Conversion and Management, 38, Suppl., S57–S62. Mofarahi M, Khojasteh Y, Khaledi H and Farahnak A (2008) ‘Design of CO2 absorption plant for recovery of CO2 from flue gases’, Energy, 33, 1311–1319. © Woodhead Publishing Limited, 2010

182


Rao A B, Rubin E S, Keith D W and Granger Morgan M (2006) ‘Evaluation of potential cost reductions from improved amine-based CO2 capture systems’, Energy Policy, 34, 3765–3772. Romeo L M, Bolea I and Escosa J M (2008) ‘Integration of power plant and amine scrubbing to reduce CO2 capture costs’, Applied Thermal Engineering, 28, 1039–1046. Rubin E S, Chen C and Rao A B (2007) ‘Cost and performance of fossil fuel power plants with CO2 capture and storage’, Energy Policy, 35, 4444–4454. Sander M T and Mariz C L (1992) ‘The Fluor Daniel Econamine FG process: past experience and present day focus’, Energy Conversion and Management, 33, 341–348. Schubert S, Grunewald M and Agar D W (2001) ‘Enhancement of carbon dioxide absorption into aqueous methyldiethanolamine using immobilised activators’, Chemical Engineering Science, 56, 6211–6216. Simbeck D R (2001) ‘CO2 mitigation economics for existing coal-fired power plants’, First National Conference on Carbon Sequestration, 14–17 May, Washington, DC. Singh D, Croiset E, Douglas P L and Douglas M A (2003) ‘Techno-economic study of CO2 capture from an existing coal-fired power plant: MEA scrubbing vs O2/CO2 recycle combustion’, Energy Conversion and Management, 44, 3073–3091. van Loo S, van Elk E P and Versteeg G F (2007) ‘The removal of carbon dioxide with activated solutions of methyl-diethanol-amine’, Journal of Petroleum Science and Engineering, 55, 135–145. Versteeg G F and van Swaaij W P M (1988a) ‘On the kinetics between CO2 and alkanolamines both in aqueous and non-aqueous solutions – I. Primary and secondary amines’, Chemical Engineering Science, 43, 573–585. Versteeg G F and van Swaaij W P M (1988b) ‘On the kinetics between CO2 and alkanolamines both in aqueous and non-aqueous solutions – II. Tertiary amines’, Chemical Engineering Science, 43, 587–591. Versteeg G F, Kuipers J A M, Beckum F P H and van Swaaij W P M (1989) ‘Mass transfer with complex chemical reactions – I. Single reversible chemical reaction’, Chemical Engineering Science, 44, 2292–2310. Versteeg G F, Kuipers J A M, Beckum F P H and van Swaaij W P M (1990) ‘Mass transfer with complex chemical reactions – II. Parallel reversible chemical reactions’, Chemical Engineering Science, 45, 183–197. Wilson M, Tontiwachwuthikul P, Chakma A, Idem R, Veawab A, Aroonwilas A, Gelowitz D, Barrie J and Mariz C (2004) ‘Test results from a CO2 extraction pilot plant at boundary dam coal-fired power plant station’, Energy, 29, 1259–1267. Yagi T, Shibuya H and Sasaki T (1992) ‘Application of chemical absorption process to CO2 recovery from flue gases generated in power plants’, Energy Conversion and Management, 33, 349–355. Yokoyama T (2004) ‘Japanese R&D on large scale CO2 capture’, in Fell C and Keller G E II (eds), Separations Technology VI: New Perspectives on Very Large-Scale Operations, ECI Symposium Series Volume RP3, Engineering Conferences International, New York, Paper 7. ZEP WG1 (2006) The final report from Working Group 1 Power Plant and Carbon Dioxide Capture, The European technology platform for zero emission fossil fuel power plants, http://www.zero-emissionplatform.eu/website/. ZEP WG1 (2008) EU demonstration programme for CO2 capture and storage (CCS), The European technology platform for zero emission fossil fuel power plants, http:// www.zero-emissionplatform.eu/website/.


6

Advanced adsorption processes and technology for carbon dioxide (CO2) capture in power plants R. M. D a v i d s o n, IEA Clean Coal Centre, UK Abstract: This chapter considers the use of solid sorbents for postcombustion carbon dioxide (CO2) capture from coal-fired power plants. First, mesoporous and microporous adsorbents are discussed: carbonbased adsorbents, zeolites, hydrotalcites and porous crystals. Attempts have been made to improve the performance of the porous adsorbents by functionalising them with nitrogen groups and, specifically, amine groups to react with CO2 and thus enhance the physical adsorption properties. Dry, regenerable solid sorbents have attracted a good deal of research. Most of the work has been on the carbonation/calcination cycle of natural limestone, but there have also been studies of other calcium-based sorbents and alkali metal-based sorbents. Key words: carbon dioxide, post-combustion capture, solid sorbents, regenerable adsorbents.

6.1

Introduction

The leading contender for post-combustion carbon dioxide (CO2) capture from coal-fired power plants is the use of solvents (mainly alkanolamines). Despite the advantages of solvent scrubbing, there are several disadvantages, especially the thermal efficiency losses due to the energy needed to regenerate the solvent by driving off the captured CO2. Other problems include the formation of degradation products formed by solvent reaction with flue gas from coal and the associated corrosion problems. There are also the difficulties of handling liquids and the energy penalties associated with the loss of solvents due to evaporation. Current technologies and modelling activities in carbon capture have been briefly reviewed by Ducroux and Jean-Baptiste (2005) who looked at the use of membranes and solid adsorbents in addition to solvent absorbents and cryogenics. They concluded that only limited evolution is expected in the development of chemical absorption based on amines but that the main fields which are subject to significant developments are adsorbents and membranes. This chapter will discuss solid adsorbents beginning with mesoporous and microporous adsorbents in which CO2 adsorption is simply a physical process controlled by the pore characteristics of the sorbent. The addition of chemical functionality such as amine groups has been studied as a means of improving the performance of porous adsorbents so this will 183 © Woodhead Publishing Limited, 2010

184


be discussed in the following section. Next, regenerable solid sorbents will be considered. These mostly involve a chemical cycle of calcination/ carbonation reactions to capture the CO2 and then release it as a pure gas while regenerating the sorbent. The timeframe for the studies included in this chapter is from the year 2000 onwards. This chapter is an edited and much shortened version of a report produced jointly by the IEA Clean Coal Centre and the IEA Greenhouse Gas R&D Programme (Davidson, 2009).

6.2

Mesoporous and microporous adsorbents

Pressure swing adsorption (PSA) is a technology used to separate some gas species from a mixture of gases under pressure according to the species’ molecular characteristics and affinity for an adsorbent material. It operates at near-ambient temperatures. Special adsorptive materials (for example, zeolites) are used as a molecular sieve, preferentially adsorbing the target gas species at high pressure. The process then ‘swings’ to low pressure to desorb the adsorbent material. The sorbents that can be used in PSA can include: ∑ carbon-based adsorbents ∑ zeolites ∑ hydrotalcite-like compounds ∑ porous crystals.

6.2.1 Carbon-based adsorbents The adsorption isotherms of the activated carbon studied by Siriwardane et al. (2001) were found to be extremely reproducible, indicating excellent reversibility of adsorption. The equilibrium adsorption capacity of the activated carbon at 25 ºC and ~2 MPa was about 8.5 mol CO2/kg of sorbent (37.4 wt%). The adsorption isotherms also indicated that a close pack monolayer of CO2 was formed at saturation with the activated carbon. So the activated carbon utilises its complete surface area to form the monolayer at higher pressures. It was possible to obtain an excellent separation of CO2 from a gas mixture containing 14.8 % CO2 and 85.2 % N2 at a flow rate of 5 cm2/min. Tang et al. (2004a,b) have pointed out that anthracites have inherent chemical properties, fine structure and relatively low price that make them excellent raw materials for the production of activated carbons. A laboratory-scale fluidised bed was used for the activation of an anthracite. The activated carbons produced had highly developed microporosity and a small amount of mesopores. The surface area could rise to 1071 m2/g with two hours activation time at 890 ºC (Maroto-Valer et al., 2005). The highest CO2 capture capacity at 30 ºC was 65.7 mg CO2/g sorbent (6.57 wt%) for © Woodhead Publishing Limited, 2010

Advanced adsorption processes and technology

185

the carbon activated for two hours at 800 ºC. Contrary to expectations, the CO2 capacity did not show any clear relationship with the surface area since this carbon had a surface area of 540 m2/g. The anthracite with the highest surface area had a CO2 capacity of only 4 wt%, probably due to only certain pore sizes being effective for CO2 adsorption.

6.2.2 Zeolites Zeolites are the aluminosilicate members of the family of microporous solids known as ‘molecular sieves’. The term molecular sieve refers to the ability to sort molecules selectively based primarily on a size exclusion process. This is due to a very regular pore structure of molecular dimensions. The maximum size of the molecular or ionic species that can enter the pores of a zeolite is controlled by the diameters of the tunnels. Zeolites occur naturally but can be synthesised. Synthetic zeolites can be manufactured in a uniform, phase-pure state. It is also possible to manufacture desirable zeolite structures which do not appear in nature. Siriwardane and her colleagues have produced several studies of zeolite; for example studies of three natural zeolites, with different major cations, were reported by Siriwardane et al. (2003). These are inexpensive materials. Studies of volumetric gas adsorption of CO2, N2 and O2 were conducted at 25 ºC up to a pressure of ~2 MPa. Preferential adsorption of CO2 was observed with all three zeolites. The differences in the observed adsorption were likely to be related to the differences in the chemical nature at the surface, specifically the major cations present, since the average pore diameters were fairly similar. Five manufactured zeolites were tested by Siriwardane et al. (2005), but they pointed out that the CO2 capture systems would be even more energy efficient if the sorbents were operational at moderate or high temperatures. However, CO2 adsorption capacities on the zeolites were less at 120 ºC than at ambient temperature. A potential problem with zeolites is that, if used in fluidised beds, attrition could cause the sorbent to be carried over (Lee et al., 2004). Zeolites 5A and 13X were found to have attrition rates 2.1–4.0-fold higher than activated carbon or activated alumina. This could result in high maintenance costs for the sorbent and problems in the operation of the fluidised bed. On the other hand, the adsorption capacities of 5A and 13X were 2.35 mmol/g (10.34 wt%) and 2.23 mmol/g (9.81 wt%), 1.5–2.7-fold higher than the other sorbents tested.

6.2.3 Hydrotalcites Hydrotalcite is a layered double hydroxide of general formula Mg 6Al2(CO3) (OH)16 @ 4H2O. Recently, they have been considered as adsorbent materials


186


for CO2. For example, Ding and Alpay (2000) studied a PSA process at temperatures up to 480 ºC and in the presence of water vapour. The conditions were chosen to depict those of steam reformer processes, but it was suggested that they were also appropriate to some flue gas CO2 recovery processes. Adsorption saturation capacities of 0.65 and 0.58 mol/kg (28.6 and 25.5 wt%) were measured at 400 and 480 ºC, respectively, under wet feed conditions. Under dry feed conditions, about a 10 % reduction of the saturation capacity was observed and also relatively rapid adsorbent degradation. Adsorbent regeneration was possible by means of a steam purge, but some irreversible loss in capacity was indicated for very long times on-stream (for example, 90 days at 400 ºC).

6.2.4 Porous crystals Metal organic frameworks (MOFs) are crystalline compounds consisting of metal ions or clusters coordinated to organic molecules to form one-, two-, or three-dimensional structures that can be porous. In some cases, the pores can be used for the storage of gases such as CO2. Millward and Yaghi (2005) investigated the viability of nine MOFs for CO2 storage at room temperature. MOF-177, composed of 1,3,5-benzenetribenzoate units and zinc clusters, was found to have a CO2 capacity of 33.5 mmol/g (147.4 wt%), far greater than that of any other porous material reported. At 3.5 MPa, a container filled with MOF-177 can capture nine times the amount of CO2 in a container without adsorbent, and about twice the amount when filled with benchmark materials like activated carbon and zeolite 13X. Arstad et al. (2007) prepared a selection of MOF adsorbents and tested them as low-temperature adsorbents for CO2. The best adsorbents reached CO2 capacity levels of 10 wt% at atmospheric pressures of CO2 and as high as 60 wt% at ~25 MPa CO2 pressure and 25 ºC. The observed capacities were proportional to the specific surface area and pore volume of the adsorbents. However, the selectivity for water adsorption was significantly higher than for CO2. The presence of water also reduced the adsorption capacity of CO2 in gas mixtures. Figueroa et al. (2008) have noted that desirable characteristics for MOFs are low energy requirement for regeneration, good thermal stability, tolerance to contaminants, attrition resistance and low cost.

6.3

Functionalised sorbents

The CO2 adsorption capacity of activated carbons and other mesoporous sorbents, governed by physical adsorption, can possibly be increased by introducing nitrogen (N) functional groups to their structure. There have been several studies in which unspecified nitrogen groups have been introduced



187

into adsorbents but, as Arenillas et al. (2005) have reported, it is apparent that, although the amount of nitrogen incorporated to the final adsorbent is important, the N-functionality seems to be more relevant for increasing CO2 uptake. A suite of high nitrogen content carbon matrix adsorbents was prepared by Drage et al. (2007a) from the activation of urea-formaldehyde (UF) and melamine-formaldehyde (MF) resins, using K2CO3 as a chemical activation agent incorporated into the resin on polymerisation. They found that the CO2 adsorption capacity displayed a hybrid adsorption between physisorption and chemisorption; it was determined to be dependent upon both textural properties and, more importantly, nitrogen functionality. Rather than simply introducing non-specific basic nitrogen functionality into mesoporous sorbents, a better approach might be to introduce amine functionality. Immobilised amine sorbents might be expected to show similar reactions to liquid amines in the typical absorption process, with the added advantages that solids are easier to handle and that they do not give rise to the corrosion problems caused by the circulation of very basic solutions.

6.3.1 Polymer and resin supported sorbents It is possible to coat solid polymers with liquid amines in order to combine the high surface area of the polymeric support with the CO2 removal efficiency of a liquid amine. Polyethyleneimine (PEI) and diethanolamine (DEA) are two amines that can be applied to support surfaces. Such sorbents were originally developed for space life support systems and are too expensive for use in post-combustion capture in power plants or other large-scale applications (Gray et al., 2003). Michael addition reaction products of ethylenediamine (EDA) and tetraethylenepentamine (TEPA) were immobilised within the pores of high surface area poly(methyl methylacrylate) solid beads by Gray et al. (2005). The primary amine sites present in the EDA and TEPA were converted to secondary amine sites by reacting them with acrylonitrile. The test results were such that it was suggested that there could be potential applications of these sorbents in the capture of CO2 from flue gas streams. Gray et al. (2008) have pointed out that, in order for the solid amine sorbents to be competitive with existing monoethanolamine (MEA) liquid systems, CO2 capture capacity must be in the range of 3–6 mol CO2/kg sorbent (13.2–26.4 wt%). This has not been achieved at the critical operating temperature of 65 ºC.

6.3.2 Carbon supported sorbents Commercial solid sorbents are very expensive so efforts have been made to seek out less expensive substances that can be converted into high surface


188


area materials easily. The unburned carbon in fly ash is one such material. However, as Maroto-Valer et al. (2008) have reported, the impregnation of the amines can significantly reduce the surface area and pore volume of the activated fly ash. High surface area powdered anthracites which had been surface treated were studied by Maroto-Valer et al. (2005) and also by Tang et al. (2004b) to assess their suitability as low-cost CO2 sorbents. The treatments included NH3 heat treatment and PEI impregnation. The NH3 heat treatment increased the surface area of the activated samples, especially at lower temperatures (650 ºC), but the PEI impregnation resulted in a dramatic decrease in surface area. This decrease was attributed to pore blockage and surface coverage by the PEI. Three alkylamines were evaluated as a potential source of basic sites for CO2 capture by Plaza et al. (2007) using a commercial activated carbon as a support. The amine coating increased the basicity and nitrogen content of the carbon but drastically reduced its mesoporous volume, the factor mainly responsible for CO2 physisorption. Thus, the capacity of the raw carbon was reduced at room temperature. However, at medium temperatures (70–90 ºC), the contribution of chemisorption associated with the incorporated amino groups may improve the performance of the carbon.

6.3.3 Silica supported sorbents Solid phase hexagonal mesoporous silica (HMS) modified using aminopropyltrimethoxysilane (APTS) and related compounds have been studied by Delaney et al. (2002). The modified silica produced very high surface area material with varied concentrations of surface bound (tethered) amine and hydroxyl functional groups which can react with CO2. Studies by Knowles et al. (2005a) showed that both the mass uptake per unit surface area and the heat evolved per unit surface area increased with the nitrogen content of the tether. Hence, their results showed that the higher the nitrogen content of the tether, the higher the CO2 capacity on the adsorbent surface. It was also reported that the materials were able to adsorb CO2 in the presence of moderate amounts of water. In the presence of water, CO2 capacity of one hybrid material was found to be enhanced, although the rate of desorption was diminished (Knowles et al., 2005b). CO2 adsorption/desorption on mesoporous silica SBA-15 grafted with g-(aminopropyl)triethoxysilane has been studied by Chang et al. (2003). The large uniform pore size of SBA-15 facilitates CO2 diffusion inside of the pore, allowing rapid CO2 adsorption on the surface amine sites. CO2 was found to adsorb on the amine sites in the form of carbonate and bicarbonate which desorb as CO2. Zheng et al. (2004, 2005) synthesised an EDA modified SBA-15



189

mesoporous silica and characterised its CO2 adsorption properties. The CO2 adsorption capacity of the EDA–SBA-15 sorbent was around 20 mg/g (2 wt%) at 25 ºC and 1 atm (~0.1 MPa) with 15 % CO2 (by volume) in N2. The CO2 adsorption capacity of the EDA–SBA-15 was not influenced by moisture. It was recognised that the adsorption capacity of the sorbent was not adequate for the reduction of CO2 emission at large scale economically; a further increase in capacity is necessary. However, Tarka et al. (2006a,b) have reported that an amine-enriched SBA-15 sorbent had a CO2 transfer capacity of 6.4 mole/kg of sorbent at 54 ºC (28.2 wt%). Liang et al. (2006, 2008) have reported the preparation and characterisation of a stepwise growth of mesoporous silica (SBA-15) bound melamine dendrimers together with their CO2 absorption capacities. The melaminebased dendrimers contain a mixture of primary, secondary and tertiary amine groups, providing a range of active sites with varied basic strength. The experimentally determined CO2 adsorption capacities were found to be similar to, but less than, the capacities predicted on the basis of interaction with the primary amine groups only. The adsorption capacities were described as not particularly impressive. The aziridine (ethylene imine) monomer was polymerised on the surface of SBA-15 by Hicks et al. (2008a,b) to prepare a hyperbranched aminosilica (SBA-HA). The sorbent was found to be capable of adsorbing CO2 reversibly with very high capacities of 3.1 mmol CO2/g material at 25 ºC (13.6 wt%). The advantage of this adsorbent over previously reported adsorbents rests in its large CO2 capacity and multicycle stability. A modified mesoporous silica-based molecular sieve (MCM-41) was investigated by Xu et al. (2002a,b, 2003) and by Song et al. (2004). They called the large pore volume MCM-41 substrate a ‘basket’ and this was turned into a ‘molecular basket’ for CO2 adsorption by immobilising the sterically branched polymer PEI into the channels of the MCM-41. A CO2 adsorption capacity of the produced MCM-41-PEI as high as 215 mg CO2/g PEI (21.5 wt%) was obtained at 75 ºC, which was 24 times higher than that of the MCM-41 alone and twice that of the pure PEI. However, it was reported by Xu et al. (2004) that it was not stable when the operation temperature was higher than 100 ºC.

6.3.4 Alumina supported sorbents A series of solid sorbents was synthesised by Plaza et al. (2008) by immobilising liquid alkylamines and alkanolamines on the surface of a mesoporous alumina. The basic nature of the sorbents was expected to be favourable for their application in the adsorption of CO2. However, impregnation had a negative effect on the texture of the resultant sorbents, as revealed by the significant decrease in the BET surface areas. This decrease is a consequence of pore


190


blockage by the amine film. CO2 adsorption on amine-impregnated alumina cannot be explained by a single process, as it can be controlled by physical/ chemical sorption or by mass transfer. Therefore, the type and load of amine used to impregnate the support and the temperature of the process need to be optimised. In the absence of water, tertiary amines cannot capture CO2 since they cannot form carbamates. Hence, CO2 adsorption can occur only by physisorption and this is less than with the raw alumina since the presence of the tertiary amine reduces the surface area. Sterically hindered amines also perform poorly. The better performing amines were those with more than one amino group in their structure and their capture capacity increased with increasing temperature.

6.3.5 Zeolite supported sorbents Zeolite 13X was impregnated with MEA with loadings of 0.5–25 wt% by Jadhav et al. (2007). The adsorbent with a 10 wt% loading showed an improvement in CO2 adsorption capacity over the unmodified zeolite by a factor of around 1.6 at 30 ºC, whereas at 120 ºC, the efficiency improved by a factor of 3.5. A small MEA loading (0.5 wt%) had very little effect on the CO2 adsorption capacity. With the highest loading (25 wt %), the CO2 capacity decreased. This was attributed to reduced surface area and pore volume and restricted access to adsorption sites for CO2 at higher loadings. The 10 wt% MEA-modified adsorbent also emerged as a potential adsorbent for CO2 in the presence of moisture. The performance of the adsorbent was also satisfactory in repeated cycles of adsorption.

6.3.6 Glass fibre supported sorbents A sorbent for CO2 capture was developed by Li et al. (2008) based on coating PEI on a glass fibre matrix using epichlorohydrin (ECH) as a crosslinking agent. The formation of a network structure between ECH and PEI was to enhance the thermal stability of the adsorbent. The maximum CO2 adsorption capacity achieved was 4.12 mmol CO2/g total adsorbent (18.1 wt%) and 13.56 mmol CO2/g PEI (59.66 wt%) on the adsorbent at 1 atm (~101 kPa), 30 ºC and about 80 % relative humidity, with 57 % utilisation of the amine compound. The sorbent had high thermal stability at about 250 ºC, a low regeneration temperature of 120 ºC and was stable in the presence of moisture.

6.3.7 Xerogel supported sorbents Derivatised aliphatic diamines bearing a silyl group have been studied by Dibenedetto et al. (2003) and by Aresta and Dibenedetto (2003). The diamines



191

were more efficient for the reversible uptake of CO2 than monoamines, but they could also be converted into silica xerogels with high porosity and surface area. The xerogels were found to absorb and desorb CO2 from a gas mixture in a very short time and for several cycles in a batch system.

6.3.8 Porous crystals Zeolitic imidazolate frameworks (ZIFs) are porous crystalline materials with tetrahedral networks that resemble those of zeolites. Imidazole is a five-membered aromatic heterocyclic ring containing two nitrogen atoms. Hayashi et al. (2007) suggest that ZIFs potentially have the advantages of both inorganic zeolites (for example, high stability) and of MOFs (for example, high porosity and organic functionality). They reported the synthesis of three porous ZIFs that are expanded analogues of zeolite A. The ZIFs had functionalised purinate cage walls and displayed a strong interaction with CO2. ZIFs have been synthesised by Banerjee et al. (2008) from either zinc(II)/cobalt(II) and imidazolate/imidazolate-type linkers. Three of the ZIFs synthesised had high thermal stability (up to 390 ºC), chemical stability and high porosity with surface areas up to 1970 m2/g. Their capacity for CO2 was described as ‘extraordinary’; one litre of ZIF-69 can hold 82.6 litres (162 g) of CO2 at 0 ºC under ambient pressure.

6.3.9 Templated sorbents Drage et al. (2007b,c) explored nanocasting or templating by which means high surface area adsorbents can be generated without the need for heat treatment or activation. In nanocasting, an inorganic diluent such as silica is used as a template to shape the growing polymer. This template is removed by dissolution, but the polymer remains as an ‘image’, inheriting the porosity of the template. Porous MF resins were synthesised in the presence of 7 nm and 14 nm fumed silica as a templating agent. The solid MF/silica product was treated with sodium hydroxide solution and then dried at 120 ºC. The adsorption capacity was highest at 25 ºC and decreased gradually with increasing temperature. The capacity and decrease in adsorption of CO2 at elevated temperature was superior to the performance of a standard commercial activated carbon. However, it was suggested that, although the nanocasting technique is effective in producing high surface area materials, exploration of different polymer types is required to generate adsorbents containing sufficiently basic nitrogen. Amines would be preferable to the amide and triazine ring nitrogen groups of the MF resin. Further carbonisation to generate the adsorbents was carried out by heating the samples at 10 ºC/min up to 400, 500, 600 and 700 ºC for one hour (Pevida et al., 2008). The carbonised adsorbents presented CO2 capture capacities up to 2.25 mmol/g at 25 ºC (9.9


192


wt%), outperforming many commercial activated carbons. Both texture and surface chemistry influenced the CO2 capture performance of the prepared adsorbents: while for adsorbent carbonised at 700 ºC, physisorption seemed to be the controlling mechanism, for the samples carbonised at 500 ºC, a favourable chemistry appeared to determine CO2 capture.

6.4

Regenerable sorbents

The role of solids in CO2 capture has been the subject of a ‘mini review’ by Harrison (2005) who pointed out that solid-based processes possess many potential advantages including: ∑ a wide range of operating temperatures ∑ reduced energy penalties ∑ avoidance of liquid wastes ∑ the relatively inert nature of solid wastes. In solid-based processes, sorbent is consumed and CO2 captured during reaction and the CO2 is liberated as the sorbent is regenerated. Recently, the term ‘chemical looping’ has been applied to this cycle. The term is probably better known in the context of chemical looping combustion in which metal oxides are used to supply oxygen to a fuel gas and are subsequently regenerated by reaction with air. CO2 looping cycles have been reviewed by Anthony (2008) and use a sorbent capable of carbonation to remove CO2 from a combustion or gasification environment:

MxO + CO2 = MxCO3

The resulting carbonate is then regenerated:

MxCO3 = MxO + CO2

6.4.1 Natural minerals From a cost point-of-view, natural minerals such as limestone (calcium carbonate, CaCO3) should have a fairly obvious advantage. However, the use of natural calcium carbonates as regenerable CO2 sorbents is limited by the rapid decay of the carbonation conversion with the number of cycles of carbonation/calcination, where calcination is the regeneration reaction (Abanades, 2002). The basic concept is shown in Fig. 6.1. Studies have shown that the recarbonation reaction is far from reversible in practice. Typically, a limestone’s carrying capacity falls from an initial value of ~79 % to only about 20–30 % after ~30 cycles (Fennell et al., 2007a). After an initial, rapid reaction period, a much slower second stage follows, controlled by diffusion in the CaCO3 layers. Thus, there is a need for a fresh feed of sorbent to compensate for the decay in activity during sorbent © Woodhead Publishing Limited, 2010


193

Low CO2 and SO2 flue gas

Flue gas Carbonator T = 650–850 °C

CaO

CaCO3

Calciner T = 900–1100 °C O2

CO2

6.1 Schematic of carbonation/calcination cycle (Davidson, 2009).

recycling. Nevertheless, Abanades et al. (2003) proposed that ‘a modest supply of fresh limestone, comparable to the use for sulphur control purposes, is enough to compensate for the intrinsic decay in sorbent activity’. They have calculated that it would be possible to operate a CO2 capture system with efficiencies of over 80 % with fresh sorbent addition comparable to those used for sulphur control in some power stations burning high sulphur fuels. This still involves large quantities of fresh sorbent in the order of 50–100 t fresh limestone/h for a 1000 MWt power plant based on coal (Grasa et al., 2007a). Based on experiments in fluidised bed carbonator–combustor systems, Rodrigo-Naharro and Clemente Jul (2008) have found that, in order to maintain the activity, a flow of 0.813 kg of fresh limestone was needed per 1 kg of coal. Volumes such as that are only acceptable because limestone has a very low unit cost and is widely available, and because the sorbent and its products are not hazardous materials. If lower CO2 capture efficiencies of 50–70 % were acceptable, then the system could run with much lower quantities of fresh sorbent. Despite the need for fresh sorbent, CO2 capture based on carbonation/ calcination cycles has some inherent advantages compared with other approaches: ∑

The efficiency penalties are intrinsically low because both the capture and sorbent regeneration processes are carried out at high temperatures. ∑ The sorbent is cheap and widely available. ∑ Most of the individual process units are commercially proven and/or there exist similar large-scale commercial processes. ∑ No hazardous materials are involved.


194


Natural minerals other than limestone have been proposed as sorbents for CO2 capture. Bandi et al. (2005) screened some and found that natural carbonates from the dolomite group were promising. The cyclic stability is dependent on their chemical composition. An increased concentration of inert matter in an absorbent had a positive influence on the cyclic stability. Limestone deactivation mechanism The mechanism by which the limestone deactivates was studied by Abanades and Alvarez (2003) who interpreted the observed conversion limits in the reaction of CO2 with lime in terms of a certain loss in the porosity associated with small pores and a certain increase in the porosity associated with large pores. In the carbonation part of every cycle, the CaCO3 fills up all the available porosity made up of small pores plus a small fraction of the large voids, limited by the thickness of the product layer that marks the onset of the slow carbonation rate. Fennell et al. (2007b) studied the effects of repeated cycles of calcination and carbonation on five different limestones using a hot fluidised bed of sand. This exposed the limestone to attrition in the bed. They found that, for most limestones, losses due to attrition amounted to less than 10 % of their mass over the course of a typical experiment, lasting about eight hours. The carrying capacity of the CaO for CO2 was found to be roughly proportional to the voidage inside pores narrower than ~150 nm in the calcined CaO before carbonation began. Morphological changes, including reduction in the pores narrower than 150 nm within a calcined limestone, were found to be responsible for much of the fall in carrying capacity with increasing number of cycles. An experimental parametric study of the CaO-based sorbent capacity under carbonation/calcination cycles has been carried out by Manovic and Anthony (2008a). The influence of various parameters was examined. These included particle size, impurities, limestone type, temperature and influence of the effective CO2 concentration surrounding reacting particles, as well as carbonation/calcination duration and temperature stresses in different reactor types. It was found that increasing temperature in the range of 650–850 ºC has a negative effect on the sorbent activity. Particle size is unimportant in terms of the sorbent CO2 carrying capacity, and any differences are likely to be a function of impurity content in samples of different particle size. Prolonged carbonation accelerates the sorbent activity decay, while an opposite effect is produced for the prolonged sorbent exposure to calcination conditions in an inert atmosphere. It was concluded that the formation and decomposition of CaCO3 is a critical step for sorbent sintering and its subsequent decay in activity. A model explaining the loss of sorbent activity has been proposed by



195

Manovic and Anthony (2008b). During cycling, two different types of mass transfer in the sorbent particles must occur in parallel: bulk diffusion connected to the formation/decomposition of CaCO3 and ion diffusion in the crystal structure of CaO. Ion diffusion in CaO stabilises its crystal structure but with no significant effect on particle morphology and corresponding carbonation conversions. Bulk mass transfer occurred during formation and decomposition of CaCO3, and this led to major changes of morphology (that is, to sintering and loss of small pores and sorbent activity). During CO2 cycles, competition occurred between ion diffusion and bulk mass transfer, and two types of structure or skeleton were formed: an internal unreacted structure and an external structure in which carbonation/calcination proceeded. The internal skeleton can be considered as a hard skeleton that stabilises and protects the particle pore structure. The external structure or skeleton can be considered to be a soft skeleton that easily changes during CaCO 3 formation and decomposition, resulting in changes of particle morphology. Manovic and Anthony (2008b) point out that it should be noted that the terms internal and external skeleton are not related to the particle but to the pores: outer or external means that this part of the skeleton is exposed to pores and surrounded by the gas that fills pores. The model predicts that, as a result of CO2 cycles, pore size distribution changes and that smaller pores transform into larger ones, leading to a loss of pore surface area and loss of activity. Improving calcium sorbent performance Manovic and Anthony (2008a), in their parametric study on the CO2 capture capacity of CaO-based sorbents, found that the grinding of the sorbent results in better reaction cycle properties for the powder obtained. They suggested that this was unlikely to be due simply to the particle size decrease and suggests that it may be possible to achieve sorbent activation by grinding. However, other methods have been studied in order to improve sorbent performance. These have included thermal activation (Manovic and Anthony, 2008b) and hydration (Manovic and Anthony, 2009). Steam hydration increases both the pore area and pore volume of CaO. It has also been reported by Fennell et al. (2007a) that it is possible to regenerate spent CaO by reaction with humid air. By reacting spent CaO overnight with moist, ambient air, the sorbent can be regenerated to ~55 % carrying capacity (moles of CO2 per mole of CaO).

6.4.2 Synthetic sorbents Synthetic calcium sorbents have been studied as a means of increasing surface area and reactivity compared with sorbents derived from naturally occurring


196


limestone. Grasa et al. (2007a,b) have reviewed some of the recent literature on synthetic sorbents that aim to overcome the decay in capture capacity of natural limestone sorbents. The object of using synthetic precursors is to obtain calcium oxides with a high surface area and a more stable pore structure. They reached the conclusion that none of the reviewed synthetic sorbents has a chance to compete with the ‘modest’ performance of natural limestones that show two competitive advantages: the maintenance of a suitable CO2 capture capacity under demanding process conditions and their intrinsic low cost. Other alkali metals have been proposed as dry, regenerable sorbents for the capture of CO2 from a gas stream. These have included sodium carbonate (Na2CO3), potassium carbonate (K2CO3), and lithium sorbents. Ryu et al. (2005, 2007) have prepared formulations containing 20–35 wt% of either Na2CO3 or K2CO3. Laboratory-scale tests indicated that two of the sorbents, designated Sorb NX30 and Sorb KX35, had almost all the requirements for a commercial fluidised bed reactor process. They showed near complete regeneration below 120 ºC with a possible regeneration window of 80–160 ºC. They had superior attrition resistance and high CO2 sorption capacity along with high bulk density. The mechanism of CO2 sorption/desorption on lithium zirconate (Li2ZrO3) was studied by Ida and Lin (2003) who found that pure Li2ZrO3 can absorb up to ~20 wt% of CO2. However, despite its high CO2 sorption capacity, the sorption rate is slow. Experiments were also carried out using Li2ZrO3 modified by additions of lithium carbonate (Li2CO3) and K2CO3. The modified Li2ZrO3 showed a dramatically improved rate of sorption, about 40 times faster than pure Li2ZrO3 at 500 ºC, although it was still low. Fauth and Pennline (2004) and Fauth et al. (2005) studied a number of binary and ternary eutectic salt modified Li2ZrO3 sorbents and found that the combination of binary alkali carbonate, binary alkali/alkaline earth carbonate ternary alkali carbonate, and ternary alkali carbonate/halide eutectics with Li2ZrO3 noticeably improved the CO2 uptake and absorption capacity. It was suggested that the formation of a eutectic molten carbonate layer on the outer surface of reactant Li2ZrO3 particles aids in facilitating the transfer of gaseous CO2 during the sorption process. The ternary K2CO3/NaF/Na2CO3 eutectic and Li2ZrO3 combination at 600 and 700 ºC produced the fastest CO2 uptake rate and highest CO2 capacity. In comparison with Li2ZrO3, Kato et al. (2002) found that: ∑

lithium orthosilicate (Li4SiO4) absorbs CO2 about 30 times faster than Li2ZrO3 at 500 ºC in 20 % CO2 gases; ∑ Li4SiO4 absorbs CO2 at 500 ºC in 2 % CO2 in which Li2ZrO3 shows little absorption; ∑ Li4SiO4 absorbs CO2 even from the ambient air at room temperature.



6.5

197


The (US) National Energy Technology Laboratory (NETL) has published a review of advances in CO2 capture technology which are part of the US Department of Energy’s Carbon Sequestration Program (Figueroa et al., 2008). The scope of that review is much broader than this chapter. Much of NETL’s research activities can be found on its website where various factsheets can be found. Some useful ones include: ∑

MOFs – http://www.netl.doe.gov/publications/factsheets/project/Proj315. pdf ∑ immobilised amine sorbents – http://www.ieaghg.org/index. php?/2009112026/high-temperature-solid-looping-cycles-network.html ∑ metal monolithic amine-grafted zeolites – http://www.netl.doe.gov/ publications/factsheets/project/Proj467.pdf ∑ sodium carbonate as a dry regenerable sorbent – http://www.netl.doe. gov/publications/factsheets/project/Proj198.pdf

An international network on high-temperature solid looping cycles was adopted as an IEA GHG network in December 2008. The aim of the network is to promote further development and scale-up of processes for CO2 capture which involve solid looping cycles operating at elevated temperatures. One application is high-temperature carbonation/calcination to remove CO2 from flue gases or reformed gas streams. An aim of the network is to expand current participation beyond the research community to include potential operators, plant designers and equipment suppliers because the technology is starting to move from the bench scale to pilot and industrial demonstration scale. Further information can be found at http://co2captureandstorage.info/ networks/looping.htm.

6.6

Conclusions

Simple porous sold sorbents such as activated carbons and zeolites are probably not well-suited to post-combustion CO2 capture. Their CO2 capacities and their CO2/N2 selectivities are not very high and they would need to use expensive pressure swing adsorption processes or variants of PSA. The much higher CO2 capacity metal organic frameworks (MOFs) and their derivatives look promising but are at an early stage of development. They would need to be produced quite cheaply on a very large scale to be used for CO2 capture in power plants and they need to be proven to work with real flue gases. Functionalised solid sorbents, especially immobilised amine sorbents, would seem to be a logical improvement over simple mesoporous adsorbents. However, the results of much research have not been very encouraging. It would appear that the increase in CO2 capacity by the functional groups seems


198


to be offset by the reduction of porosity caused by the functional groups filling the pores. Knowles et al. (2005b) admit that the CO2 adsorption capacities of aminopropyl-functionalised materials to date are not outstanding. However, they argue that there is insufficient data in the literature to distinguish whether higher CO2 adsorption capacities can be achieved. The development of functionalised amine sorbents will depend on improving characteristics such as high thermal stability, excellent CO2 stability, high CO2 adsorption capacity, easy CO2 desorption and reversible regeneration Research on dry, regenerable, solid sorbents is currently at the pilot plant demonstration stage. They have the great advantage of being cheap, especially if they are based on natural limestone. Even their loss of capacity with cycling does not seem to be that worrying if it remains at about 20–30% after ~30 cycles (Fennell et al., 2007a). This capacity is still higher than that of the simple and the functionalised porous adsorbents. However, the requirements for fresh limestone feed seem to be of the same order as the coal requirements of a plant. This would certainly limit their use for retrofitting plants unless there is a great deal of stockpile space available and probably a nearby cement plant to take the calcined waste lime produced. The potential success of solid sorbents will depend on whether it is true that only limited evolution is expected in the development of chemical absorption based on amines and that significant developments will be made in the development of adsorbents. Certainly, there is scope for continued research, especially concerning their behaviour in real coal-fired power plant conditions. As Harrison (2005) has pointed out, the key factors in the further development of solid processes are the cost and durability of the reactive solids, along with the development of technology to manage the large solid circulation rates.

6.7

References

Abanades J C (2002) The maximum capture efficiency of CO2 using a carbonation/ calcination cycle of CaO/CaCO3. Chemical Engineering Journal; 90 (3); 303–306. Abanades J C and Alvarez D (2003) Conversion limits in the reaction of CO2 with lime. Energy and Fuels; 17 (2); 308–315. Abanades J C, Oakey J E, Alvarez D and Hämäläinen J (2003) Novel combustion cycles incorporating capture of CO2 with CaO. In: Gale J and Kaya Y (eds). Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Oxford, UK, Elsevier (Pergamon), Vol. I, 181–186. Anthony E J (2008) Solid looping cycles: a new technology for coal conversion. Industrial and Engineering Chemistry, Research; 47 (6); 1747–1754. Arenillas A, Drage T C, Smith K and Snape C E (2005) CO2 removal potential of carbons prepared by co-pyrolysis of sugar and nitrogen containing compounds. Journal of Analytical and Applied Pyrolysis; 74 (1/2); 298–306. Aresta M and Dibenedetto A (2003) New amines for the reversible absorption of carbon dioxide from gas mixtures. In: Gale J and Kaya Y (eds), Proceedings of the sixth



199

international conference on greenhouse gas control technologies: GHGT6, Oxford, UK, Elsevier Pergamon, Vol. II, 1599–1602. Arstad B, Fjellvåg H, Kongshaug K O, Swang O and Blom R (2007) MOFs as low temperature adsorbents for carbon dioxide. In: Gale, J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the 8th International Conference on Greenhouse Gas Control Technologies, GHGT:8 Oxford, UK, Elsevier, paper 01_10_03.PDF (CD-ROM). Bandi A, Marquand-Moellenstedt T, Sichler P, Axmann P, Specht M, Ioannides T and Stucki S (2005) High temperature absorbents for CO2 capture. In: Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7. Cheltenham, UK, IEA GHG, Vol. II – Part 2, 1759–1763. Banerjee R, Phan A, Wang B, Knobler C, Furukawa H, O’Keeffe M and Yaghi O M (2008) High-throughput synthesis of zeolitic imidazolate frameworks and application to CO2 capture. Science; 319 (5865); 939–943. Chang A C C, Chuang S S C, Gray M and Soong Y (2003) In-situ infrared study of CO2 adsorption on SBA-15 grafted with g-(aminopropyl)triethoxysilane. Energy and Fuels; 17 (2); 468–473. Davidson R M (2009) Post-combustion carbon capture – solid sorbents and membranes. Report CCC/144, London, UK, IEA Clean Coal Centre. Delaney S W, Knowles G P and Chaffee A L (2002) Hybrid mesoporous materials for carbon dioxide separation. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 47 (1); 65–66. Dibenedetto A, Aresta M and Girardi R (2003) New amines for carbon dioxide separation from gas mixtures. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 48 (1); 167–168. Ding Y and Alpay E (2000) Equilibria and kinetics of CO2 adsorption on hydrotalcite adsorbent. Chemical Engineering Science; 55 (17); 3461–3474. Drage T C, Arenillas A, Smith K M, Pevida C, Piippo S and Snape C E (2007a) Preparation of carbon dioxide adsorbents from the chemical activation of urea-formaldehyde and melamine-formaldehyde resins. Fuel; 86 (1/2); 22–31. Drage T C, Pevida C and Snape C E (2007b) Templated polymeric materials as adsorbents for the post-combustion capture of CO2. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 52 (1); 7–8. Drage T C, Pevida C and Snape C E (2007c) Templated polymeric materials as adsorbents for the post-combustion capture of CO2. In: 2007 international conference on coal science and technology, London, UK, IEA Clean Coal Centre, paper 7C5 (CD-ROM). Ducroux R and Jean-Baptiste P (2005) Technologies, methods and modelling for CO2 capture. In: Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7: Cheltenham, UK, IEAGHG, Vol. II – Part 2, 1835–1839. Fauth D J and Pennline H W (2004) CO2 scrubbing with novel lithium zirconate sorbents. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 49 (1); 310–311. Fauth D J, Frommell E A, Hoffman J S, Reasbeck R P and Pennline H W (2005) Eutectic salt promoted lithium zirconate: novel high temperature sorbent for CO2 capture. Fuel Processing Technology; 86 (14/15); 1503–1521. Fennell P S, Davidson J F, Dennis J S and Hayhurst A N (2007a) Regeneration of sintered limestone sorbents for the sequestration of CO2 from combustion and other systems. Journal of the Energy Institute; 80 (2); 116–119. © Woodhead Publishing Limited, 2010

200


Fennell P S, Pacciani R, Dennis J S, Davidson J F and Hayhurst A N (2007b) The effects of repeated cycles of calcination and carbonation on a variety of different limestones, as measured in a hot fluidized bed of sand. Energy and Fuels; 21 (4); 2072–2081. Figueroa J D, Fout T, Plasynski S, McIlvried H and Srivastava R D (2008) Advances in CO2 capture technology – the U.S. Department of Energy’s Carbon Sequestration Program. International Journal of Greenhouse Gas Control; 2 (1); 9–20. Grasa G, González B, Alonso M and Abanades J C (2007a) Comparison of natural and synthetic precursors of CaO as CO2 sorbents. In: 2007 International Conference on Coal Science and Technology, London, UK, IEA Clean Coal Centre, paper 9C2 (CD-ROM). Grasa G, González B, Alonso M and Abanades J C (2007b) Comparison of CaO-based synthetic CO2 sorbents under realistic calcination conditions. Energy and Fuels; 21 (6); 3560–3562. Gray M L, Soong Y, Champagne K J, Pennline H, Baltrus J, Stevens R W Jr, Khatri R, Chuang S S C and Khan S (2003) Performance of solid amine sorbents. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 48 (1); 164–165. Gray M L, Fauth D J, Soong Y, Champagne K J, Pennline H, Baltrus J and Filburn T (2005) Temperature effect on the capture of carbon by immobilized amine sorbents. In: Fourth Annual Conference on Carbon Capture & Sequestration, Alexandria, VA, 2–5 May, paper 239, available at: http://www.netl.doe.gov/publications/proceedings/05/ carbon-seq/Tech%20Session%20Paper%20239.pdf (accessed December 2009). Gray M L, Champagne K J, Fauth D, Baltrus J P and Pennline H (2008) Performance of immobilized tertiary amine solid sorbents for the capture of carbon dioxide. International Journal of Greenhouse Gas Control; 2 (1); 3–8. Harrison D P (2005) The role of solids in CO2 capture: a mini review. In: Wilson M, Morris T, Gale J, Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Cheltenham, UK, IEA GHG, Vol. II – Part 1, 1101–1106. Hayashi H, Côté A P, Furukawa H, O’Keeffe M and Yaghi O M (2007) Zeolite A imidazolate frameworks. Nature Materials; 6 (7); 501–506. Hicks J S, Drese J H, Fauth D J, Gray M L, Qi G and Jones C W (2008a) Designing adsorbents for CO2 capture from flue gas – hyperbranched aluminosilicas capable of capturing CO2 reversibly. Journal of the American Chemical Society; 130 (10); 2902–2903. Hicks J S, Drese J H, Fauth D J, Gray M L and Jones C W (2008b) Stable aminopolymersilica hybrid material capable of capturing CO2 from simulated flue gas reversibly. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 53 (1); 227–228. Ida J-I and Lin Y S (2003) Mechanism of high-temperature CO2 sorption on lithium zirconate. Environmental Science & Technology; 37 (9); 1999–2004. Jadhav P D, Chatti R V, Biniwale R B, Labhsetwar N K, Devotta S and Rayalu S S (2007) Monoethanol amine modified Zeolite 13X for CO2 adsorption at different temperatures. Energy and Fuels; 21 (6); 3555–3559. Kato M, Yoshikawa S and Nakagawa K (2002) Carbon dioxide absorption by lithium orthosilicate in a wide range of temperature and carbon dioxide concentrations. Journal of Materials Science Letters; 21; 485–487. Knowles G P, Delaney S W and Chaffee A L (2005a) Amine-functionalised mesoporous silicas as CO2 adsorbents. Studies in Surface Science and Catalysis; 156; 887–896. Knowles G P, Graham J V, Delaney S W and Chaffee A L (2005b) Aminopropyl-



201

functionalized silicas as CO2 adsorbents. Fuel Processing Technology; 86 (14/15); 1435–1448. Lee S-S, Yoo J-S, Moon G-H, Park S-W, Park D-W and Oh K-J (2004) CO2 adsorption with attrition of dry sorbents in a fluidized bed. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 49 (1); 314–315. Li P, Ge B, Zhang S, Chen S, Zhang Q and Zhao Y (2008) CO2 capture by polyethyleniminemodified fibrous adsorbent. Langmuir; 24 (13); 6567–6574. Liang Z, Fadhel B, Schneider C J and Chaffee A L (2006) Mesopore-bound melaminetype dendrimers and their CO2 adsorption properties. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 51 (1); 159–161. Liang Z, Fadhel B, Schneider C J and Chaffee A L (2008) Stepwise growth of melaminebased dendrimers into mesopores and their CO2 adsorption properties. Microporous and Mesoporous Materials; 111 (1/3); 536–543. Manovic V and Anthony E J (2008a) Parametric study on the CO2 capture capacity of CaO-based sorbents in looping cycles. Energy and Fuels; 22 (3); 1851–1857. Manovic V and Anthony E J (2008b) Thermal activation of CaO-based sorbent and self-reactivation during CO2 capture looping cycles cycles. Environmental Science & Technology; 42 (11); 4170–4174. Manovic V and Anthony E J (2009) Improvement of CaO-based sorbent performance for CO2 looping cycles. Thermal Science; 13 (1); 89–104. Maroto-Valer M M, Tang Z and Zhang Y (2005) CO2 capture by activated and impregnated anthracites. Fuel Processing Technology; 86 (14/15); 1487–1502. Maroto-Valer M M, Zhe L, Zhang Y and Tang Z (2008) Sorbents for CO2 capture from high carbon fly ashes. Waste Management; 28 (11); 2320–2328. Millward A R and Yaghi O M (2005) Metal-organic frameworks with exceptionally high capacity for storage of carbon dioxide at room temperature. Journal of the American Chemical Society; 127 (51); 17998–17999. Pevida C, Drage T C and Snape C E (2008) Silica-templated melamine-formaldehyde resin derived adsorbents for CO2 capture. Carbon; 46 (11); 1464–1474. Plaza M G, Pevida C, Arenillas A, Rubiera F and Pis J J (2007) CO2 capture by adsorption with nitrogen enriched carbons. Fuel; 86 (14); 2204–2212. Plaza M G, Pevida C, Arias B, Fermoso J, Arenillas A, Rubiera F and Pis J J (2008) Application of thermogravimetric analysis to the evaluation of aminated solid sorbents for CO2 capture. Journal of Thermal Analysis and Calorimetry; 92 (2); 601–606. Rodrigo-Naharro J and Clemente Jul C (2008) The carbonation-calcination cycles as a system to capture CO2 in post-combustion. Poster presented to: 7th European Conference on Coal Research and its Applications, Cardiff, UK, 3–5 September. Ryu C K, Lee J B, Eom T H, Oh J M and Yi C K (2005) Development of Na and K-based sorbents for CO2 capture from flue gas. In: Fourth Annual Conference on Carbon Capture & Sequestration, Alexandria, VA, USA, 2–5 May, paper 113, available at: http://www.netl.doe.gov/publications/proceedings/05/carbon-seq/Tech%20Session%20 Paper%20113.pdf (accessed December 2009). Ryu C K, Lee J B, Baek J-I, Eom H M and Yi C K (2007) CO2 capture from flue gas using dry regenerable sorbents. In: Gale J, Rokkc N, Zweigel Pard Svenson H (eds), Proceedings of the 8th International Conference on Greenhouse Gas Control Technologies: GHG78. Oxford, UK, Elsevier, paper P01_09_11.PDF, (CD-ROM). Siriwardane R V, Shen M-S, Fisher E P and Poston J A (2001) Adsorption of CO2 on molecular sieves and activated carbon. Energy and Fuels; 15 (2); 279–284. Siriwardane R V, Shen M-S and Fisher E P (2003) Adsorption of CO2, N2, and O2 on natural zeolites. Energy and Fuels; 17 (3); 571–576. © Woodhead Publishing Limited, 2010

202


Siriwardane R V, Shen M-S, Fisher E P and Losch J (2005) Adsorption of CO2 on zeolites at moderate temperatures. Energy and Fuels; 19 (3); 1153–1159. Song C, Xu X, Andresen J M, Miller B G and Scaroni A W (2004) Novel nanoporous ‘molecular basket’ adsorbent for CO2 capture. Studies in Surface Science and Catalysis; 153; 411–416. Tang Z, Maroto-Valer M M and Zhang Y (2004a) CO2 capture using anthracite based sorbents. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 49 (1); 298–299. Tang Z, Zhang Y and Maroto-Valer M M (2004b) Study of the CO2 adsorption capacities of modified activated anthracites. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 49 (1); 308–309. Tarka T J, Ciferno J P, Gray M L and Fauth D J (2006a) CO2 capture systems utilizing amine enhanced solid sorbents. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 51 (1); 104–106. Tarka T J, Ciferno J P, Gray M L and Fauth D (2006b) CO2 capture systems using amine enhanced solid sorbents. In: Fifth Annual Conference on Carbon Capture & Sequestration, Alexandria, VA, 8–11 May, paper 152, available at: http://www. netl.doe.gov/publications/proceedings/06/carbon-seq/Tech%20Session%20152.pdf (accessed December 2009). Xu X, Song C, Andresen J M, Miller B G and Scaroni A W (2002a) Novel polyethyleneiminemodified mesoporous molecular sieve of MCM-41 type as high-capacity adsorbent for CO2 capture. Energy and Fuels; 16 (6); 1463–1469. Xu X, Andresen J M, Song C, Miller B G and Scaroni A W (2002b) Preparation of novel CO2 ‘molecular basket’ of polymer modified MCM-41. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 47 (1); 67–68. Xu X, Song C, Andresen J M, Miller B G and Scaroni A W (2003) Preparation and characterization of novel CO2 ‘molecular basket’ adsorbents based on polymer-modified mesoporous molecular sieve MCM-41. Microporous and Mesoporous Materials; 62 (1/2); 29–45. Xu X, Song C, Andresen J M, Miller B G and Scaroni A W (2004) Adsorption separation of CO2 from simulated flue gas mixtures by novel CO2 ‘molecular basket’ adsorbents. International Journal of Environmental Technology and Management; 4 (1/2); 32–52. Zheng F, Tran D N, Busche B, Fryxell G E, Addleman R S, Zemanian T S and Aardahl C L (2004) Ethylenediamine-modified SBA-15 as regenerable CO2 sorbents. Division of Fuel Chemistry, American Chemical Society – Preprints of Symposia; 49 (1); 261–262. Zheng F, Tran D N, Busche B J, Fryxell G E, Addleman R S, Zemanian T S and Aardahl C L (2005) Ethylenediamine-modified SBA-15 as regenerable CO2 sorbent. Industrial and Engineering Chemistry, Research; 44 (9); 3099–3105.


7

Advanced membrane separation processes and technology for carbon dioxide (CO2) capture in power plants A. B a s i l e and A. I u l i a n e l l i, Italian National Research Council, Italy, F. G a l l u cc i, University of Twente, the Netherlands and P. M o rr o n e, University of Calabria, Italy Abstract: (CO2) capture by membrane separation in order to produce clean fuel from a gas mixture (from coal gasification or steam reforming processes) can use either polymeric or inorganic membranes. Inorganic membranes are very attractive for CO2 removal in integrated gasification combined cycle (IGCC) power plants even though their costs are very high. Gas separation using polymeric membranes is commercially available. However, CO2 capture using polymeric membranes in large-scale power production is still inadequate, due to their poor high-temperature stability. CO2 capture using membrane technology is an innovative solution that could be applied to all types of power plant off-gases, its main benefit being the possibility of using membranes in combination with small-scale modular fuel cells. The main drawback of removing CO2 using commercially-available membranes is the higher energy penalties on power generation efficiency compared to conventional chemical absorption processes. Key words: CO2 capture, polymeric membrane, inorganic membrane, membrane module.

7.1

Introduction

The scientific community is becoming increasingly interested in technologies that capture and store carbon dioxide (CO2). This is confirmed by several ongoing research projects co-funded by various industrial companies and the many scientific papers and reviews on this field. Carapellucci and Milazzo1 point out that fossil-fuel power plants generate the largest proportion of CO2 emissions (33–40 % of the total CO2 emitted into the atmosphere), so it is important to separate CO2 from the flue gas of these types of power plants and capture it for storage. However, capture and storage of CO2 must be considered as a mid-term solution, to be used only until hydrogen or other renewable energy technologies are considered mature enough to be industrially exploited. Many papers and reviews have been published on the current status of carbon capture and storage technologies, see for example References 2–7. CO2 separation and capture can be subdivided into three main technological 203 © Woodhead Publishing Limited, 2010

204


pathways: post-combustion and pre-combustion processes and oxyfuel combustion power cycles. Other emerging technologies that are under study include chemical-looping combustion (for details see, for example,9) and membrane gas separation for CO2 removal in power generation.8,10–14 The principles of the three main CO2 capture options are described below and represented schematically in Fig. 7.1. ∑

Post-combustion capture: CO 2 is separated from other flue gas constituents (such as N2, NOx and SO2) that derive either from the air

Post-combustion

Fuel Air

Fuel conversion

Flue gas

CO2 separation

Steam turbines

CO2

Nitrogen

Power (a) Pre-combustion

Fuel O2 Air

Air Fuel conversion

CO2 separation

H2

Energy conversion

CO2

Air separation

Steam cycle

N2 (b) Oxy-combustion

Flue gas

Fuel O2 Air

Air separation N2

Fuel conversion

Power

CO2

Steam turbines Power (c)

7.1 Principles of three main CO2 capture options, (a) postcombustion, (b) pre-combustion, (c) oxy-combustion.


Advanced membrane separation processes and technology

205

or were produced during combustion. For combustion, existing power plants use air (composed of almost four-fifths N2) and produce a flue gas at atmospheric pressure with a CO2 concentration below 15 %. The driving pressure for CO2 capture is low: typically less than 0.15 atm. The traditional technique for post-combustion capture is chemical absorption, for example using monoethanolamine (MEA). This technique, widely used in the natural gas industry for over 60 years, produces a relatively pure CO2 gas stream. Amine solvents for CO2 are available in three forms (primary, secondary and tertiary), each one having its own advantages and disadvantages. This method suffers from a number of problems: (i) capital costs are high; (ii) the operation is complex and usually requires full-time supervision; (iii) maintenance is expensive and labour-intensive. The installation of membrane plants using CO2-selective cellulose acetate membranes began around 1980, mainly at small gas processing plants (less than 6000 Nm3/h), where amine systems would be too complex and expensive. Research is mainly aimed at improving the performance of the membranes by increasing selectivity and permeability, and also reducing their cost.15 ∑ Pre-combustion capture, for gasification or reforming processes: CO2 is removed from the fuel before combustion. The fuel is first converted into a mixture of CO2 and H2 via a reforming process, or to a mixture of CO and H2 via coal gasification process which is subsequently fed to a shift reaction section. Traditionally, chemical (such as MEA absorption) or physical processes (such as pressure swing adsorption) are used to capture CO2. New technologies being studied fall within the process categories of wet scrubbing with physical sorption, or chemi- or physic-sorption with solid sorbents. Membranes can also be used for this type of separation.16 Apart from membranes, all these techniques need regeneration. ∑ Oxyfuel combustion power cycles: the fuel is burned in an O2 stream that contains no (or little) N2. Oxygen is used as a fuel oxidising agent instead of air. In these plants, the main separation step is O2 from N2: pure O2 is first separated from air and then sent to the energy conversion unit. The combustion takes place in an O2/CO2 environment and the resulting flue gas is a high-purity CO2 stream. Although experience with O2 fuel combustion in the glass, steel and aluminium industries is well known, the oxyfuel concept has not yet been applied to power plants. For example, this option could not be considered viable for pulverised coal-fired power plants in the near future. Membranes can also be used to separate O2 from air in oxyfuel combustion power cycles, but this is outside the scope of this chapter. Some important aspects of these three processes are summarised in Table 7.1a. Pre- and post-combustion processes and oxyfuel combustion power © Woodhead Publishing Limited, 2010

Advantages

Disadvantages

Other considerations


Post- ∑ Natural gas-fired Chemical or ∑ Applicable to the Flue gas is diluted in CO2 ∑ Low-temperature process combustion combined cycle physical majority of existing at ambient pressure ∑ Production of relatively ∑ Coal-fired steam solvents coal-fired power resulting in: pure CO2 stream power plant scrubbing plants ∑ low CO2 partial pressure ∑ Both chemical and physical ∑ IGCC ∑ Retrofit technology ∑ CO2 produced at low absorption are mature option pressure compared to technologies ∑ It is considered a storage requirements ∑ Commercial polymeric promising long-term ∑ higher circulation membranes are available option (the required volume required for (polyimide, polysulfone, technology is under high capture levels polyether-polyamide development) copolymer, etc.) and can be used with an adsorption liquid (e.g. MEA) Pre- Natural gas-fired Chemical or Synthesis gas is ∑ Applicable mainly to ∑ combustion combined cycle physical concentrated in CO2 at new plants, as only with partial solvents high pressure and so few plants are now in oxidation or scrubbing results in: operation ∑ Barriers to commercial with methane ∑ high CO2 partial steam reforming pressure application of and subsequent ∑ high driving force gasification are common ∑ WGS reaction for separation to pre-combustion ∑ Oxygen blown ∑ more technologies capture: IGCC with available for s availability ∑ water gas shift separation s cost of equipment reactor s extensive supporting system requirements ∑

Technology for CO2 capture similar to post– combustion processes but smaller in size, potentially less expansive than postcombustion processes IGCC power plants applying this process are more efficient than pulverised coal-fired plants Both chemical and physical absorption are mature technologies The same commercial polymeric membranes used in post-combustion


Process Type of power plant CO2 capture method

206

Table 7.1a An overview of the three CO2 separation and capture processes

capture and also hydratebased separation can be used

Condensing of ∑ Very high CO2 ∑ Large cryogenic O2 ∑ Oxyfuel ∑ O2/CO2 combustion combustion water in the concentration in production requirement power in natural gas- exhaust gases, flue gas may be cost prohibitive ∑ cycles fired combined remaining dry ∑ Retrofit and ∑ Cooled CO2 recycle cycle with gas consists repowering required to maintain exhaust gas mainly of CO2 technology option temperatures within recirculation limits of combustor ∑ O2/CO2 materials combustion of ∑ Decreased process coal with efficiency exhaust gas ∑ Added auxiliary load recirculation ∑ This technique is ∑ ∑ Chemical- simpler and less looping chemically intensive combustion than post-combustion capture, but is less ∑ mature ∑ ∑

Oxygen from air separation unit The exhaust stream is free of N2, sulphur components and particulates. Exhaust gas stream is about 90 % CO2 (vol) on a dry basis Further separation of CO2 is not necessary and CO2 can be compressed for storage or transportation The main advantage is the elimination of NOx control equipment and the CO2 separation step Boiler size and SO2 scrubber are reduced The technology is not yet considered mature Polymeric air separation membranes can be used for the separation of oxygen from nitrogen (commercialise, for example, by DuPont, Air Liquid, etc.)



207

208


cycles all have significantly lowered thermal efficiency. Both material and operative costs for a power plant based on these processes would be quite high. Membranes can be applied in all the processes considered. Both polymeric and inorganic membranes can be used to produce clean fuel from a mixture gas (from coal gasification or steam reforming processes) via membrane separation.17 Inorganic membranes are very attractive for CO2 removal in IGCC power plants due to their simplicity, flexibility, ability to maintain high CO2 pressure and their potential for carrying out separations with low energy penalties. The main aspects of inorganic membranes used for H2, O2 and CO2 separations are shown in Table 7.1b. The costs of inorganic membranes are generally very high, so only polymeric membranes have been used commercially for gas separation to date. However, CO2 capture using polymeric membranes in large-scale power production is inadequate. Syngas is delivered at pressures and temperatures that depend on the type of fuel processor used, but polymeric membranes lack high-temperature stability and so the flue gases have to be cooled. Polymeric membranes also show inadequate performance in terms of both permeability and actual selectivities, which are lower due to plasticization effects.1

7.2

Cryogenic carbon dioxide (CO2) capture

Among the different CO2 capture technologies proposed in the literature, an interesting technique is to freeze CO2 out at low pressure. This method is called cryogenic CO2 capture. However, there are few papers dealing with this topic. Eide et al.18 described the process as ‘anti-sublimation’, where the CO2 is frozen on a cold surface at atmospheric pressure. Sublimation is the phase change from solid to gas, whereas the reverse phase change from gas to solid does not have well defined designation, thus ‘anti-sublimation’ seems to be an appropriate definition. The freezing point of CO2 at atmospheric pressure is –78.5 °C. Practical applications have to take into account the fact that the freezing temperature depends on the concentration of CO2 in the feed gas. Table 7.2 can be used to help with estimating the temperatures at which cryogenic CO2 capture should be performed. In post-combustion flue gas, the CO2 content is roughly 15 %, so to capture 99.99 % of the CO2 in the flue gas the process should be performed between –99.3 °C and –155.8 °C. The scheme proposed by Eide and co-workers refers to CO2 capture using a refrigerant which is then cooled down and reused in the process. Recovering all the energy from the re-sublimation of CO2 and so on is quite complex. In practice, the system requires that the flue gas is first cooled to near ambient temperature, which leads to a first capture of water by condensation. Then the flue gas is cooled further, to near 0 °C, by a (or additional) refrigeration


Table 7.1b Main characteristics of inorganic membranes applied to H2, O2 and CO2 separations Transport mechanism

Flux and selectivity

Commercial use

Notes


Inorganic membranes are much more expansive than the polymeric ones. Vice versa, due to their greater flexibility, inorganic membranes are considered more promising for the CO2 capture Non-palladium-based metal alloy materials are under development for reducing the cost, increasing the flux, and improving the durability of H2 separation membranes The high-temperature operation makes the robust design of membrane support, module and housing more challenging The tolerance to contaminants in the coalderived syngas is a big problem, especially the stability of these membranes in the presence of CO2 Dense perovskite membranes are still in an early stage of development

209

Dense Pd-based and H2 ∑ Solution-diffusion The H2 flux Small-scale ∑ membranes refractory separations mechanism: is quite low, palladium metals in hydrogen diffuses in while the H2 membranes are group IV (Ti, the atomic form selectivity is commercially Zr, Hf) and ∑ The H2 flux through a infinite available for group V thick Pd-based hydrogen purification (Nb, Ta, V) membrane generally in electronic elements conforming to industry Richardson’s law ∑ Dense O2 or H2 ∑ When a H2- mixture is The H2 (or O2) Oxygen production ∑ electrolytes separation introduced from one flux is very from air is currently and mixed side of the membrane, low. used in many conducting H2 dissociates into The H2 (or O2) applications: (ionic and proton (H+) and electron selectivity is medical devices, electronic) (e+) on the surface. very high steel industry, ∑ membranes These species are chemical transported to the manufacturing, etc. opposite side of the membrane and recombine to H2 molecule ∑ ∑ In the case of perovskites, oxygen


Membrane Materials Separation

210

Table 7.1b Continued Transport mechanism

Flux and selectivity

∑

ions and electrons are transported in a counter-current mode and no external electrons are needed: the oxygen partial pressure difference through the membrane acts as a driving force Dense oxide membranes require high temperatures (> 700 °C) for good O2 fluxes

Commercial use

Notes

High flux but Examples of ∑ Microporous A microporous H2 or CO2 There are many composite thin top layer separation transport mechanisms: less selective commercial membranes is casted on a (Oxygen Poiseuille, Knudsen, porous inorganic porous support. separation molecular sieving, membranes are: Alumina, is relatively surface diffusion, ceramic membranes carbon, glass, difficult in multilayer diffusion, (alumina, silica), silicon carbide, porous capillary condensation glass and porous titania, zeolite membranes) metals (stainless and zirconia steel and silver) are mainly ∑ used on substrates such as a-Al2O3,

These membranes are generally cheaper than inorganic dense ones. They are able to operate at high temperatures, so potentially interesting for the separation of CO2 from H2 in syngas processes at high temperatures Inorganic silica and zeolite membranes, showing hightemperature and highpressure stability, hold the potential for full and near-



Membrane Materials Separation

term industrial implementation Due to their ease of fabrication, and scalability silica membranes are good candidates for hydrogen separation, being also less expensive than metals and not susceptible to H2 embrittlement. Similarly, zeolite membranes have inherent chemical, mechanical, and thermal stability.



g-Al2O3, zirconia, zeolite, stainless steel ∑ ∑

211

212

Developments and innovation in CCS technology Table 7.2 Frosting temperatures as a function of the CO2 concentration in the flue gases CO2 Conc. v/v

Tc (°C)

100 15 10 5 2 1 0.1 0.002

–78.5 –99.3 –103.1 –109.3 –116.7 –121.9 –136.7 –155.8

system, which recovers more water by condensation. The flue gas next enters a heat exchanger in which the temperature is decreased to about –15 °C (more water is frosted out). The exchange takes place with the ‘clean’ flue gas, which needs to be heated up before leaving the recovery system. The temperature is lowered to –40 °C by a refrigerant (at this temperature all the remaining water is recovered). The temperature is further reduced in a new heat exchanger (with clean flue gas), down to approximately –100 °C. The flue gas now enters the low temperature frost evaporators (LTFE). Different LTFEs (at least two) are operated in parallel in order to perform a continuous operation. When one LTFE is in the frosting cycle, the other one is in the defrosting cycle. Different refrigerants and cycles can be used. As reported by Eide, the typical energy penalty of this type of CO2 capture route is around 4–8 % for a pulverised coal boiler. The energy penalty can be reduced substantially if cryogenic CO2 capture is coupled with, for example, LNG re-gasification sites in order to exploit the cold available in these plants, as suggested by Tuinier et al.19 The authors also proposed a novel concept for separating CO2 from flue gas using dynamically operated packed beds. With this concept, the authors claim to be able to circumvent the drawbacks of the process described above, which are essentially due to possible plugging (by ice) of the cooling units with subsequent increase in pressure drop. Process efficiency is also reduced because the rate of heat transfer inside the LTFE decreases as solid CO2 builds up on the exchange surfaces. Last but not least, the heat exchangers need to be operated at different temperatures (between capture and regeneration), so they are exposed to large mechanical stresses. Cryogenic CO2 capture is based on the difference in dew and sublimation points of the different gases in a mixture. In practice, the gas is fed into a very cold packed bed. As the gas reaches the bed, it starts to refrigerate and the packing starts to heat up. At a certain point water starts to condense, thus there is a condensation front moving inside the bed. However, the packing containing condensed water continues to heat up and the water starts to



213

evaporate again, so there will also be an evaporation front moving through the bed. The condensation front is faster than the evaporation front due to the difference in enthalpy during condensation/evaporation. Over a certain length of bed, the water is completely condensed and the gas is further cooled so that CO2 starts to freeze on the surface of the bed, and a desublimation front also moves through the bed. When the desublimation temperature is reached in the bed (and CO2 is deposited on the packing) the bed is switched and a pure CO2 stream is used to sublimate (and recover) the CO2 on the bed. When the CO2 is recovered, then the bed is switched to a cooling cycle and the process can start again. Water condensation and evaporation avoid any problem of excessive pressure drop in the bed. The authors proposed a theoretical study followed by an experimental demonstration of the principle. More details can be found in Tuinier, et al.19 These two examples show that it is possible to effectively separate CO2 from flue gas using a cryogenic route. Whether this route is more advantageous than membrane separation can only be decided based on energetic (and exergetic) analysis coupled with a cost analysis, which is not yet possible because the two separation techniques are at very different stages of research. Since water plugging is the most critical aspect of cryogenic CO2 capture, membranes can be used to improve this technique. In particular, a hydrophilic membrane separator can be used before the cryogenic system in order to recover the water contained in the flue gas. A zeolite membrane with pore size comparable to the molecular diameter of water can be used for this, and if the pore size of the zeolite is close to the molecular diameter of water, the latter will condense in the pore via capillary condensation. The water condensed in the pores will prevent permeation of gases so that only water will permeate through the membrane, resulting in dehumidification of the flue gas. By separating water from the flue gases, cryogenic CO2 capture will be easier and the energy penalty reduced.

7.3

Performance of membrane systems

Gas separation using membranes is a highly attractive, energy efficient technique for CO2 capture.20, 21 A membrane is a physical device able to remove selectively one or more components from a mixture while rejecting others. Membrane gas separation shows different advantages over conventional processes and has been well described in many excellent reviews (see for example References 22–24). Membrane separation processes are used today in bulk chemistry as well as in the petrochemical sector. Membranes offer several advantages, including their small size, simplicity of operation and maintenance, compatibility, diversity and lack of pollutant by-products. The main membrane separation techniques are: reverse osmosis, nano- ultra- and microfiltration, pervaporation, gas separation, vapour © Woodhead Publishing Limited, 2010

214


permeation and electrodialysis. The ‘standard’ membrane processes (reverse osmosis, nano- ultra- and microfiltration) are now reasonably commonplace in the majority of chemical sectors.25 Gas separation membranes are used in many industrial processes, such as the production of air enriched with oxygen, separation of CO2 and H2O from natural gas, purification of H2 and so on. Various reviews and books can be found in the literature.21, 26–29 In 1992, studying the application of polymeric membranes for recovering CO2 from the flue gas of a power plant, Feron et al.30 and Van Der Sluijs et al.31 showed that up to 76 % CO2 removal is achievable and, moreover, that the economic competitiveness of the process depends on the membrane used, in particular, on its selectivity to gas transport. In separating one or more gases from a feed mixture and generating a specific gas-rich permeate, a membrane acts as a ‘filter’ that allows the preferential passage of certain substances. In particular, a membrane will separate gases only if some components of the mixture are able to pass through the membrane more rapidly than others. In other words, the flux of the gas to be separated (in our case CO2, preferentially in CO2-rich feed mixture) should be higher than all the others (under the same conditions). Membranes can be separated into two types: porous and non-porous (or dense) membranes. Porous membranes separate gases through small pores in the membrane based on molecular size. Non-porous or asymmetric membranes, which separate based on solubility and diffusivity, are more commonly used for gas separation, e.g. in natural gas applications. For both porous and non-porous membranes, there are many possible separation mechanisms, but only six of them are considered important for gas separation: Knudsen diffusion, molecular sieving, surface diffusion, facilitated transport, capillary condensation and solution-diffusion separation.28, 32–36 Among them, solution-diffusion is the most appropriate process for CO2 separation in polymeric membranes.35 In dense membranes, the gas transport is based on a solution-diffusion mechanism and results in selective transport of gases and, consequently, their separation. It is interesting to observe that there is a trade-off between selectivity and permeability: membranes with a high selectivity show low permeability, and vice versa.37 Figure 7.2 indicates the ideal CO2/N2 selectivity versus fast component CO2 permeability of polymeric membranes. Compared to other separations, such as O2–N2 and CO2–CH4 mixtures, CO2–N2 mixture appears to be an easier separation. In 1991, Robeson suggested that this trade-off possesses an upper bound.37 Figure 7.3 shows an example of this upper bound, for a range of glassy and rubbery membranes involved in CO2/CH4 separations. Since Robeson’s paper, only a few examples of polymeric membranes that exceed the upper bound have been published, and overcoming it is the focus of many recently awarded patents in polymeric membranes. The aim is to achieve both high CO2 permeability and selectivity. For example, the polybenzimidazole © Woodhead Publishing Limited, 2010


215

50

CO2/N2 perm-selectivity

40

30

20

10

0

0

100

200

300 400 CO2 permeability

500

600

7.2 CO2/N2 selectivity versus the CO2 permeability of polymeric membranes.

CO2/CH4 perm-selectivity

100

10

Glass Rubber 1 1e-18

1e-17

1e-16 1e-15 1e-14 1e-13 CO2 permeability, [mol m/(m2sPa)]

1e-12

1e-11

7.3 Comparison of Robeson’s curve for CO2/CH4 separation by glass (o) and rubber () membranes.

membrane exceeds the Robeson upper bound for H2/CO2 selectivity versus H2 permeability in the temperature range 100–400 °C. Recently, Koros and Mahajan suggested the possibility of exceeding the upper bound using mixed-matrix membranes.38


216


Apart from permeability and selectivity, other membrane properties are also very important, such as their thermal, chemical and plasticisation resistance, and ageing affects when continued performance is required over long periods of time. Cost-effectiveness in manufacturing standardised membrane modules is also important. Considerable experimental research has been addressed to meeting these aims. An extensive review describing original polymeric and inorganic membrane patents was recently published by Scholes et al.39 with particular attention paid to CO2 separation through polymeric membrane systems in flue gas applications. This review is particularly interesting because it focuses on recent novel approaches in polymeric membranes that achieve separation performance above Robeson’s upper bound and therefore are potentially more commercially competitive than present membrane gas separation technologies. Another important extensive review of polymeric CO2/N2 gas separation membranes for the capture of CO2 from power plant flue gases was recently published by Powel and Qiao.36 In this review, a chemist’s perspective is taken, i.e. the gas permeability properties of dense membranes are seen in the light of their chemical structure. The Carapellucci and Milazzo1 paper already cited above investigates the engineering aspects of using membranes for CO2 separation in flue gases. Some important aspects of these three papers regarding polymeric membranes will be considered here.

7.4

Carbon dioxide (CO2) membrane materials and design

As reported recently by Baker29 in his excellent review, the market for membrane acid gas separation systems can be classified into three categories: very small systems (< ~ 140k m 3/d): membrane units are very attractive; ∑ small systems (~ 140k–1133k m3/d), where two-stage membrane systems are used to reduce methane loss – amine and membrane systems compete in this range and the choice depends on site-specific factors; ∑ medium to large systems (> ~ 1133k m3/d), where membrane systems are generally too expensive to compete with amine plants.

∑

A low-cost alternative to amine plants could involve combining membrane systems to remove the bulk of the CO2 with amine plants to act as polishing systems, but the savings in capital cost are largely offset by the increased complexity of the plant due to the presence of two different separation processes.40 It must be concluded that membrane systems cannot compete with current amine systems for most CO2 removal applications. The main problem is



217

the relatively low selectivity and permeability of current membranes. For example, under normal operating conditions, cellulose acetate membranes show an ideal gas selectivity (defined as the ratio of the pure CO2 and CH4 permeabilities) aCO2/CH4 = 12–15, which is quite low and reflects plasticisation effects by CO2. There are other issues associated with the capture of CO2 from flue gas that limit the use of membranes: ∑

CO2 concentration is low, and consequently a very large amount of flue gas must be processed. ∑ Flue gas is at a high temperature and membranes are easily destroyed, so it is necessary to cool the gas mixture below 100 °C before sending it to the membranes for separation. ∑ Membranes must be chemically resistant to aggressive gases present in the flue gas mixture. ∑ In order to improve the performance of the membrane, it is necessary to improve the pressure difference across the membrane itself and this requires a significant amount of power.

A number of polymers for CO2–N2 membrane separation have been studied, some of which are shown in Table 7.3, which reports performance in terms of single gas permeance and CO2/N2 selectivity. These membranes could be used for separating post-combustion flue gas mixtures with CO2 and N2 as the main components. Considering that PCO2 = S CO2·DCO2, (permeability = solubility * diffusivity) and in order to increase the permeability of CO2 in a polymeric membrane, Table 7.3 Performance of polymeric membranes in terms of CO2 permeance and CO2/N2 selectivity. Data adopted from refs [13, 37, 41] Material

CO2/N2 selectivity

CO2 permeance [m3·m–2·Pa–1·s–1]

Polydimethylsiloxane Polydimethylphenilene oxide Poly(4-vinylpyridine)/polyetherimide Polyethersulfone Polyacrylonitrile with ethylene glycol Polysulfone Polyimide Poly(ethylene oxide) Poly(amide-6-b-ethylene oxide) Polyvinyl alcohol (cross-linked) Vinyl alcohol / acrylate copolymer – FT Polyvinyl alcohol (cross-linked formaldehyde)

11.4 19 20 25 28 31 43 52 61 170 1417 1782

32001 2750 52 665 91 450 735 52 608 82781 24001 3381

1

Data are in Barrer. FT = facilitate transport membrane.


218


research is exploring increases in CO2 diffusivity through the polymeric structure and increases in CO2 solubility in the membrane matrix.39 Another possible avenue is the preparation of new membranes, such as mixed-matrix, hybrid, facilitated transport and membrane contactors.42–44

7.4.1 Mixed-matrix membranes Membrane separation processes have several advantages over conventional separation techniques, such as energy-saving, space-saving and being easy to scale-up. Unfortunately, a membrane able to combine high flux, high selectivity and high stability is still not realistic. For CO2 separation, mixed-matrix membranes could be the best solution to these problems in the near future. This new type of membrane, which combines the advantages of polymeric and inorganic membranes, seems to indicate superior performance. However, there are numerous problems regarding their processing ability. For this reason, Nomura et al.45 modified silicalite membranes to incorporate solid particles within a polymeric membrane, obtaining some improvements in the selectivity of the molecular sieve membranes. Powell36 reported that the presence of solid particles in a polymeric matrix can modify the permeability via three different effects: ∑

by acting as molecular sieves (and so altering the permeability of the preceding polymeric membrane); ∑ by disrupting the polymeric structure (and so increasing permeability); ∑ by acting as barrier (and so reducing permeability). This new polymer–zeolite research area includes the following pairs:

∑ ∑ ∑ ∑ ∑ ∑ ∑

poly(ethylene oxide)–various nanoparticles46, 47 polydimethylsiloxane–silicalite;48 polyimide–carbon molecular sieve;49 polyimide–silica;50 Nafion® (Du Pont)–zirconium oxide;42 HSSZ-13–polyetherimide;51 acrylonitrile butadiene styrene-activated carbon.52

Two interesting papers on these membranes were published by Zimmerman et al.53 and Mahajan et al.54 a decade ago, in which more details can be found.

7.4.2 Hybrid membranes Hybrid membranes offer some advantages over porous inorganic ones. In this case, the surface of a porous inorganic support material is chemically



219

modified to improve affinity with CO2, i.e. a surface-modified inorganic membrane. The working concept of these membranes is quite simple: the porous support has a low resistance to CO2 flux transport, whereas the chemical modifications are devoted to better controlling the selectivity.43, 55 This new chemical-inorganic support research area includes the following pairs: ∑ ∑ ∑ ∑ ∑ ∑ ∑

trichlorosilane–g–alumina;43 polyether–silica;44 organosilane–Vycor® (Corning Inc.) glass;56 tetrapropylammonium–silica;57 trimethoxysilane–titania;58 trimethoxysilane–g–alumina;58 aminopropylhydroxysilyl–hexagonal mesoporous silica.59, 60

The potential of hybrid membranes for CO2 capture in an integrated gasification combined cycle was evaluated recently by Luebke et al43. Unfortunately, the authors concluded that, despite the high performance of their membranes being attractive, selectivity was not good enough to make them competitive with other processes.

7.4.3 Facilitated transport membranes The first facilitated transport membrane patent was awarded to General Electric in 1967.39 Facilitated transport membranes show high selectivity and high flux55 and represent one way to circumvent the Robeson flux/selectivity trade-off. These membranes are obtained by incorporating a carrier (pure water, glycine, etc.) in a membrane able to react with the gas of interest. This reaction is reversible, so the gas at first dissolves in the membrane and then is selectively transported by the carrier via diffusion from one face (feed side) of the membrane to the opposite face (downstream side). Here, the gas is released, while the carrier agent is recovered then diffused back to the feed side. The driving force for gas transportation is the partial pressure difference across the membrane. The carrier increases both the permeability and selectivity of the membrane. It can be either fixed within the polymeric membrane or in a mobile matrix. Facilitated transport membranes take the form of fixed carrier membranes, solvent-swollen polymer membranes and mobile carrier membranes.55 Table 7.4 compares the performance of these membranes with others in terms of CO2 permeability and CO2/N2 selectivity. Among the problems associated with these membranes are poor chemical stability of the carriers, evaporation and degradation related to immobilised liquid membranes and short lifetime (at best a month, according to the literature).


220

Developments and innovation in CCS technology Table 7.4 Specific surface area of some contactors Contactor

Specific surface area (m2/m3)

Free dispersion column Packed column Packed/trayed column Mechanically agitated column Membrane contactor

1–10 100~800 10~100 50~150 1500~3000

Liquid-gas interface Solvent inlet

CO2

Membrane

Flue gas inlet

7.4 CO2-membrane gas absorption principle.

7.4.4 Membrane contactors Membrane contactors are now considered one of the best ways of controlling CO2 emissions using membrane technology.61 The best-established among CO2 separation techniques (chemical and physical absorption, solid adsorption, carbon molecular sieve adsorption, cryogenic distillation and membrane separation) is that which uses absorption into alkanolamine solutions using conventional contactor equipment (for example, packed columns or tray columns). Unfortunately, this method is energy-consuming and not easy to operate because of problems due to foaming, flooding, channelling and so on. Membrane gas absorption technology combines membrane separation and chemical absorption technologies, i.e. the separation role is fulfilled by the absorption liquid. A conceptual schema of a membrane contactor is shown in Fig. 7.4. In an ideal situation, all the pores of the membrane should be completely filled by gas in order to minimise mass transfer resistance, and so the membrane



221

itself does not separates the gases. As a consequence, the reactive absorption liquids in these systems are preferably physically reactive liquids because of their higher absorption rate and capacity62. The main advantages of these contactors over conventional column contactors are the reduction in size, operational flexibility, elevated mass transfer rate and linear scale-up. For example, the membrane surface area of commercial hollow fibre membrane modules ranges between 1500–3000 m2/m3 of contactor volume,63 whereas in conventional contactors (bubble column, packed and plate columns), it is in the range 100–800 m2/m3, as displayed in Table 7.4.61 The higher membrane surface area of commercial hollow fibre membrane modules therefore results in a significantly smaller contactor. These advantages over conventional contactors were seen by Gabelman and Hwang64 as another interesting potential possibility for recovering CO2 from flue gas, natural gas and gas streams from other industrial processes. Membrane contactors need to be tested on a larger scale and for long-term, stable operation. For the latter, it is important that the pores of the membrane remain completely gas filled (i.e. non-wetted); only in this way can an increase in overall mass transfer resistance be prevented. To overcome wetting, new absorption liquids for CO2 removal using membrane contactors are being studied. For example, Kumar et al.63 investigated a new absorption liquid based on the alkaline salts of amino acid in a single membrane gas–liquid contactor. The liquid prevents the wetting of commercial polypropylene membranes and the results seem very promising for CO2 removal. An interesting review of CO2 absorption using chemical solvents in a hollow fibre membrane contactor was recently published by Li and Chen. 65

7.5

Membrane modules

In order to use a membrane in a separation process, it must be packed into a proper device, known as a membrane module. A conceptual scheme of a membrane module is shown in Fig. 7.5. The requirements that a membrane module must meet are: low production costs, high packing density, low energy consumption and good control of concentration polarisation. In the practical applications of membrane separation processes, the modules are quite different.66–69 Stirred batch cells or deadend filter cells are used in small-scale laboratory applications, whereas two types of configurations are used in large-scale industrial applications: flat (plate-and-frame, spiral wound and disc tube modules) and tubular (tubular, capillary and hollow fibre modules). The membrane modules used in gas separations are spiral wound, capillary and hollow fibre and are shown schematically in Figs 7.6–7.8. As can be observed, the design of these modules is quite different as are their modes of operation, production costs and energy requirements (mainly owing to © Woodhead Publishing Limited, 2010

222

Developments and innovation in CCS technology Membrane separation process PH

Retentate

Feed

PL < PH

Permeate

Membrane

Permeability

7.5 Conceptual scheme of a membrane module.

7.6 Capillary module. Extruded or spun fibres with membrane on inside circumference (often also on outside).



Retentate

223

Drain for backflush

Permeate

Entry for backflush

Feed

Drain for backflush

7.7 Hollow fibre module. Fine fibres with shell side feed.

pressure drop inside the membrane module). Table 7.5 lists the principal characteristics of some CO2 separation membrane modules. It should be pointed out that there is no single membrane module that can solve all the problems. The commercial membrane modules available today are each designed for a specific membrane process application.

7.6

Comparing membrane modules

Many factors must be considered in choosing the most suitable module, including the particular CO2 separation to be performed, manufacturing cost, concentration polarisation and fouling (if any), packing density, cleanability, permeate side pressure drop, high-pressure operation and limitation of type of material. Depending on the membrane processes considered, several types of modules can be used (see Table 7.6). The following trend in packing density is observed for commerciallyavailable polymeric membrane modules: hollow fibre > capillary > spiral wound > plate and frame > tubular. The packing density of commerciallyavailable inorganic membrane modules is summarised in.70 No commercial inorganic membranes are yet available in the high packing density, spiral wound and hollow fibres. © Woodhead Publishing Limited, 2010

solu

224

d Fee

tion

Feed channel spacer

Anti-telescoping device


Pem

eat

c Con

Membrane

e

ent

rate

Permeate spacer

Permeate collection material Membrane

Permeate collection tube

Feed channel spacer Outer wrap

Section

Outer cover Feed spacer

Permeate flow

Arrows indicate permeate flow

7.8 Spiral-wound module. Membrane and permeate channel wound around a permeate pipe.


Perforated central tube


225

Table 7.5 Membrane equipment configuration Membrane module

Characteristics

Capillary (Fig. 7.6)

∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑

Cleanable by reversing permeate flow Well-developed equipment Low area cost Low holdup volume Pre-treatment required to prevent plugging Intolerant of capillary rupture Operating power range: 200 watts/m2 Flux range: 20–50 l/m2 · hr Energy requirement: 4–10 watt · hr/lpermeate

Hollow fibre (Fig. 7.7)

∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑

Low area cost – (best) Compact – (best) Low holdup volume Well developed Somewhat tolerant of fibre rupture High pressure capability Sensitive to dirty feed – (worst) Poor cleanability – (worst) Energy requirement assuming no power recovery: pressure (psi) watt · hr 4 · 10–3 · · liter permeate conversion

Spiral wound (Fig. 7.8)

∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑ ∑

Inexpensive, mature hardware – (best) Compact, low holdup Wide pressure range High temperature possible Pre-treatment required Difficult to clean Operating power range: 20–70 watts/m2 Flux range: 10–50 l/m2 · hr Energy requirement: 1–6 watt· hr/lpermeate

It is difficult to correctly quantify the cost of a module because module design varies widely depending on the application. Generally, hollow fibre modules are cheaper than the others even if produced for very high-volume applications in order to justify the expense in developing and building the spinning and module fabrication equipment. In Table 7.7 the module manufacturing cost per square meter of membrane is shown (selling costs are usually two to five times higher than manufacturing costs). High-pressure modules are more expensive than low-pressure or vacuum modules. Membrane modules with high packing density, i.e. hollow fibre or spiral-wound, are most commonly used for gas separation.

7.7

Design for power plant integration

Simulation models for gas separation membranes have been developed by a number of investigators. For example, Corti et al.71 recently simulated © Woodhead Publishing Limited, 2010

226


Table 7.6 Commercially available CO2-separation membrane modules, their costs, control of concentration polarization and applications Membrane Cost Characteristics Packing module density [m2/m3]

Control of Applications concentration polarisation

Spiral Low Flat sheet 700–1000 Good wound Capillary Low i.d. 110 bar 3

607

8 130 30

30

0 –40 000 –30 000 –20 000 –10 000

700

707 7

Internal regeneration

Recycled CO2

142

6

O2

Fuel

Fuel LHV = 45 MJ/kg Preheat = 40 bar Pupper = 110 bar TiT = 1300 °C TET = 700 °C h = 45 %

299

127 30

127 1

306 H2O + (dissolved CO2)

0 10 000 20 000 30 000 40 000 50 000 60 000 S (J/kmol.K) 1–2: intercooled staged compressor; 2–3: upper pressure cycle; 3–4: HP combustor chamber; 4–5: HP expander; 5–6: LP combustion chamber; 6–7: LP expander; 7–8 internal regeneration. 8–1: water cooler/separator.

9.6 T–S diagram of a gas turbine Matiant cycle with regeneration and reheat.

In order to make efficient use of the heat at a temperature higher than 700 °C at the turbine exhaust, a solid oxide fuel cell (SOFC) was integrated into the cycle (Demaret and Mathieu, 2001). A coal-fired cycle with integrated gasification (IGCC–Matiant) has also been developed (Mathieu and van Loo, 2005). At 1250 °C and 120 bar the calculated efficiency is 44.8 %, which is rather high for a coal-fired ZEPP because it benefits from the already existing air separation unit (ASU) in an IGCC and uses the most advanced CC. The Graz cycle The Graz cycle was first introduced in 1995 (Jericha et al., 2003) and has been continually developed at Graz University of Technology (Austria) and adapted from a hydrogen/oxygen cycle. The Graz cycle is a mixed Brayton/ Rankine cycle (see Fig. 9.7). However the proportion of water in the mixture is higher: 55 % steam by volume and 45 % CO2 at the combustion chamber exhaust. The stream at 40 bar is then expanded in two turbines: a first one at high temperature from the turbine inlet temperature (TIT) (1300–1400 °C) down to 600–700 °C/1 bar; after that, the gas flows through a heat recovery boiler HRSG and is expanded in a second turbine at low temperature, from 160 °C/1 bar down to the condenser at 0.25 bar. There, the steam is condensed and the water fraction generated in the combustion is extracted. The remaining water is pumped at high pressure (180 bar) and sent to the


300

Developments and innovation in CCS technology Steam

Fuel

Combustion Flue chamber gas

HTT

1 bar 642 °C

HRSG

160 °C

LPT

(1400 °C) CO2 recirculation

O2

567 °C 180 bar

HPT

0.25 bar

Feed pump Deaerator

CO2

H2O Cooler Condenser Excess H2O removal

CP 3

Regenerator

CO2 to storage

CP 2

CP 4

CP 1

Excess CO2 removal

9.7 Layout of a Graz cycle-based power plant with typical figures (HPT = high pressure turbine; HTT = high-temperature turbine; LPT = low-pressure turbine).

HRSG where it is vaporised and superheated (550–600 °C), then expanded in a steam turbine down to 40 bar and sent back to the combustion chamber. The other fraction, gaseous CO2, is compressed in a staged compressor and sent back to the combustion chamber. The CO2 generated in combustion is extracted, behind the first compressor (1 bar). Steam is used for turbine cooling and O2 is generated externally in an ASU. The efficiency of this cycle is about 49 %. This CO2 has still to be compressed up to 110 bar for transportation. Steam is recycled to the combustion chamber, similarly to CES cycle but with a higher complexity. Jericha et al. (2003) presented the design of 600 MWe Graz cycle power plant.

9.3.5 Oxyfuel combustion in gas turbines with internally generated oxygen The advanced zero emission plant (AZEP) In this cycle, the working fluid of the gas turbine is air, compressed in the compressor, heated in heat exchangers and expanded in a turbine. Here, the O2 is generated, inside the cycle, in a high-temperature ceramic mixed conducting membrane operating at about 900 °C (Betz et al., 2009). On one


Oxyfuel combustion systems and technology

301

side of that membrane, compressed air flows to a turbine. On the other side, a fraction of the hot flue gas CO2 + H2O coming from the combustion of natural gas flows along the membrane at high temperature and is recycled to the combustion chamber (Bouwmeester et al., 2002; Bredesen et al., 2004). This hot gas warms up the membrane to the right temperature at which the ionisation of O2 takes place so that a fraction of the O2 in the air flow migrates through the membrane (Griffin et al., 2003). N2 remains entirely in the ‘air side’. The O2 so pumped from air is carried by the flue gas to the combustion chamber where it oxidises the gas fuel (see Fig. 9.8a). The combustion chamber of the gas turbine is replaced by a reactor comprising the membrane and heat exchangers before and behind it, in order to ensure the heat transfer between the hot flue gas and the compressed air which is heated up while its O2 partly crosses the membrane. The depleted air is then expanded in the turbine. Its heat is given up in a heat recovery boiler HRSG to a steam cycle. The second fraction of the flue gas, composed of the combustion products, which is not sent back to the reactor, is expanded in a CO2/steam turbine and then separated into water and CO2 in a condenser (see

CO2 + H2O + O2

Sweep gas CO2 + H2O

Porous carrier

900 °C

O2–

Dense membrane Air

450 °C 10 % O2 Air 400 °C 20 % O2

O2

Depleted air

Oxygen ionisation

950 °C HX 1

O2 1050 °C

O2 MCM

900 °C

HX 2 1000 °C

Flue gas 1300 °C ~0 % O2 Depleted air 1250 °C 10 % O2

(a)

9.8 (a) Mixed ceramic membrane ionizing oxygen and conducting the ions through a porous carrier from compressed air into hot exhaust gas. (b), AZEP cycle using a mixed conductive membrane (MCM) and heat exchangers HX. The air bled from the compressor is divided into two fractions heated up separately by the combustion flue gas and remixed at the turbine inlet. The flue gas is also divided into 2 fractions, one being used as the sweep gas in the membrane MCM and the other one being cooled for water removal.


302

Developments and innovation in CCS technology CO2

CO2 + H2O

Excess CO2 + H2O

Cooler Condenser

CP

Steam cycle

HRSG

HX

Fuel CO2 + H2O + O2

Combustion chamber

H 2O

CO2 to storage

O2 O2 MCM

HX 1

Flue gas CO2 + H2O HX 2 Air Fuel

CP

SF Turbine

Steam cycle

Oxygen depleted air

HRSG

SF: with a supplementary firing (AZEP 85 %): Air Fuel

SF CO2 + H2O + N2 (b)

9.8 Continued

Fig. 9.8b). CO2 is compressed up to the delivery pressure in the transportation pipes. In order to avoid the use of a CO2 gas turbine, a modified version uses a heat recovery boiler HRSG feeding the steam cycle. A fraction of the compressed air goes to the membrane and is depleted while the other fraction goes to a heat exchanger where it is heated by the CO2/H2O flue gas stream. Both heated air fractions are remixed at turbine inlet and flow through the heat recovery boiler (see Fig. 9.8a). The efficiency of this cycle is about 50 % LHV, taking CO2 compression into account. In order to take full advantage of the performance of the most advanced gas turbines, the TIT is increased by adding an afterburner behind the reactor, fired with natural gas in air. The efficiency climbs then to 52 %, but the system is now no longer ‘zero emission’, some 15 % of the CO2 generated in combustion is not captured and released at the stack.



303

Chemical-looping combustion (CLC) The basic idea of chemical-looping combustion is to bring O2 into contact with the fuel using the oxidation of a metal (Fe, Ni, Cu, Mn…) followed by the reduction of its oxide which plays the role of an O2 carrier from air to fuel (Cho et al., 2002). Similarly to a membrane, the O2 is transferred internally in the cycle. In one variant of CLC (see Chapter 11), a metal is oxidised in contact with compressed air at a temperature of about 900 °C (generally in the range 800–1200 °C) in a reactor. The depleted air (still about 14 % O2) is then expanded in a turbine, similarly to the AZEP cycle (see Fig. 9.9). At these temperatures the formation of NOx remains low. The metal oxide, generally fluidised in small particles, goes into a second reactor where the fuel reduces the metal oxide into CO2 and water and the original metal. This last is recycled to the oxidation reactor, making a loop in which the metal carries O2 picked up from air into contact with the fuel, whereas the CO2/H2O mixture is expanded in a turbine (Brandvoll and Bolland, 2004). The combustion chamber is here replaced by the reduction/oxidation reactors. In the oxidation reactor, heat is generated by the exothermic reaction of metal with O2 and transferred to the reduction reactor where the fuel reduction is generally endothermic. The depleted air and the flue gas from the reduction reactor, heated at high temperature, are expanded into air and CO2 gas turbines, respectively. The temperatures and heat transfers are controlled by the recycle rate of solid particles between the two reactors and the average solids residence time in each reactor. The result of separating the combustion process into two reactors is that CO2 is not diluted in N2 so that it can be extracted from the mixture with water by cooling and condensation as in the previous oxyfuel cycles. CO 2 is then dried and compressed up to 110 bar. Air is used for turbine cooling. An improvement may be obtained by using the heat at the depleted air turbine exhaust in a heat recovery boiler generating steam in a Rankine cycle. Calculations show that the efficiencies in this case are about 45 % LHV when the temperatures in the reactors are 900 °C. When two oxidation/reduction systems with a single reheat are used, the first one operates at 1200 °C and high pressure, and the depleted air coming out of it is reheated in a second low-pressure oxidation/reduction system. The heat of the flue gas CO2/steam exiting out of the reduction reactor is used to preheat the fuel. The efficiency then climbs to 53 %. Another configuration is to operate the two reactors at atmospheric pressure and use the heat of the flue gas CO2/steam exiting out of the reduction reactor, on the one hand, and the heat of the depleted air from the oxidation reactor, on the other hand, to generate steam in a Rankine cycle, as in an oxyfuel boiler. The efficiency in this steam cycle configuration is about 43 % (Stroehle et al., 2009). © Woodhead Publishing Limited, 2010

304

MeO CO2 + H2O

R E D

Metal


loop Depleted air

O X

Cooler Condenser

Me Air

Turbine

CP

Depleted air

HRSG

Water Steam cycle

9.9 Layout of a chemical-looping combustion system.

CO2– steam turbine

H 2O

CO2 to storage


Fuel


305

9.3.6 Oxyfuel integrated gasification combined cycle (IGCC) In an IGCC, coal is converted into a CO + H2 mixture inside a gasification unit fed with pure oxygen and steam as the gasification agent. The so-called syngas CO + H2, is cooled, cleaned and shifted with steam into a CO2 + H2 mixture. CO2 and H2 are separated in a physical absorption process using solvents like Rectisol or Selexol or using a membrane (Jordal et al., 2004). The system may be modified in such way that the cleaned syngas, CO + H2 is used to feed any oxyfuel GT cycle considered in this chapter. The CO2 and H2O resulting from the combustion are separated in a condenser, without needing a solvent, as in the pre-combustion capture process. Performance calculations on an IGCC configuration have been carried out for the Matiant cycle (Mathieu and van Loo, 2005). Since the production of O2 is already required in an IGCC for the gasification, the coupling of a gasification unit delivering a syngas CO + H2, with an oxyfuel combustion cycle using that syngas as fuel is less penalising than the other capture options. The shift reactor and separation of CO2 from H2 are no longer needed. Here, the syngas is burnt in the oxyfuel combustion and the products, CO2 and water, are separated by cooling and condensation.

9.3.7 Oxyfuel solid oxide fuel cell (SOFC–CC) In this cycle, a SOFC, which is not yet commercially available, replaces the combustion chamber of a gas turbine (Dijkstra and Jansen, 2003). If the system is pressurised, the voltage and, consequently, the power output increases. Integrating the SOFC into a gas turbine is one possible way to pressurise the SOFC (see Fig. 9.10). The SOFC under pressure at high temperature, 850–1000 °C, generates a direct current from an electrochemical reaction of a gaseous desulphurised fuel at the anode and the oxygen of air at the cathode. In one specific fabrication, the cell has a tubular shape and is a hollow cylinder inside which air flows along the cathode whereas the fuel flows outside along the anode. The solid electrolyte is between the two electrodes. Oxygen is ionised, conducted through the electrolyte in contact with the fuel and oxidises it. If the fuel is H2, the product is steam. The efficiency of the SOFC at 1 bar is 45–50 %, whereas it goes up to 55–60 % at 3 bar. If the fuel is natural gas, it is reformed into CO and H2, and CO is shifted with steam into CO2 and H2 inside the cell. Typically, the fuel utilisation is 85 % while the oxygen utilisation is 25 %. The depleted air stream from the cathode is expanded in an air turbine, quite similar to the AZEP cycle. The flow from the anode consists of CO2, water and the non-converted fuel (CH4, CO, H2) which represents, typically, 15 % of the flow. This fuel is


306

Developments and innovation in CCS technology CO2 + H2O + unused fuel (+ Air1) SOFC generator

AC

Air 2 or O2

Fuel Air1 or O2

CO2 to storage

CO2 CP

CP

CO2 + H2O + air Cooler Condenser H 2O

Afterburning CO2 + H2O + air

DC

Turbine

HRSG

Only if fed with oxygen Steam cycle

9.10 Solid oxide fuel cell (SOFC) plays the role of the combustion chamber in a gas turbine cycle. The fuel is internally reformed in the cell at 850 °C/7 bar. The flue gas CO2 + steam + depleted air may give up its heat in a heat recovery boiler after the turbine exhaust, in a combined cycle configuration.

then burnt in an afterburner in air so that CO2 and water are now diluted in N2. Here the concentration of CO2 is much higher than in the flue gas of a gas turbine and theoretically it should be less penalising to capture it. The mixture at the cell temperature is then expanded in a turbine, (which has yet to be designed), just as in oxyfuel systems. The heat of the flue gases after expansion is given up either to the fuel and air streams (see Fig. 9.10) or to a steam cycle in a CC configuration. A hybrid system will provide efficiencies as high as 55 % for a power output of 200–400 kW and 60 % for 1 MW. At 2–3 MW, the efficiency goes up to some 70 % when using staged reheat GT cycles (without CO2 capture). This cuts the CO2 emission by about 50 % compared to the gas turbine alone. If a capture unit were installed in the flue gas of such a system, even with an efficiency penalty of 10 percentage points, the resulting efficiency would still be around 55–60 %, that of the best current CC. An oxyfuel system based on the use of a SOFC/GT would use O2 instead of air as the oxidiser in the afterburner (Demaret and Mathieu, 2001). Then the exhaust is a mixture of CO2 and water which, after giving up their heat in a recovery boiler, are cooled and the water removed by condensation.

9.3.8 Other oxyfuel combustion cycles To our knowledge, the first zero emission power cycle was proposed by D. Hochstein from Odessa Polytechnic Institute, who designed a cycle operating



307

on high-pressure CO2 as the working fluid in a Rankine cycle (Hochstein, 1940) at a time when the greenhouse effect was still unknown. (Yantovski, et al., 1992) presented a ZEPP with combustion of natural or coal-derived gas in an O2/steam mixture, triple turbine expansion, CO2 separation for sequestration and water recirculation. A calculation of the cycle efficiency using methane as the fuel, recycling of water and temperatures of 750 °C at the inlet of the three turbines provides a 37 % efficiency. This is an ancestor of the CES cycle. At the highest temperature of 1300 °C at the steam turbines inlet, which is unrealistic without cooling, the efficiency does not exceed 40 %, due the inability to recover the large enthalpy of steam condensation. Yantovski proposed the so-called COOPERATE cycle (CO2 prevented emission recuperative advanced turbine energy) which is a combination of a high-pressure Rankine cycle and a low-pressure Brayton cycle, both using CO2 as working fluid. The cycle uses a CO2 condenser to pump liquid CO2 at high pressure back to the combustor. A major problem was the release of non-condensable gases in the CO2 condenser, meaning that some CO2 escaped with inert gases. Suppressing the condenser and operating above the CO2 saturation line, the COOPERATE cycle became the Matiant cycle with a reheat (Mathieu, 1998). Ruether et al. (2000) described an integrated system with oxygen-blown dry coal entraining gasification and providing syngas fuel to the Matiant cycle. Oxygen for both the gasifier and the Matiant cycle is generated using an ion transport membrane (ITM) instead of a conventional cryogenic air separation unit. The thermal efficiency of the overall cycle is about 41 % LHV and 99.5 % of the CO2 produced is captured. Similar to the AZEP cycle, the Milano cycle (Romano et al., 2005) is another coal-fired ZEPP incorporating a membrane. As in the most recent versions of the AZEP cycle, there is no CO2 turbine but the efficiency is limited to about 42 % by the operating temperature of the membrane reactor. The bottoming cycle is a steam cycle, whereas the air turbine drives the compressor only. Here, again, the technical issue is the development of the membrane. The Oxycoal-AC cycle, developed at Aachen University in Germany (Renz et al., 2004), includes a high-temperature membrane unit, in which oxygen is mixed with CO2 and water vapour (Engels et al., 2009). Pulverised coal is burned in this mixture in a boiler to provide heat to a steam cycle. Both of these AZEP-like cycles show efficiencies using coal of around 41 %.

9.4

Advantages and limitations

9.4.1 Oxycoal boilers With regard to oxycoal boilers, the technology appears to be a competitive option for CO2 capture since none of the technical ‘challenges’ is prohibitive © Woodhead Publishing Limited, 2010

308


or shows any barriers. However, a better understanding of the combustion processes in a CO2/water atmosphere is still needed and is expected to arise from pilot installations. Once the models are validated, scaling up to large size boilers will be the next step on the road to commercial plants (Adams and Fry 2009; Erfurth et al., 2009; Toporov et al., 2009). The status of oxyfuel combustion may be summarised as follows: ∑ ∑ ∑ ∑

∑ ∑ ∑

Oxyfuel technology applied to supercritical coal-fired power plants uses a technology predominantly based on well-proven, commercially-available equipment which can be delivered at a low risk. The relative simplicity of the process is likely to have a lower impact on the availability of the plant compared to other options for CO2 capture. Oxyfuel technology in bituminous and lignite-fired plants to capture CO2 requires an additional capital investment estimated to be in the range 45–60 % (based on net power output). An efficiency penalty of around 8–9 percentage points (36.4 % for oxyfuel againt 45 % for air) is expected with the currently available technologies, that is about 20 % reduction of power output for a fixed fuel mass flow rate. Further improvements in oxygen production technologies are expected to lead to a major reduction in the efficiency penalty. Some 20 % improvement in specific energy consumption is in view when the design of the ASU is adapted to power generation and is fully integrated in the power system. The cost of electricity (44 7/MWh for oxyfuel against 31 for air) increases by some 45 % based on a 1.6 7/GJ bituminous coal price. The CO2 avoidance cost is around 20 7/t, which is the target of the European projects (CASTOR, ENCAP) in the Framework Programmes FP6 and FP7. CO2 emissions are reduced by 90 % by volume (95 against 774 g/ kWhe net) which is also the target of the above-mentioned European projects.

Vattenfall (Strömberg et al., 2009) has taken the initiative to construct a 30 MWth oxyfuel pilot plant at Schwarze Pumpe Power Station, Germany. It has been in operation since mid-2008 and is a step towards the construction of a 200 MWe demonstration power plant, at minimum risk, generating ‘CO2free’ electricity from 2015, under commercial conditions.

9.4.2 Oxyfuel combustion gas turbine (GT) cycles With regard to oxyfuel combustion GT cycles, there are several unsolved issues discussed in this section. The technology of gas turbines operating on CO2/H2O working fluids needed for the oxyfuel/Matiant CC, the water cycle



309

and the Graz cycle is not yet developed. This working fluid, differing from air especially in its molecular weight (heavier) and adiabatic exponent (Cp/ Cv) (less compressible), requires radically different design of compressors from those currently in use in 50 Hz power generation gas turbines. In existing gas turbines, the molecular weights of the gases in the compressor and turbine are close to that of air (28.8). In the case of oxyfuel combustion with CO2-recycle, the fluid at the compressor inlet has a molecular weight of about 43 whereas at the turbine inlet it is about 40. The change in working fluid from air to a CO2-rich gas (SCOC/Matiant and Graz cycles) results in a number of changes in properties that are important for the design of the compressor, the combustor and the hot gas path including the turbine (Kvamsdal et al., 2005): ∑ The speed of sound is 80 % of air. ∑ The gas density is 50 % higher than air. ∑ The adiabatic exponent, the specific heat ratio, is lower than that of air, resulting in a lower temperature change in adiabatic compression or expansion processes. The length and profile of the CO2/H2O compressor are different compared to air. More stages are needed and the height of the blades is lower at the exhaust. The longer rotor may result in rotor dynamics problems. Ways of reducing the overall length of the compressor should be investigated in future developments. The working fluid through the turbine imposes a new design of the expander too, i.e. existing power generation turbines would not be suitable. An oxyfuel gas turbine in a CC has a higher optimal pressure ratio, typically 30–35 compared to 15–18 used with air in the same cycle. Using TITs complying with materials resistance, this higher pressure ratio results in an exhaust gas temperature of about 600 °C, which perfectly matches the steam cycle. As for the cycles including an oxygen-transport membrane (AZEP and oxyfuel-fired lignite/coal cycles with integrated gas turbine and oxygen transport membrane (OTM) module), the technology is not yet developed to a level where it can be applied in power generation. All the concepts involving membranes (i.e., the AZEP and the SOFC/ GT) need further development. Due to their complexity, it is anticipated that several technological breakthroughs are needed to demonstrate their technical feasibility. For the AZEP cycles, an oxygen conducting membrane and a high-temperature heat exchanger have to be further developed. Even though a 220 kW SOFC + GT plant without CO2 capture has been demonstrated, the development of a novel afterburner technology is required to demonstrate the feasibility of the SOFC + GT system. Adaptation to the use of oxygen instead of air is still to come.


310


Considering the CLC cycles, most of the components are state-of-the-art and commercially available except for the reactor system. However, the use of solid fuels, or even of gasifier off-gases, is currently under development so that this option cannot really be used to decarbonise coal-fired power, the main CO2 emitter. The CLC cycles with the pressurised reactors (integrated in GT cycles) present a higher degree of novelty than other cycles in terms of the key components design as well as plant control. In contrast and as a term of comparison, post-combustion capture technology based on chemical absorption using amines may be considered as a mature technology even though it has still to be up-scaled to the level of several Megawatts power plants. This technology is, therefore, considered as a base case or a reference to evaluate the merits of oxyfuel systems (see Fig. 9.11). Comparison of the oxyfuel concepts All the three oxyfuel concepts–oxyfuel CC, water cycle, and Graz cycle–involve recirculation of exhaust gases. This gives a closed loop, i.e. a close integration of components, so that the system is vulnerable to startup, shutdown and load changes. Since the Graz cycle involves two recycle loops, it is anticipated that the operational challenges are higher than in case of the other two. Furthermore, all the concepts involving new and emerging technologies (AZEP, CLC) involve a high degree of subsystem integration, complex recycling of streams and/or a large amount of relative heat transfer. Oxyfuel CC, Graz, AZEP and CLC, can be considered as not mature with significant technology development challenges. However, the anticipated challenges relating to development of new gas turbines are less than those for large-scale membrane-based units as the large-scale turbine manufacturers have a lot of experience with similar development. Thus the oxyfuel or Matiant CC and the Graz concepts are less challenging than the others (AZEP, CLC). The operational challenges are considered as: ∑ low for the amine chemical absorption; ∑ medium for the oxyfuel/Matiant CC and the WC cycles; ∑ high for the AZEP and the CLC systems. It has been observed that the oxyfuel/Matiant CC and the WC cycles have a higher level of maturity and lower operational challenges than the AZEP and the CLC systems. By contrast, the latter have higher net efficiencies than amine absorption, WC, oxyfuel/Matiant CC and Graz cycles. To summarise, the concepts such as AZEP and CLC which show a higher net plant efficiency than WC, oxyfuel/Matiant CC, Graz and amine-based capture are those having a low level of technical maturity and a high degree of operational challenges. They will require technological breakthroughs if


56–59 % with O2 from ASU 70 66 % with air 65

Efficiency (%)

57 % 55 50

48 %

45

54 %

Oxygen produced externally

44 %

44 %

Water cycle

Matiant CC

51 %

51 %

43 %

40 35 30 CC (base case) Amine-based

Graz cycle

AZEP 100 %

AZEP 85 %

CLC

SOFC GT

311

9.11 Efficiency for post- and pre-combustion capture and oxyfuel GT cycles for coal and natural gas (WC = water or CES cycle; oxyfuel CC = Matiant cycle in a combined cycle configuration; Graz).



Oxygen produced internally

60

312


they are to be feasible. If the most promising concept is to be determined, it will be crucial to look more closely at the potential for plant efficiency improvement as well as other important factors, such as the risk of not succeeding in the technology development, the relative cost reduction potential and the timeframe until plant realisation. On the other hand, most of the cycles considered might be capable of integrating a high-temperature conductive membrane to generate oxygen (Anantharaman et al., 2009).

9.4.3 Capture readiness A power plant is called ‘capture ready’ when it can incorporate a CO2 capture system as soon as the regulations or the economic drivers intended to reduce carbon accumulation as well as ‘stranded assets’ are in place. The essential requirements are (Davison, 2009): 1. A study of various options for CO2 capture retrofit and for potential pre-investments. 2. The inclusion of sufficient space and of access for future additional facilities such as the ASU, the flue gas recycle (typically composed of 3 % O2, 2–17 % N2 depending on air in-leakage, 95 % CO2 by volume dry) and the CO2 purification and compression unit. Water cooling and water treatment systems have to be adaptable after retrofit. The de-NOx selective catalytic reactor is generally not needed anymore. The stack may also be suppressed and replaced by the vent of inert gases. 3. The identification of routes for CO2 storage, including transport facilities (such as pipelines). Pre-investments may be justified to reduce the downtime and the cost of capture retrofit and to optimise the plant performance after retrofit (Davison, 2009). In the case of oxyfuel combustion retrofit, these pre-investments have the objectives: 1. to minimise in-leakage of air into the boiler; 2. to design air ducts and fans to enable them to be re-used for flue gas recycle after retrofit; 3. to make FGD, if still needed, adaptable to new gas flows and compositions; 4. to modify the steam cycle to utilise low-grade heat to heat up boiler feedwater. 5. to add new capacity to maintain the site production. The risks of pre-investments are linked to uncertainties (such as the future value of CO2 emission permits, regulatory requirements, future availability of better capture technologies at the time of the retrofit) and to economic discounting, especially if the retrofit is not made soon after the construction.



9.5

313

Future trends

No technical barriers or show-stoppers have been identified for the implementation of oxyfuel combustion applied to furnaces, process heaters, boilers and power generation systems. Early use of this particular capture technology is likely to address applications involving boilers in power generation and process heating, since they require minimal modification of technologies and infrastructure that have already been developed for the combustion of hydrocarbon or carbonaceous fuels in air. However, several novel applications proposed for direct heating in steam turbine cycles or GT cycles for power generation still require the development of new components such as oxyfuel combustors, higher-temperature tolerant components such as CO2- and H2O-based turbines with blade cooling, CO2 compressors and high-temperature ion transport membranes for oxygen separation. As for CLC, it remains at an early stage of development. The potential for thermal efficiencies for oxyfuel cycles with CO2 capture, assuming the current state of development in power plant technology, may be summarised as shown on Fig. 9.11. The calculations show that the efficiency penalty for oxyfuel combustion boilers in power plants is around 9 percentage points compared to the same plant operating with air and using coal as fuel. This penalty is of the same order of magnitude as the penalty when using post-combustion capture. However, when oxy-combustion is applied in a natural gas-fired CC, the penalty is higher and, in addition, is much more sensitive to the fuel price. Oxyfuel combustion is consequently best suited to coal-fired plants (IPCC, 2005; Kluger et al., 2009; Ritter et al., 2009a,b). The oxyfuel GT cycles with external oxygen production (water cycle, Matiant or oxyfuel CC and Graz cycles) have similar efficiencies of around 44 %, while those with internal oxygen production using high-temperature membranes, such as AZEP (with 100 %) and CLC, have similar efficiencies at around 51 %. AZEP with 85 % CO2 retention rate when afterburning is used and the hybrid SOFC/GT, both without CO2 capture show the highest efficiencies. Pulverised coal-fired power generation systems, using supercritical steam conditions presently operate at efficiencies around 45 % (LHV), while predicted efficiencies for 2010–2020 are above 50 % (IEA, 2004) for plants using ultra-supercritical steam conditions. An increase in efficiency of more than 5 percentage points can therefore be expected for future oxyfuel capture systems based on coal firing; this could potentially match the best efficiencies realisable today for pulverised coal-fired plants without CO2 capture. Similarly, natural gas-fired CCs will have efficiencies of 65 % in 2020, 7–10 percentage points above current efficiencies (between 55 and 58 %), which will enable plant efficiencies for natural gas-fired oxyfuel cycles with CO2 capture to be above 50 %.


314


Current technology development envisages very high-efficiency separation of NOx, SOx and Hg, as part of the CO2 compression and purification system (Ritter et al., 2009a,b; Santos, 2009; White and Fogash, 2009; Yan et al., 2009). Improved separation efficiencies of these contaminants are possible based on further process and heat integration in the power cycle. The energy penalty for producing oxygen is by far the most important cause of reduced efficiency in an oxyfuel cycle compared to a conventional power plant. Current cryogenic oxygen technology shows continuing cost reduction based on improved compressor efficiencies, more efficient process equipment and larger scale plants (Beysel, 2009). The new high-temperature oxygen membrane could significantly improve power generation efficiency and reduce capital cost. Future oxyfuel demonstration plants could be based on retrofits to existing equipment such as process heaters and boilers, in order to minimise development costs and achieve early market entry. In this respect, power systems of relevance for oxyfuel combustion capture are mainly the steambased pulverised coal and natural gas-fired plants that currently represent up to 1468 GWe or 40 % (IEA, 2004) of the existing global infrastructure. Several demonstration units may be expected within the next few years, particularly in Europe, the USA, Canada and Australia, where active research initiatives are currently underway. As these developments proceed and the technologies achieve market penetration, they may become competitive compared to alternate options based on pre- and post-combustion CO2 capture. In 2008, a major milestone in the development of oxyfuel combustion technologies was achieved when Vattenfall and Alstom successfully completed the commissioning of the world’s first full chain oxyfuel pilot plant facility with CO2 capture. In Europe, pilot/demonstration power plants using oxyfuel boilers are already in operation or will be in operation within the next few years. Eight projects are listed below: 1. the lignite-fired plant 30 MWth at Schwarze Pumpe (Vattenfall, Berlin, Germany), startup in 2008 (Strömberg et al., 2009; Hultqvist et al., 2009); 2. a 30 MWth natural gas-fired industrial boiler in Lacq (Total-Fina-Elf, Pau, France), startup in July 2009 (Perrin and Cauley, 2009); 3. Callide (CS Energy, Australia), 90 MWth (30 MWe) coal-fired boiler, startup expected in 2010; 4. CIUDEN CCS (Spain), 20 MWth anthracite-fired boiler and 30 MWth circulating fluidised bed, startup expected in 2010; 5. ENEL (Brindisi, Italy) 50 MWth coal-fired boiler, startup to be decided; 6. Jamestown oxycoal technology (USA), 150 MWth (50 MWe) coal CFB, startup expected in 2013; 7. Vattenfall (Janschwalde, Germany) 1000 MWth (250 MWe) lignite fired boiler, startup expected in 2015; © Woodhead Publishing Limited, 2010


315

8. Korean Electric Power Research Institute (Youngdong, Korea), 400 MWth (100 MWe) coal-fired boiler, startup expected in 2016. Work on a fluidised bed using oxyfuel combustion at lab scale is ongoing in the framework of the European ENCAP project (Alstom), at Chalmers, Cambridge, Imperial College (part of the EU FLEXGAS project) and Vienna Universities and a number of other locations in the world. Regarding the oxyfuel combustion gas and steam turbines cycles, a pilot of 5 MWe is operating in California based on rocket propulsion technology from NASA (Clean Energy Systems Ltd, USA). The other cycles (Matiant, Graz, AZEP, and others) are still at the conceptual level. Experiments in pilot/demo plants should enable technical issues to be overcome and provide a better knowledge aimed at validating the models. These technical issues in oxyfuel combustion are linked to: 1. combustion: flame stability (Toporov et al., 2009), ingnition, char combustion rate, devolatilisation kinetics, reactivity of coal particles, formation of sulphur oxides and sulphur in ash, air ingress in boilers; 2. radiative and convective heat transfers either with or without water (O’nions et al., 2009); 3. flue gas cleaning processes such as the removal of water, halides, SOx and NOx particles and Hg (Thébault et al., 2009); 4. removal of impurities in the compression/purification of CO2 product (White and Fogash, 2009); 5. corrosion of materials and steels (Hünert et al., 2009); 6. ash slagging, fouling and deposition and the influence of flue gas recycling (Adams and Fry, 2009; Bradley and Fry, 2009). In the IEA roadmap (IEA, 2004), post-combustion capture systems using solvents such as amines or ammonia should be the first to appear on the market, since they are able to achieve the required emission reduction targets as defined in the IPCC assessment report, (IPCC, 2005). Oxyfuel combustion will follow shortly after while other technologies will come later, so creating a portfolio of capture technologies able to reduce CO2 emissions by 90 % or more. A significant incentive to the development of oxyfuel combustion technology, as well as for pre- and post-combustion capture technologies, is the introduction of environmental requirements and/or fiscal incentives to promote CO2 capture and storage.

9.6

References

Adams B and Fry A (2009), Impacts of oxy-combustion flue gas recycle on coal ash slagging and fouling, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany.


316


Anantharaman R, Bolland O and Åsen K I (2009), Novel cycles for power generation with CO2 capture using OMCM technology, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 335–342. Anderson R, Brandt H, Doyle S, Pronske K and Viteri F (2003), Power generation with 100 % carbon capture and sequestration, Second Annual Conference on Carbon Sequestration, 5–8 May, Alexandria, VA. Anderson R E, Doyle S E and Pronske K L (2004), Demonstration and commercialization of zero-emission power plants, 29th International Technical Conference on Coal Utilization & Fuel Systems, 18–22 April, Clearwater, FL. Beysel G (2009), The proven cryogenic air separation process adapted to the needs of oxyfuel combustion and first results from Schwarze Pumpe Pilot Plant, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Betz M, Baumann S and Meulenberg W A (2009), Ceramic membranes for oxyfuel power plants, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Bouwmeester H J M and Van Der Haar L M (2002), Oxygen permeation through mixedconducting perovskite oxide membranes, Ceramic Transactions, 127, Materials for Electrochemical Energy Conversion and Storage, 49–57. Bradley A and Fry A (2009), Impacts of oxy-combustion flue gas recycle on coal ash slagging and fouling, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Brandvoll Ø and Bolland O (2004), Inherent CO2 capture using chemical looping combustion in a natural gas fired power cycle, ASME Paper No. GT-2002-30129, ASME Journal of Engineering for Gas Turbines and Power, 126, 316–321. Bredesen R, Jordal K and Bolland O (2004), High-temperature membranes in power generation with CO2 capture, Journal of Chemical Engineering and Processing, 43, 1129–1158. Cauley K J, Alexander K C, McDonald D K, Perrin N and Tranier J (2009), Commercial demonstration of oxy-coal combustion clean power technology, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Chatel-Pelage F, Ovidiu M, Carty R, Philo G, Farzan H and Vecci S (2003), A pilot scale demonstration of oxy-fuel combustion with flue gas recirculation in a pulverised coal-fired boiler, 28th International Technical Conference on Coal Utilization & Fuel Systems, 10–13 March, Clearwater, FL. Cho P, Mattisson T and Lyngfelt A (2002), Reactivity of iron oxide with methane in a laboratory fluidised bed – application of chemical-looping combustion, 7th International Conference on Circulating Fluidised Beds, Niagara Falls, Canada, 5–7 May, 599–606. Croiset E and Thambimuthu K V (2000), Coal combustion in O2/CO2 mixtures compared to air, Canadian Journal of Chemical Engineering, 78, 402–407. Davison J (2009), CO2 Capture ready plants, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Dechamps P, Distelmans M, Mathieu P and Pirard N (1994), Performances of combined cycle power plants using CO2 gas turbines, Flowers’94 Conference, Florence, Italy, 671–682. Demaret M and Mathieu P (2001), Integration of a high temperature fuel cell (SOFC) in a near zero CO2 emission power cycle, ECOS 2001, 4–6 July, Istanbul, Turkey. © Woodhead Publishing Limited, 2010


317

Dijkstra J W and Jansen D (2003), Novel concepts for CO2 capture with SOFC, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Elsevier (Pergamon), Oxford, UK, Vol. 1, 161–166. Dillon D J, Panesar R S, Wall R A, Allam R J, White V, Gibbins J and Haines M R (2005), Oxy-combustion processes for CO2 capture from advanced supercritical PF and NGCC power plant, in Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 1, 211–220. Dubuisson R, Houyou S, Mathieu P and Nihart R (2000), A near zero emission O2/CO2 combined cycle with CO2 reuse and sequestration: efficiency, costs, optimisation, simulation and environmental aspects of energy systems, ECOS 2000, 5–7 July, Twente, the Netherlands. Engels S, Modigell M and Beggel F (2009), OXYCOAL-AC: integration of high temperature membranes for air separation in oxyfuel power plants, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Erfurth J, Toporov D, Förster M and Kneer R (2009), Numerical simulation of a 1200 MWth pulverised fuel oxy-firing furnace, Fourth International Conference on Clean Coal Technologies; CCT2009, 18–21 May, Dresden, Germany. Fiaschi D, Manfrida G, Tempesti D and Mathieu P (2008), Predicting the performance of a supercritical CO2 power cycle including sequestration, Energy–The international Journal, 34(12), 2240–2247. Foy K and Yantovski E (2006), History and state-of-the-art of fuel fired zero emission power cycles, International Journal of Thermodynamics, 9(2), 37–63. Grahl S, Hiller A, Löser J, Weigl S, Wilhelm R and Beckmann M (2009), ADECOS II advanced development of the coal-fired oxyfuel process with CO2 separation. A Research Project of the COORETEC Programme, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Griffin, T, Sundkvist S G, Aasen K and Bruun T (2003), Advanced zero emissions gas turbine power plant, ASME Turbo Expo Conference, 16–19 June, Atlanta, GA, paper GT-2003-38120. Hochstein D P (1940), Carbon dioxide power cycle, Soviet Boiler and Turbine Construction, 10, 420–423. Hultqvist D, Glausch M, Meyer H, Radunsky D and Witter T (2009), Vattenfall oxyfuel power plant development – engineering of a coal fired power plant with oxyfuel technology, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Hünert D, Schultz W, Österle W, Saliwan-Neumann R, Oder G, Urban I and Kranzmann A (2009), Ageing and corrosion of steels in CO2 rich flue gases, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. IEA (2004), World Energy Outlook, International Energy Agency, OECD/IEA, Paris, France. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, de Coninck H C, Loos M and Meyer, L A (eds), Cambridge University Press. Cambridge, UK. Jericha H, Göttlich E, Sanz W and Heitmeir F (2003), Design optimisation of the Graz cycle power plant, ASME Turbo Expo Conference, 16–19 June, Atlanta, GA, paper GT-2003-38120.


318


Jordal K, Bolland O and Klang A (2004), Aspects of cooled gas turbine modeling for the semi-closed O2/CO2 cycle with CO2 capture, Journal of Engineering for Gas Turbines and Power, 126, 507–515. Kluger F, Monckert P, Krohmer B, Stamatelopoulos G-N, Jacoby J and Burchardt U (2009), Oxyfuel pulverized coal steam generator development 30 MWth pilot steam generator commissioning and testing, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Kvamsdal H, Maurstad O, Jordal K and Bolland O (2005), Benchmarking of gas-turbine cycles with CO2 capture, in Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 1, 233–242. Marin O, Bourhis Y, Perrin N, DiZanno P, Viteri F and Anderson R (2003), High efficiency zero emission power generation based on a high temperature steam cycle, 28th International Technical Conference on Coal Utilization and Fuel Systems, 10–13 March, Clearwater, FL. Mathieu P (1998), Presentation of an innovative zero-emission cycle for mitigation the global climate change, International Journal of Applied Thermodynamics, 1, (1–4) 21–30. Mathieu P (2003), Mitigation of CO2 emissions using low and near zero CO2 emission power plants, Clean Air, International Journal on Energy for a Clean Environment, 4, 1–16. Mathieu P and De Ruyck J (1993), The CO2 gas turbine option for recovery of CO2 in CC and IGCC plants, IGTI, 8, 77–83. Mathieu P and van Loo F (2005), Modeling of an IGCC plant based on an oxy-fuel combustion combined cycle, Clean Coal Technologies 2005, 10–12 May, Cagliari, Sardinia, Italy. Mathieu P and Yantovski E (1997), Highly efficient zero emission CO2 based power plants, Energy Conversion and Management, 38, (suppl 1), S141–146. Nsakala N, Liljedahl G, Marion J, Bozzuto C, Andrus H and Chamberland R (2003), Greenhouse gas emissions control by oxygen firing in circulating fluidised bed boilers, Second Annual National Conference on Carbon Sequestration, 5–8 May, Alexandria, VA. O’nions P, Riley G, Jamieson E and Smart J (2009), Radiation and convection heat transfer in oxyfuel combustion, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Perrin N and Cauley K (2009), Commercial demonstration of oxy-coal combustion clean power technology, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Renz U (2004), Entwicklung eines CO2-emissionsfreien Kohleverbrennungs-processes zur Stromerzeugung in einem Verbundvorhaben der RWTH Aachen, XXXVI. Kraftwerkstechnisches Kolloquium: Entwicklungpotentiale fuer Kraftwerke mit fossile Brennstoffen, 19–20 October, Dresden, Germany. Ritter R, Schödel N, Ponceau M, Winkler F (2009a), Advancement of the CO2 compression and purification plant integrated in the oxyfuel technology, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany Ritter R, Holling B, Porsche P and Biele M (2009b), First experience in the commissioning of the CO2-plant in Schwarze Pumpe: generation, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany.



319

Romano M, Napoletano S, Chiesa P and Consonni S (2005), Decarbonized electricity production from coal by means of oxygen transport membranes, Fourth Annual Conference on Carbon Sequestration, 2–5 May, Alexandria VA, 1402. Ruether J, Le P and White Ch (2000), A zero-CO2 emission power cycle using coal, Technology, 7S, 95–101. Santos S (2009), Fate of sulphur in coal during oxyfuel combustion with recycled flue gas (review of the current state of understanding), Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Simonsson N, Eriksson T and Shah M (2009), Circulating fluidized bed boiler technology – a competitive option for CO2 capture through oxyfuel combustion?, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Stadler H, Beggel F, Persigehl B, Kneera R, Modigellb R and Jeschke P (2009), The OXYCOAL-AC process: component behaviour and thermodynamic efficiency, Fourth International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Stroehle J, Lombarte A, Orth M and Epple B (2009), Simulation of a chemical looping combustion process for coal, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Strömberg L, Lindgren G, Jacoby J, Giering R, Anheden M, Burchhardt U, Altmann H, Kluger F, Stamatelopoulos G-N (2009), Update on Vattenfall’s 30 MWth oxyfuel pilot plant in Schwarze Pumpe, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 581–589. Tan, Y, Douglas M A, Croiset E and Thambimuthu K V (2002), CO2 capture using oxygen enhanced combustion strategies for natural gas power plants, Fuel, 81, 1007–1016. Thébault C, Yan J, Biele M, Ritter R, Jacoby J, Anheden M and Kuehnemuth D (2009), Behaviors of NOx and SOx in CO2 compression/purification processes – experience at 30 MWth oxy-coal combustion CO2 capture pilot plant, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Toporov D, Förster M and Kneer R (2009), Oxycoal swirl flame stability as a function of flue gas recycling ratio, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. White V and Fogash K (2009), Purification of oxyfuel-derived CO2: current developments and future plans, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Wilkinson M B, Simmonds M, Allam R J and White V (2003), Oxy-fuel conversion of heaters and boilers for CO2 capture, Second Annual Conference on Carbon Sequestration, 5–8 May, Alexandnia, VA. Yan J, Faber R, Schmidt T, Ross G, Jacoby J, Anheden M, Geiting R and Kosel D (2009), Flue-gas cleaning processes for CO2 capture from oxyfuel combustion – experience of FGD and FGC at 30 MWth oxyfuel combustion pilot plant, First Oxyfuel Combustion Conference, 8–11 September, Cottbus, Germany. Yantovskii E I, Zvagolsky K N and Gavrilenko V A (1992), Computer exergonomics of power plants without exhaust gases, Energy Conversion and Management, 33(5–8), 405–412.


10

Advanced oxygen production systems for power plants with integrated carbon dioxide (CO2) capture

S. C. K l u i t e r s, R. W. v a n d e n B r i n k and W. G. H a i j e, Energy research Centre of the Netherlands, the Netherlands Abstract: This chapter deals with advanced oxygen generation technologies and oxyfuel power plant systems with integrated oxygen selective membranes. Section 10.2 shows the main air separation technologies, existing and under development: cryogenic air separation units, pressure and vacuum swing adsorption, chemical looping, the ceramic autothermal recovery (CAR) process and oxygen selective membranes. Section 10.3 presents materials for oxygen selective membranes in oxyfuel applications. Section 10.4 shows oxyfuel power plant schemes with integrated oxygen selective membranes and discusses their technical performance. The following sections show advantages and limitations; future trends; and sources of further information about advanced oxygen generation technologies before concluding. Key words: oxygen selective membranes, chemical-looping combustion, cryogenic air separation unit (ASU), oxyfuel power plant with integrated oxygen selective membrane, pressure/vacuum swing adsorption (PSA/VSA).

10.1

Introduction

At present, both hydrogen and power are mainly produced from fossil fuels. It has now generally been accepted that the anthropogenic CO2 emissions coming with power generation are the primary cause of the global warming through the greenhouse gas effect (IPCC, 2007). Add to this the speed with which developing countries are closing the gap in per capita energy consumption with developed countries (in China alone every week a 500 MW coal-fired plant is being commissioned to meet the increasing energy demand). It is clear that there is an enormous requirement for technology development for clean power and hydrogen production from fossil fuels including carbon dioxide (CO2) capture and subsequent storage in empty gas or oil fields and saline aquifers. Generally speaking, there are three options for CO2 capture. The first is oxyfuel combustion, where fuels are combusted with pure oxygen producing only water and CO2. Water can easily be condensed and pure CO2 is available for storage or sequestration. The second option is post-combustion capture. In this case the CO2 is captured from the flue 320 © Woodhead Publishing Limited, 2010

Advanced oxygen production systems for power plants

321

gasses (current technology would use liquid chemical sorbents). This is a huge task in view of all the nitrogen present in this stream. The third option is pre-combustion capture. In this case the hydrogen is separated from the fossil fuel before combustion. This can, at present, be done with physical liquid sorbents at low temperatures. All three options have in common that a great deal of energy is needed to run them. In this way, the power production efficiency is severely reduced, by about 10 % points. To give an indication, a new coal-based power plant will suffer an efficiency penalty of about 10–14 percentage points, i.e. from about 46 % to about 36 % efficiency (All efficiency numbers mentioned in this chapter are LHV based.) (Jansen et al., 2008). This is the origin of the statement that for every four power plants built with CO2 capture, a fifth is needed to compensate for the energy losses. Natural gas-based power plants enjoy higher efficiencies, but also fall foul of larger penalties for CO2 capture on a per tonne CO2 captured base. In simple terms, the less carbon-rich fuel produces a lower concentration level of CO2, which is thus more difficult to remove. In this field, Energy research Centre of the Netherlands (ECN) is focusing on oxyfuel and pre-combustion capture technologies for natural gas- and coalbased power plants, thereby aiming at lower efficiency penalty and cost as compared to conventional capture technologies. The use of high temperature solid CO2 sorbents or CO2 or hydrogen selective membranes instead of lowtemperature physical liquid sorbents for pre-combustion capture and the use of advanced membrane-based oxyfuel processes hold the promise to reduce both the efficiency penalty and the cost. This chapter focuses on the production of oxygen for use in power production with integrated CO2 capture. Obviously, production of oxygen is needed for the oxyfuel route, where it is used for combustion of coal or natural gas. Oxygen production is also important in the pre-combustion CO2 capture route. Coal gasification to produce syngas is usually carried out using pure oxygen. Moveover, oxygen-blown autothermal reforming (ATR) of natural gas is one of the preferred ways of converting natural gas into syngas. In this chapter the demands on the oxygen production in terms of volume, purity and pressure are discussed. Currently, for those applications cryogenic air separation units (ASU) are used. This chapter describes cryogenic ASU and several other oxygen production technologies and discusses whether these technologies meet the demands set. Subsequently, the chapter goes into more detail on the most promising oxygen production technology under development, oxygen selective membranes (OSM), also referred to as oxygen transport membranes (OTM), ion transport membranes (ITM), mixed conducting membranes (MCM) or mixed ionic electronic conducting (MIEC) membranes. Recent advances in membrane materials and membrane modules are discussed, before the


322


focus shifts to the integration of OSM in power production. Several possible applications are analysed for their power production efficiency.

10.2

Technologies for air separation

In industry, oxygen has been produced by a variety of technologies for decades. Cryogenic air separation, pressure and vacuum swing adsorption (PSA and VSA) and polymeric membranes all serve different segments of the oxygen market. The interest in oxygen combustion as an enabling technology of CO2-lean power production from fossil fuels has spurred the development of advanced air separation technologies, such as chemical-looping combustion (CLC), ceramic autothermal recovery (CAR) and OSMs. These technologies will be briefly described below, and their suitability for use in an oxygenfired fossil fuel fed power plant with CO2 capture will be discussed.

10.2.1 Demands for oxygen production technologies for zero-emission power plants Oxygen can be used in several ways in a power plant equipped with CO2 capture. Oxyfuel power plants use oxygen to combust all fuel. In these power plants, air separation can be integrated into the power plant to varying degrees. A pulverized coal supercritical boiler and a natural gas combined cycle using recycled CO2 as working medium are examples of oxyfuel power plants with non-integrated air separation. The demand for oxygen is 19.5 t/day/MWe for the coal-fired boiler and 15.5 t/day/MWe for the natural gas combined cycle (IEA, 2007b). This translates to 19 500 t day for a 1000 MWe coal-fired power plant. To limit the dilution of the CO2 product after combustion, the purity of the oxygen should be at least 95–97 % (IEA, 2007b). Chemical-looping combustion (as shown in Chapter 11) and the OSM systems shown in the remainder of this chapter are examples of oxyfuel systems with integrated ASUs supplying oxygen but not in the form of a high-purity oxygen flow. In principle, nitrogen dilution is very low in these systems. The oxygen demand is equivalent to systems with non-integrated air separation. For a 1000 MWe natural gas-fired power plant, the amount of air separated is equivalent to approximately. 15 000 t/day of pure oxygen production. A special oxyfuel case is the use of oxygen to combust left over hydrogen and carbon monoxide in solid oxide fuel cell (SOFC) off-gas. SOFCs are not expected to be used in the hundreds of MWe range. For a 1 MWe SOFC, the amount of oxygen needed is less than 1 t per day. Pre-combustion systems with CO2 capture can use oxygen for partial oxidation of the fuel. Examples are production of syngas by coal gasification or natural gas ATR. As this is only a partial oxidation, the oxygen demand



323

is lower than for oxyfuel combustion: 5000–8500 t/day for a 1000 MWe power plant. Also oxygen purity specifications are somewhat lower: 90–97 % (IEA, 2007b). In pre-combustion CO2 capture systems with gas turbines, the hydrogen needs to be diluted before combustion in the gas turbine and, in many cases, nitrogen from the ASU will be used for that purpose. This means that the oxygen content of the nitrogen should be below the lower flammability limit for hydrogen: i.e., below 1 % (IEA, 2007b). Summarising, the oxygen demand for CO2 capture power plants is characterised by high volumes of oxygen and relatively high purity, in the range 90–97 %, especially when oxygen is used for oxyfuel systems.

10.2.2 Cryogenic air separation Cryogenic ASU installations have already been used for oxygen production for over a century in various industrial processes such as steel production and refineries. The largest ASU plants in use produce 4000 t/day of oxygen and are used in the production of synfuels from natural gas. Also, the currently existing oxygen-blown coal gasifiers and the 30 MWth pilot plant for oxyfiring of coal by Vattenfall in Schwarze Pumpe, Germany, use cryogenic air separation systems (Strömberg et al., 2009). In a cryogenic air separation plant, the more volatile components of the air (nitrogen and argon) leave the distillation column at the top. The less volatile oxygen is collected at the bottom, either as a gas or a liquid (IEA, 2007b). The purity of the oxygen is determined by the column design and the number of stages. Purities of higher than 99 % are achievable. For use in power production, oxygen purities can be lower, which reduces the energy use: a specific design for an ASU for oxycoal combustion by Air Liquide showed a 20 % lower energy use than a standard ASU (Darde et al., 2009). Still, even for state-of-the-art cryogenic ASUs, the power consumption for oxygen production is relatively high. A power use of 87 MW for oxygen production has been reported for a 600 MW oxy-fired pulverised coal plant (Dillon, cited in IEA, 2007b). Main power consumption is in the air compressor and the low-pressure column. Moreover, capital costs are relatively high: between 20 and 30 % of total equipment costs for natural gas oxyfuel cycles depending on the type of cycle (Rezvani et al., 2009). When using a cryogenic ASU in an integrated gasification combined cycle system (IGCC), air bleed can be used from the gas turbine, which reduces the duty of the air compressor of the ASU. Nitrogen is needed to dilute the hydrogen before the gas turbine in an IGCC equipped with CO2 capture. This implies that the oxygen content in the nitrogen should be low (less than 1 %) and that the nitrogen needs to be compressed up to the inlet pressure of the gas turbine (approximately 15–25 bar depending on the gas turbine). Summarising, cryogenic ASUs are available at or near the size needed for


324


an oxyfuel power plant, and can deliver oxygen at the right purity. However, the costs are high (approximately 20–30 % of the capital costs for natural gas oxyfuel cycles (Rezvani et al., 2009)) and the high power consumption of approximately 14 % of the power produced (IEA, 2007b) leads to an importantly lower efficiency of an oxyfuel power plant.

10.2.3 Pressure and vacuum swing adsorption Both VSA and PSA for air separation use a number of columns filled with sorbent to separate the air at ambient temperature. The columns are filled with at least two different sorbents, a pre-treatment sorbent to remove water and CO2 from the air and a main sorbent that preferentially adsorbs nitrogen over oxygen. So, in the feed step air is fed and oxygen is produced and in the regeneration step both sorbents are regenerated and water, CO 2 and nitrogen are released. In VSA of air, the feed step takes place at atmospheric pressure, while the sorbent is regenerated at pressure well below ambient (200 to below 20 mbar pressure depending on the desired oxygen purity) (Kumar, 1996). In PSA, the feed pressure is well above ambient, while regeneration takes place at atmospheric pressure. In PSA oxygen recovery is lower, because the oxygen co-adsorption increases with pressure. It also requires the total feed air to be compressed, compared to only the nitrogen for VSA. This means that PSA is not suited for plants larger than 15 t/day oxygen (Kumar, 1996). VSA systems are commercially sold for up to 120 t/day, like the PRISM® technology sold by Air Products and Chemicals (Air Products and Chemicals 2002). Purity is 90–93 % oxygen. For continuous and efficient oxygen production by PSA and VSA, several vessels are needed, the number depending on the cycle design. The typical size of VSA systems is two orders of magnitude smaller than needed for a large-scale oxyfuel power plant.

10.2.4 Chemical-looping Chemical-looping technologies (see also Chapter 11) do not produce ‘free’ oxygen, but make use of metallic oxygen carriers to transfer the oxidizing agent to a fuel for full or partial combustion. The process is usually carried out in a connected fluidized bed system, with an air reactor, in which a reduced metal is being oxidized. The product of the air reactor is nitrogen with several percent of oxygen. The oxidized metal is transferred to the fuel reactor, where a fuel is combusted and the metal oxide is reduced. The product of the fuel reactor is a mixture of steam and CO2 (Lyngfelt et al., 2001). Chemical-looping can also be used for hydrogen production (Ryden and Lyngfelt, 2006) and pre-combustion CO2 capture. Systems for CLC of coal have been described in the patent literature (Lyon, 1996).



325

The chemical-looping technology is still in the research phase, and the largest CLC plant in existence today is a natural gas fed 120 kWth pilot plant at Vienna University of Technology in Austria. This size is equivalent to less than 1 t of oxygen per day. Large-scale plants have been described in literature (Naqvi and Bolland, 2007). They performed system calculations and came up with a power plant efficiency of about 53 %. As already mentioned, chemical-looping is not an oxygen production technology per se, but requires a completely integrated process.

10.2.5 The ceramic autothermal recovery (CAR) process The CAR process for oxygen production was developed by the BOC group (currently part of Linde) in the 1990s. The CAR process uses the oxygen storage capacity of perovskite materials in a fixed bed reactor. The use of at least two reactors makes the process cyclic and ensures a continuous production of oxygen (Acharya et al., 2006). The CAR process can also be described as a PSA process carried out at high temperature (600–800 °C). The feed gas is air and the purge gas could be low-pressure steam or recycled flue gas. For a large-scale plant (15 000 t per day of oxygen) a ten-bed cycle was proposed, each reactor having an internal diameter of 2.5 m and a length of 25.6 m (Acharya et al., 2006). Technical and economic analysis of the CAR process in an oxyfuel 1000 MW natural gas combined cycle plant has been performed. Both power consumption and capital costs for air separation by CAR are reported to be approximately half of the values for a cryogenic ASU (Acharya et al., 2006). CAR produces near 100 % pure oxygen and, in the study mentioned above, is compared to an ASU producing 99 % oxygen. It must be noted that 95 % oxygen purity may be enough for oxy-firing, and that ASU costs and power consumption increase strongly when purities higher than 95 % are required (Darde et al., 2009). CAR is still in the research phase. The largest scale reported in literature is an approximately 1 t/day process development unit (PDU) that is currently operated by the Western Research Institute in Wyoming, USA (WRI, 2007).

10.2.6 Oxygen selective membranes Membranes are another technology in development for large-scale oxygen production. Several types of membranes for oxygen production exist, but in this chapter only dense membranes are considered. Polymer membranes are used for air enrichment for medical purposes on a small scale, so they will not be discussed further. The working principle of dense membranes is discussed in Section 10.3.


326


Here, only the use of oxygen selective membranes for CO2 capture is discussed. The high operating pressure and temperature make OSMs a good match with an IGCC power plant. However, several advances in gas turbine technology are necessary for full integration of these membranes and IGCC (IEA, 2008). Compared to cryogenically produced oxygen, IGCC power plant efficiency is 1.2 % higher when OSMs are used for oxygen production, and the total power production costs are 6–8 % lower (IEA, 2007b). In principle, OSMs are 100 % selective and therefore deliver high oxygen purities in practice. For IGCC, the oxygen content of the nitrogen product is generally too high to use the nitrogen as a diluent of the gas turbine fuel. The oxygen can be removed from the nitrogen by reacting it with a fuel (IEA, 2007b). For integration of OSMs in oxyfuel power production, the cost per tonne CO2 avoided is approximately 25 % lower when OSMs are used as compared to cryogenic ASU (IEA, 2007b). The largest scale at which OSMs have been demonstrated is 5 t/day, and a design has been presented for a 2000 t/day plant by Air Products (IEA, 2007b).

10.3

Oxygen selective membrane technology for oxyfuel power plants

In the previous sections we have seen that advanced oxygen generation can be applied in both pre-combustion and oxyfuel systems with CO2 capture. As pre-combustion processes have been discussed in Chapters 5–8 of this book, we will focus on oxyfuel power plants. Also, previous sections showed that only cryogenic air separation is currently available to supply the required oxygen quantities and purities for large-scale power production, albeit at a substantial cost in terms of energy and system economics. Of the advanced air separation technologies, chemical-looping is discussed in Chapter 11. CAR is a promising technology, but has only been demonstrated on a small scale. PSA and VSA technologies are viable only in applications much smaller than hundreds of megawatts. Consequently, the remainder of this chapter focuses on oxyfuel power plants with CO2 capture using OSMs. Gas separation membranes can be used to separate oxygen from air by creating an oxygen partial pressure difference across the membrane. The best known class of materials that can transport ionic oxygen through the bulk is that of the perovskites, with the general formula ABO3 – d (Fig. 10.1), the A cations being Ba, La, Sr, etc., and the B cations Co, Fe, Cr, etc. It exhibits the largest oxygen fluxes, while the constituents necessary for the production of perovskites are abundantly present in the earth’s crust. Since perovskites are fully oxidic materials, it is expected that these materials are stable in an oxidizing environment at elevated temperatures (700–900 °C). High oxygen transport rates have been achieved, which relates to the high



ABO3–d

327

A2BO4 + d

10.1 Structure of the compounds ABO3–d and A2BO4+d. The octahedra contain the B cation, the A cations are represented by the large spheres (Courtesy of ECN).

concentrations of mobile oxygen vacancies in the perovskite oxides at elevated temperatures. The presence of multivalent cations ensures a high, often predominating, electronic conductivity. Crucial to the performance as OSM is that the materials are capable of rapid oxygen exchange and bulk diffusion of oxygen, and maintain structural and chemical integrity under the conditions of application. As stated before a 1000 MWe natural gas-fired power plant, requires 15 000 t/day of pure oxygen production. This means that with a target flux for economic viability of 10 ml/cm2·min (Bredesen et al., 2004), a surface area of about 80 000 m2 is needed, stable in all aspects, with no pinholes, cracks and such like. The mechanism of oxygen conductance through perovskites is shown in Fig. 10.2. Here oxygen is dissociated into two atoms that travel through the perovskite as oxide ions along oxygen vacancies from the feed (left) side to the permeate (right) side where they recombine to oxygen molecules. Transport occurs provided that the pressure (more correctly chemical potential) at the feed side pf, is larger than that at the permeate side, pp. At the same time, electrons are transported from right to left. The materials thus have to be conductive in both oxide anions and electrons and are therefore called MIEC (mixed ionic electronic conductive) materials. SrCo0.8Fe0.2O3–d (SCF) is reported (Vente et al., 2006a) to provide one of the largest membrane oxygen fluxes in the series La1–xSrxCo1–yFeyO3–d (0 < x, y < 1). However, undesirable ordering of SCF into a brownmillerite type structure, Sr2Co1.6Fe0.4O5, has been reported (McIntosh et al., 2006) to occur below 800 °C at very low oxygen partial pressures. This ordered state reduces the oxygen flux, whereas the associated lattice expansion leads to large mechanical stresses across the membranes. Literature reports have demonstrated an increase in oxygen flux and an apparent increase in the stability of the cubic perovskite phase upon 50 % substitution of Ba for Sr, Ba0.5Sr0.5Co0.8Fe0.2O3–d (BSCF). © Woodhead Publishing Limited, 2010

328

Developments and innovation in CCS technology Ion transport membrane

Sweep

Oxygen

Air

O2 + 4e–

O2–

O2 + 4e–

2e– P fO 2

>

P pO 2

10.2 Schematic of a MIEC membrane (Courtesy of ECN).

Tubular membranes are prepared by extrusion of perovskite pastes with additions of organic compounds that, after burning out during sintering, provide tubular porous supports. These tubes are coated by film coating techniques, e.g. dip coating, with a dense top layer of the same material. This layer provides excellent selectivity provided that it does not contain defects. The advantage of an all-perovskite membrane configuration over coating of such a layer on other ceramics is that unwanted reaction products that could reduce the oxygen flux are prevented and furthermore a close similarity of the thermal expansion coefficient is guaranteed. The choice at ECN for tubular membranes is based on a study by Vente et al. (2006b), where monoliths, hollow fibres, tubes and fin-tube type geometries were compared in relation to surface density (m2/m3), taking into account the limited size of module (lack of maximum close packing), maximum gas velocity of 25 m/s and an oxygen flux of 10 ml/cm2·min. The monolith geometry is a good second when surface area density is considered. It is, however, quite a job to connect this monolith, in which the air and oxygen channels form a checkerboard pattern to optimally profit from the surface area density, to a manifolding system, let alone to seal the individual channels. The tubes are calculated to be 2500 mm long and have a diameter of 10–20 mm. A possible disadvantage of tubular membranes is the occurrence of fatigue caused by vibrations induced by vortices on the outside (Allam, 2009).



329

ECN membranes are made of a porous support coated with a dense layer of the same material. Pore-forming compounds used are starch and graphite particles. Additionally, some binder and lubricant are required. Each of these organic compounds burns out at its specific temperature before the actual sintering starts. In Fig. 10.3, the sinter shrinkage (solid), the mass loss during sintering (chain) and the heat caused by burning the organics (dot) is shown. The mass loss steps coincide with the heat effects of the burning of starch, polyethylene glycol (PEG) and graphite, in order of increasing temperature. These effects can also be seen in the dilatometer curve as small bumps because, at that point, momentarily the temperature is higher than that of the controlled furnace. At about 750 °C, the sintering starts and during cool down the now sintered sample shrinks proportional to its thermal expansion coefficient. Careful sintering procedures clearly must to be applied in order to prevent unwanted heat excursions during sintering and concomitant mechanical stress. These excursions, caused by catalytic combustion with the perovskite material of the pore formers, easily result in severe damage to the membranes during preparation. Small adaptations of the sintering trajectory by well-chosen dwell periods result in the desired membranes (Fig. 10.4). Photo (a) shows what happens if no proper dwell periods are used: through self-ignition of the organics the tube explodes; photo (b) shows that, with

DTA SCF/organics

TGA SCF/organics

Dilatometry SCF/organics

100.0

2.0

90.0

–3.0

80.0

–8.0

70.0

–13.0

50.0

–18.0

(dL/Lo) %

Mass (%)

60.0

40.0

–23.0

30.0

PEG A4M/starch

20.0

Graphite (SFG6)

–33.0

10.0 0.0

–28.0

0

100 200

300

400 500 600 700 800 Temperature (°C)

–38.0 900 1000 1100 1200

10.3 Sinter shrinkage, weight loss and heat evolution (not to scale) during sintering of perovskite samples with pore forming compounds (unpublished result, Courtesy of ECN).


330


5 cm

(a)

20 cm

(b)

10.4 Damaged perovskite membrane during fabrication (a) and an undamaged coated tube (b) as a result of proper selection of the sintering procedure (Courtesy of ECN).

proper dwell periods, one obtains a perfect metallic looking coating on a sintered porous support tube. The materials discussed above can be well applied in pure oxygen production. However, under reducing conditions, as are present in (partial) oxidation reactions, these compositions are not stable, especially due to their high Co content (Yaremchenko et al., 2007). Increasing the iron content is a



331

well known strategy, but also the use of the perovskite like A2BO4 + d (Fig. 10.1) is a route presently under investigation. Furthermore the presence of CO2 excludes the use of barium (e.g. Yan et al., 2007). This leads to the choice for Sr0.97Ce0.03Fe0.8Co0.2O3–d (SCFC). Figure 10.5 shows a test on a 5 cm ¥ 5 cm, 200 mm thick porous SCFC support with a 40 mm thick dense layer of the same material on top. It is clear that the flux goes up with temperature. Similarly, when the He sweep (black dots) is replaced by a reducing gas mixture, e.g. CO2, CO, H2 and H2O (grey), the flux increases. Instantaneous combustion of this mixture reduces the oxygen partial pressure and thus increases the driving force and the flux. During 400 hours on stream, there is no significant degradation. The preferred material after intensive testing of both La2Ni0.9Co0.1O4+d (LNC) and SCFC is the latter. It has a higher flux, about twice as high, and it does not contain the carcinogenic NiO. The higher the flux the smaller the required membrane area and the lower the costs.

10.4

Power generation systems integrated with oxygen selective membrane (OSM) units

From a systems point of view, the essence of air separation in oxyfuel systems is to prevent mixing of nitrogen and CO2, which would normally happen when a hydrocarbon fuel is burnt in air. As a result, oxyfuel power plants have a part where nitrogen-containing gases are present, but no gases containing any form of carbon; and a part where carbon is present, but no Oxygen flux versus time for SCFC without catalyst

5

T cell (°C) 2

O2 (ml/min*cm ) He

Tcel (°C)

950

2

O2 (ml/min* cm ) 1D

4.5 4 3.5 3

900

O2 flux (ml/mln*cm2)

1000

2.5 2

850

1.5 1

800

0.5 750

0

50

100

150

200 t (h)

250

300

350

0 400

10.5 Oxygen flux of Sr0.97Ce0.03Fe0.8Co0.2O3-d as a function of time and temperature under different sweep conditions: a He sweep and a sweep stream containing a mixture of CO2, CO, H2 and H2O (1D) (unpublished result, Courtesy of ECN).


332


nitrogen-containing gases. The interface between both parts is the oxygen generator, which can be a cryogenic ASU or an OSM (see Fig. 10.6). The choice for an oxygen generator and its demands in terms of pressures and temperatures is an important determinant in total power plant layout. A stand-alone ASU takes in air at atmospheric conditions and delivers oxygen to the carbon side at atmospheric conditions. In this case, the nitrogen side of the power plant is basically the ASU itself, without any integration with the carbon side other than the oxygen supply. Figure 10.7 shows the general system set up for OSM-based power plants with CO2 capture. The membrane forms the interface between the carbon side and the nitrogen side. Combustion of the carbon-containing fuel takes place on the carbon side. The amount of oxygen allowed to permeate to the carbon side should be close to stoichiometric. Excess of oxygen will have to be separated out from the CO2 before storage. (Depending on the carbon storage option, small quantities of excess oxygen may not be a problem. For enhanced oil recovery or enhanced coal bed methane production, it is likely to be a problem.) Less oxygen will mean incomplete combustion leading to lower efficiencies and higher exhaust gas emission levels. To prevent temperatures from becoming unacceptably high, numerous systems proposed in literature apply recycle flows of CO2 and/or steam. The pressure level on the carbon side is a trade-off between efficiency, which is favoured by a high pressure, and membrane permeation driving force, which is favoured by a low partial oxygen pressure. The current generation of OSMs requires high temperatures (in the order of 700–1000 °C) and preferably, depending on the application and economic considerations, high feed side pressures (this can vary according to the system but will typically be in the 5–40 bar range). In other words, on the nitrogen side compression and heating of the air flow is required. To maximize efficiency, the heat and pressure remaining in the nitrogen-rich retentate flow should be used as much as possible. As the heat release takes place on the carbon side, integration between the carbon side and nitrogen side and a more elaborate system on the nitrogen side is desired. Heat transfer from the carbon side to the nitrogen side can take place in heat exchangers or through fuel decarbonization, i.e., converting (part of) the fuel to hydrogen, if full CO2 capture is desired. Numerous oxyfuel power production systems proposed in literature have non-integrated ASUs, without an explicit specification of the type of ASU, as discussed in Chapter 9 (see also Kvamsdal et al., 2005; Foy and Yantovski, 2006b; IEA, 2007b; and many others). Although these ASUs can be thought of as OSM units, only systems with integrated OSM units will be studied in the remainder of this section as these provide more potential for efficiency and cost improvement than non-integrated OSM units. This is not to say that non-integrated OSM units are not a good option to replace a cryogenic


Fuel conversion and power generation

CO2, H 2O

Water removal

CO2

CO2 compression, transport and storage Carbon side Nitrogen side

O2 Air

Air separation

Depleted air

10.6 General system set up oxyfuel systems with CO2 capture and storage.



Fuel

333

334

O2

Fuel conversion and power generation

Membrane Air

O2

CO2, H 2O

Water removal

CO2

CO2 compression, transport and storage Carbon side

O2 Depleted air

10.7 General system set up oxygen selective membrane systems with CO2 capture and storage.

Nitrogen side



Fuel


335

ASU. In fact, Air Products has proposed a system comprising an OSM and gas turbine delivering high-purity oxygen that can replace cryogenic ASUs, delivering substantial cost and energy savings compared to cryogenic ASUs (see IEA, 2007b). The power generating systems to be discussed will be subdivided into four groups depending on the degree of integration of fuel conversion into the OSM unit: systems with all fuel oxidation taking place inside the membrane unit; systems with partial oxidation of the fuel in the membrane unit; systems without any fuel oxidation in the membrane unit; and systems with postoxidation of off-gas from a preceding combustion process in the membrane unit. The demands as to reduction stability of the membrane materials to be used in these systems range from severe to none, i.e. just replacing a cryogenic ASU.

10.4.1 Systems with full oxidation in the membrane unit Advantages of having full oxidation in the membrane unit are; that permeate side oxygen partial pressures can be low, increasing the driving force for permeation; and a relatively compact system design. However, membrane unit design will be complicated by a number of potential problems, such as hot spots, high temperature gradients, carbon formation and membrane degradation because of contaminants in the fuel. Careful operation is required to prevent membrane damage. Because of these difficulties, gaseous fuels are the fuel of choice for these membrane units. On a systems level, relatively low allowable maximum temperatures inside the membrane unit (typically 1000 °C) and low allowable temperature increases over the membrane unit (typically 100–200 °C) reduce electrical efficiencies, making these systems more suitable for heat (or steam) production than power production. The main examples of systems with complete oxidation in the membrane reactor are natural gas-fired oxyfuel boilers using the heat released to generate steam. For examples, readers are referred to Foy and Yantovski (2006b), IEA (2007a,b) and Switzer et al. (2005). If this steam were used to produce power, efficiencies would be limited depending on the steam cycle efficiency. Foy and Yantovski (2006b) report a 35.7 % electrical efficiency for a steam cycle with a modest steam temperature and reheat. Using ultra-supercritical steam cycles would increase this number. IEA (2007b) also mentions a Norsk Hydro system with complete oxidation in the membrane unit. The nitrogen side of this system consists of a blower and heat exchanger, and the heat released in the membrane reactor on the carbon side is used to drive a gas turbine (with CO2 as its working medium). The limited allowable membrane temperature means electrical efficiencies for this system will not be much higher than for oxyfuel boilers. Early versions of the advanced zero emission power plant (AZEP) system (to be discussed


336


in greater detail below) also showed all oxidation taking place inside the membrane unit (Sundkvist et al., 2001). However, this approach has now been abandoned in favour of systems with a separate combustion chamber and no oxidation taking place in the membrane unit.

10.4.2 Systems with partial oxidation in the membrane unit Membrane units with partial oxidation can have reducing or oxidising conditions on the permeate side, depending on the amount of fuel and oxygen supplied. If the membrane unit is primarily an oxygen generator with only some fuel added to heat the unit and supply a sweep flow, oxidizing conditions are to be expected. If the membrane unit is primarily a syngas generator with all fuel added but only limited oxygen permeation, reducing conditions are to be expected. The latter case resembles the full oxidation in the membrane unit cases in that the reducing conditions on the permeate side will lead to high driving forces for oxygen permeation. This means that an absolute pressure drop over the membrane may not be required. As for a membrane unit with complete oxidation, though, careful design and operation of the partial oxidation membrane unit will be required to avoid hot spots, high temperature gradients, local carbon formation and membrane contamination. Gaseous clean fuels will be the fuels of choice. Also, a recycle and steam reforming inside the membrane unit may be used to cool it (IEA, 2007b). If partial oxidation and steam reforming simultaneously take place, the process is generally called autothermal reforming. The case with the membrane unit acting as an oxygen generator does not benefit from high permeation driving forces. Still, the design problems mentioned in the previous paragraph are expected to be present, only reforming is now unlikely to be an option as there is an (overall) excess of oxygen. In other words, if this concept is technically feasible at all, the amount of fuel will be very limited and a large sweep will be required, which means it will have no major advantages over systems without oxidation in the membrane unit (which will be studied below), but it will have the disadvantage of a more complex OSM unit. Several authors propose schemes with OSM in autothermal reformers producing a syngas, which can be shifted to produce hydrogen and/or carbonfree power in a fuel cell or gas turbine combined cycle (see for example, Shah et al., 2001; Shah and Drnevich, 2002; Chen, 2003; Bredesen et al., 2004; Mundschau et al., 2004; IEA, 2007b). It is noteworthy that Mundschau (2004) considers biomass and coal as feasible fuels. Bredesen et al. (2004) plan to shift the syngas and separate the hydrogen in a water gas shift membrane reactor with hydrogen selective membrane. Shah and Drnevich



337

(2002) separate hydrogen by a hydrogen selective membrane in a separate reactor and Shah et al. (2001) propose integration of a hydrogen selective membrane into the ATR with oxygen selective membrane. Figure 10.8 shows an example of a system with a membrane unit acting as a syngas generator studied at ECN, partly based on the system proposed by Bredesen et al. (2004). In this and the following figures, heat streams are denoted by a dashed line and the letter Q. On the nitrogen side, a combined cycle power plant is used, based on a Siemens SGT5-4000F gas turbine. The depleted air leaving the membrane reactor is heated to the regular turbine inlet temperature by combusting a hydrogen-containing mixture coming from the carbon side. On the carbon side, the fuel is pre-reformed and mixed with a CO2–steam recycle before entering the membrane unit. The syngas produced is cooled quickly from 800 °C to 400 °C against steam cycle feedwater to prevent metal dusting.1 Subsequently, an enhanced water gas shift reactor is applied, which could, for example, be a membrane reactor or sorption enhanced reactor. An advantage of using an enhanced water gas shift reactor over a classical water gas shift reactor is that the hydrogen yield can be acceptable at higher exit temperatures. Higher exit temperatures in turn enable high-efficiency power production on the carbon side of the system. Depending on the water gas shift reactor exit conditions, a combustion step may be required to oxidize the remaining hydrogen and carbon monoxide. This combustion would need to take place in pure oxygen to avoid dilution of the CO2 flow with nitrogen. If the fractions of carbon monoxide and hydrogen are small (catalytic) post-combustion can take place in air without significant dilution of the CO2 flow. Two cases have been analysed: the system with a water gas shift membrane reactor with a hydrogen selective membrane; and the system with a sorption enhanced water gas shift reactor with a CO2 selective sorbent. For reference, a standard gas turbine combined cycle without CO2 capture based on the same gas turbine without CO2 capture would yield a 380 MW power output at 57.1 % efficiency. For the water gas shift membrane reactor case, the cycle efficiency came out at 51.0 % at a CO2 capture ratio of 99.6 % with a net power output of 406 MW. The post-combustion in air led to an 8.2 mole % nitrogen concentration in the compressed CO2. For the sorption enhanced water gas shift reactor, the efficiency was 50.0 % at a CO2 capture ratio of 90.7 % and power output of 357 MW. The CO2 capture ratio is lower 1

In practice, this membrane reactor is more than a syngas generator. As it heats the prereformer exit flow on the carbon side and the air flow on the nitrogen side, the heat release inside the reactor is substantial and, consequently, the resulting syngas has a lower CO/ CO2 ratio than typical syngases, which means that the metal dusting tendency may not be as bad as for other syngases.


338

Feedwater pre-heat and evaporation

H2, H2O, …

T = 800 °C Recycle blower

T= 800 °C

CO2 H 2O Postcombustion in air

CO2, H 2O turbine

CO2–H2O steam cycle

Water removal

T = 1000 °C Carbon side

Membrane

Nitrogen side

O2 Q

O2

T= 1000 °C

Q

Air

CO2 compression, transport and storage

Partial oxidation

Air compressor


Fuel and steam

Prereformer

Enhanced water gas T= shift reactor 400 °C

10.8 Partial oxidation power cycle.

Air turbine Depleted air steam cycle

Depleted air


H 2O


339

than for the water gas shift membrane reactor case because the CO2 slip through the sorbent is expected to be larger than CO2 permeation through the membrane. This CO2 ends up in the nitrogen side gas turbine combustion chamber. These systems have capture penalties of 6–7 percentage points, which is mostly due to the inefficient use of heat during the quick syngas cooling (to prevent metal dusting) and to the fact that the carbon side is not a full combined cycle (for both cases only a CO2–H2O turbine was used, no CO2–H2O steam cycle). Further optimization is expected to improve these numbers for this type of cycle.

10.4.3 Systems without oxidation in the membrane unit Membrane unit design will be easier if no oxidation takes place inside. A drawback of this arrangement, however, is a significant oxygen partial pressure on the permeate side, necessitating an absolute pressure drop over the membrane and/or a large sweep flow on the permeate side. To accommodate a pressure drop over the membrane, compression of air on the nitrogen side in a gas turbine (possibly combined cycle) arrangement seems a particularly favourable option to efficiently recapture part of the energy from the depleted air leaving the membrane unit. The most extensively studied power plant scheme in this category is arguably the AZEP cycle, originally proposed by Norsk Hydro. Several variants of this scheme have been proposed (see, for example, Bücker et al., 2005; Fredriksson Möller et al., 2006 and Anantharaman et al., 2009). The most basic one, with full CO2 capture, is shown in Fig. 10.9. On the nitrogen side, this system has a gas turbine combined cycle. In the MCM (mixed conducting membrane) reactor, oxygen permeation takes place. Extensive heat exchange is used to heat the air flow on the nitrogen side. On the carbon side, a large recycle, mainly consisting of CO2 and steam, is used to promote oxygen permeation. This oxygen is led to a separate combustion chamber. The heat in the ongoing flow of combustion products is partly transferred to air in the bleed gas heat exchanger and, following that, it is turned into power in a steam cycle. The AZEP cycle uses natural gas as fuel. Cycles without combustion in the membrane unit are better suited to combust more challenging fuels, such as coal or biomass that would as a solid or a slurry mechanically deteriorate the membrane when combusted inside the membrane unit. The AZEP concept is not well suited to burning these fuels either, as the recycle needs to be clean to avoid membrane degradation. High-temperature gas cleaning, a far from mature technology, could in time overcome this problem. For current low-temperature gas cleaning to be applied, extensive cooling and extensive reheating would be required for the combustion exhaust gases to serve as recycle, strongly decreasing system efficiency.


340

Recycle blower

MCM reactor

Q

Combustion chamber

Membrane

Q

Air

Air compressor


Q

O2 Q


Water removal

Carbon side

Q

Nitrogen side

Q O2 Bleed gas heat exchanger



10.9 The AZEP cycle, full CO2 capture (Anantharaman et al., 2009, adapted).

Depleted air


Fuel preheater

Fuel


341

A drawback of this cycle is a reduced turbine inlet temperature (compared to natural gas-fired gas turbines) on the nitrogen side due to heat exchanger temperature limits. Modern combined cycle gas turbines have turbine inlet temperatures in the range 1300–1500 °C, whereas gas–gas heat exchanger temperatures over 1000 °C are challenging, requiring expensive hightemperature materials. This reduction in turbine inlet temperature will lead to a substantial reduction in cycle efficiency. Aspen Plus simulations performed at ECN, using a Siemens SGT5-4000F gas turbine combined cycle on the nitrogen side, forecast a 50.4 % efficiency for this system at full CO2 capture (electric output would be 267 MW).2 This 6.7 percentage point efficiency penalty is quite acceptable compared to the penalties for other power schemes with CO2 capture. Other sources come to comparable capture penalties: Anantharaman et al. (2009) 7.9 percentage points; and Frederiksson Möller et al. (2006) 8.3 percentage points. Kvamsdal et al. (2005) report 5.6 percentage points but use a turbine on the carbon side instead of a CO2–H2O steam cycle. An important variant addresses the lower turbine inlet temperature on the nitrogen side by combusting natural gas in the nitrogen side gas turbine. In-house simulations show that efficiency can go up to 52.7 % (CO2 capture penalty 4.4 percentage points); at the expense of a lower CO2 capture ratio of 89.5 % (power output would be 311 MW). Anantharaman et al. (2009) report a 3.9 percentage points penalty (at 84.3 CO2 capture); Frederiksson Möller et al. (2006) 4.5 percentave points (at 85 % CO2 capture); Kvamsdal et al. (2005) report a 2.4 percentage points penalty (at 85 % CO2 capture), but for a system with a turbine on the carbon side instead of a CO2–H2O steam cycle. This efficiency is still considerably lower than the 57.1 % obtained without CO2 capture, partly because the power production part on the carbon side of the system is only a steam cycle. To extend the carbon side to a full combined cycle a turbine was added, expanding the CO2–steam flow before passing through the steam cycle (see Fig. 10.10). Simulations showed this could improve efficiency to 53.9 % with a CO2 capture ratio of 90.1 % and a power output of 343 MW. Another example of a power cycle without oxidation in the membrane unit is the ZEITMOP (zero emission ion transport membrane oxygen power) cycle (see Fig. 10.11) (Foy and Yantovski, 2006b). Like the AZEP cycle, this system has a separate combustion chamber directly following the membrane unit on the carbon side. Only here, the recycle going back to the membrane unit is not split from the combustion chamber outlet flow, but from the CO2 flow after compression. This has some important consequences. One is that the combustion chamber temperature is no longer limited by heat exchanger 2

In these simulations the heat exchanger exit temperature was assumed to be 1250 °C. If an exit temperature of 1000 °C were to be assumed, the efficiency would be lower.


342

Recycle blower

CO2, H 2O turbine

Combustion chamber


Q Membrane

Q

Air

Air compressor


MCM reactor

O2

Q

Water removal

Q

Carbon side Nitrogen side

Q O2 Bleed gas heat exchanger



Depleted air

10.10 The AZEP cycle, with carbon side turbine and additional firing on nitrogen side.


Q

Fuel pre-heater

Fuel


Q

Fuel CO2, H 2O turbine

CO2, O2

CO2 Membrane O2

Air compressor

CO2

CO2 transport and storage


O2 Q

Air

Water removal and CO2 compression

Air turbine Depleted air

10.11 The ZEITMOP cycle (Foy and Yantovski, 2006b, adapted).


CO2, turbine

343

344


considerations but by turbine considerations. Modern gas turbine technology has increased turbine inlet temperatures to considerable levels, which continue to be increased. Also, the recycle flow is cooled to low temperatures before CO2 compression, enabling low-temperature gas cleaning. It is not likely, though, that this will mean increased tolerance for contaminated fuels as turbines and heat exchangers are generally sensitive to these contaminants. Furthermore, the recycle and sweep flow through the membrane unit consist of pure CO2 instead of a CO2–steam mixture. An important drawback of not splitting the recycle until after CO 2 compression is that a large part of the system will be substantially larger than for the AZEP cycle for the same power output (including heat exchangers and the CO2 compressor), leading to higher cost. Another important difference with the AZEP cycle is that less heat is transferred to the nitrogen side, which means a lower turbine inlet temperature on that side and no combined cycle arrangement. (Of course, this could be circumvented by installing a combustion chamber on the nitrogen side but, compared to AZEP, more fuel would need to be combusted to increase the turbine inlet temperature to comparable levels.) Still, Foy and Yantovski (2006b) report a 46 % efficiency for this system at a turbine inlet temperature of 1300 °C, and even a 56 % efficiency at a turbine inlet temperature of 1500 °C, which compares well with the AZEP system’s efficiency. Several authors propose coal-fired oxyfuel combustors/boilers with external OSM units (see, for example, Nsakala et al., 2004; Switzer et al., 2005; IEA, 2007a; Beggel et al., 2009; Leo et al., 2009; Stadler et al., 2009). Compared to oxyfuel boilers with the membrane integrated into the boiler, these systems have enhanced fuel flexibility although, as for the systems discussed in the previous paragraphs, if recycles are used as sweep flows, dirty fuels require high-temperature gas cleaning to maximize electrical efficiency. These efficiencies are expected to be in the same range as their counterparts with full combustion in the membrane unit (typically 35–42 %, depending on the steam cycle). Figure 10.12 shows a well-studied example of a power plant cycle with a coal-fired oxyfuel boiler, the OXYCOAL–AC process, proposed by RWTH Aachen University. Like the standard AZEP configuration and ZEITMOP configuration, heat transfer to the nitrogen side is limited reducing cycle efficiency. Stadler et al. (2009) report a 37.4 % cycle efficiency, which means an 8.5 percentage-point capture penalty relative to their reference coal boiler cycle with 45.9 % efficiency. Further optimization and more optimistic assumptions could lift the efficiency to 41.1 %. If a natural gas-fired combustion chamber would be included on the nitrogen side to increase the turbine inlet temperature, efficiency was shown to increase to 46.2 % even (Stadler et al., 2009). Beggel et al. (2009) studied a modified version of this process with a vacuum pump on the OSM permeate side instead of a flue


Steam

Recycle blower

O2

Air

High-temperature gas cleaning

Steam cycle

Water removal

Membrane

Q

Air compressor


Fuel combustor/ boiler



O2

Air turbine

Hot feedwater

Feedwater pre-heat

10.12 The OXYCOAL–AC cycle (Stadler et al., 2009, adapted).

Depleted air


Fuel

345

346


gas recycle. Their motivation for this was observed membrane degradation under influence of the flue gas recycle. Optimization showed that (relative to the same reference boiler) the capture penalty could be reduced to 6.2 percentage points, in other words, an efficiency of 39.7 %. Another oxyfuel combustor/boiler cycle with external OSM unit is the so-called Milano cycle (Foy and Yantovski, 2006b). This cycle is largely comparable to the OXYCOAL–AC cycle with the main exception that the Milano cycle uses a fluidized bed combustor. The expected efficiency of the cycle is around 41–42 % (Foy and Yantovski, 2006b). Nsakala et al. (2004) propose a circulating moving bed reactor combusting coal in oxygen supplied by a separate OSM unit. Foy and McGovern (2006) propose a set-up combining an SOFC with an OSM unit. The main fuel oxidation takes place in the SOFC (which keeps nitrogen-containing air separate from fuel and flue gas flows), but not all fuel can be oxidized in the SOFC to prevent locally oxidizing conditions. The OSM uses a CO2–steam recycle to provide oxygen for post-combustion of the SOFC flue gas in a separate combustion chamber. This system is largely comparable to systems with post-oxidation in the membrane unit as discussed in the following section.

10.4.4 Systems with post-oxidation in the membrane unit OSM units can also be used to oxidize fuel tailings, left over from another (partial) oxidation process. Depending on whether the oxidation will be slightly oxygen rich or lean, oxidizing or reducing conditions can be present at the permeate side. In the latter case, driving forces for oxygen permeation will be higher. Care should be taken to avoid problems such as hot spots, high temperature gradients and carbon formation. As these membrane units are positioned after the main combustion processes, the systems may be fired with fuels such as coal and biomass as long as the syngas flow has been thoroughly cleaned before entering the membrane unit. Since the required membrane unit inlet temperatures are high, this gas cleaning may need to take place at high temperature to achieve high electrical efficiencies. Numerous authors propose OSM units to post-oxidize SOFC anode off-gases (see for example Riensche et al., 2000; Shockling et al., 2001; Maurstad et al., 2005; Foy and Yantovski, 2006a; IEA, 2007a). Granovskii et al. (2006) report a scheme with an OSM unit post-oxidizing SOFC flue gases, to produce a syngas to be fed to a combined cycle power plant. This scheme will need an additional separation step to separate the CO2 from the syngas flow. SOFCs are a promising technology for power production with CO2 capture thanks to their high efficiencies and inherent separation of nitrogen and carbon, yet their fuel utilization is limited (usually 85 % is proposed) to prevent oxidizing conditions locally in the SOFC. OSMs allow



347

using the remaining heat in the carbon-containing anode off-gas flow to be used without mixing it with the nitrogen-rich cathode off-gases. Most schemes proposed combine the SOFC and OSM with a gas turbine cycle; their main difference lies in heat integration. High efficiencies can be expected from this set-up. Maurstad et al. (2005) find 64.8 % at near complete CO2 capture and Riensche et al. (2000) 64.0 %, also at around full CO2 capture. Figure 10.13 shows an example of such a power cycle combining an SOFC with an OSM and gas turbine studied in-house. In the actual model analysis, two stacked SOFCs were used. Stacking reduces the required air flows and therewith increases system efficiency. All the heat production in this system takes place on the carbon side; heat transfer to the nitrogen side takes place in the SOFCs and the OSM unit. Both sides are pressurized. Neither side has a steam cycle as the turbine outlet temperatures are low and the heat available in the turbine outlet flows is largely required in other parts of the system. On both the nitrogen and carbon side, there is extensive heat transfer, mainly due to the stringent inlet temperature and maximum temperature increase requirements of the SOFC and OSM unit. In line with practical developments in SOFC sizing the system size was chosen at 1 MWe electrical power output. Consequently, a micro turbine with a lower pressure ratio was selected instead of a heavy-duty gas turbine on the nitrogen side. The thermodynamic efficiency of this cycle was found to be 68.6 % at full CO2 capture. This value is not too far from the numbers from literature mentioned above. Pre-combustion decarbonization systems separating hydrogen from a syngas flow usually have carbon monoxide and hydrogen left in the shifted syngas flow, making them prime candidates to apply post-oxidation OSM units. In terms of efficiency, these systems may be promising, although they do combine pre-combustion decarbonization technology with OSM burner technology, which are both currently unproven on a large scale. Finally, IEA (2007b) shows a Praxair system using OSM reactors to postoxidize a syngas produced by an oxygen-blown gasifier. In the first step, the syngas is partially oxidized followed by expansion. The second step is an OSM boiler producing steam. It is not straightforward, however, that separate OSM reactors are a better option here than enlarging the gasifier’s oxygen supply, a cryogenic ASU in this particular system, which would benefit from economies of scale. As the heat release is mainly used for steam production, the system efficiency is limited to 34.5 %. The system contains a syngas expander which will deliver some power too. The performance characteristics of the various systems described are given in Table 10.1.

10.5

Advantages and limitations

Based on the preceding sections, the techno-economic viability will be assessed. Technological gaps to be closed are still many. The most urgent


348

CO2, H 2O turbine

Q

Desulphurization

Prereformer

SOFC

O2

Water removal

Residual fuel oxidation

Anode Cathode

Q O2

Q

10.13 SOFC–GT post-oxidation power cycle.

Air turbine

Q

Air



Membrane

O2

Air compressor


Natural gas

Depleted air


Q Q

Table 10.1 Performance comparison of systems mentioned in this section Electrical Carbon capture efficiency (%) ratio (%)

System size Carbon capture Reference (MWe) penalty (%pt)

No capture reference cases SGT5–4000F combined cycle base case SOFC–GT base case

57.1 70.9

380 1.0

N/A N/A

Own numbers Own numbers

Full oxidation in membrane unit Oxyfuel boiler

35.7

~100

Unknown

Unknown

Foy and Yantovski (2006b)

Partial oxidation in membrane unit CO2 sorbent SE–WGS reactor case H2 membrane WGSMR reactor case

50.0 51.0

91 100

357 406

7.1 6.1

Own numbers Own numbers

0.0 0.0

No oxidation in membrane unit AZEP 100 % capture 50.4 100 267 6.7 49.6 100 248 8.3 49.3 100 299 7.9 AZEP N-side combustion chamber, 52.7 90 311 4.4 no carbon side turbine 53.4 85 300 4.5 53.3 84 393 3.9 AZEP N-side combustion chamber, 53.9 90 343 3.2 with carbon side turbine (and steam cycle) ZEITMOP cycle 46–56 ~100 Unknown Unknown OXYCOAL–AC cycle standard case 37.41 90 501 8.5 OXYCOAL–AC cycle vacuum pump case 39.7 ~90 Unknown 6.2 Milano cycle 41.9 Unknown 415 Unknown

Own numbers Frederiksson Möller et al. (2006) Anantharaman et al. (2009) Own numbers Frederiksson Möller et al. (2006) Anantharaman et al. (2009) Own numbers

Foy and Yantovski (2006b) Stadler et al. (2009) Beggel et al. (2009) Foy and Yantovski (2006b)



System description

349


Post-oxidation in membrane unit SOFC–GT with oxygen selective membrane Maurstad SOFC–GT with oxygen selective membrane Riensche SOFC–GT with oxygen selective membrane in-house Praxair POx reactor and oxygen boiler after oxygen blown gasifier 1

Electrical Carbon capture efficiency (%) ratio (%)

System size Carbon capture Reference (MWe) penalty (%pt)

64.8

~100

20

5.7

Maurstad et al. (2005)

64.0

~100

Unknown

Unknown

Riensche et al. (2000)

68.6

100

1.0

2.3

Own numbers

34.5

~100

Unknown

Unknown

IEA (2007b)

Efficiencies up to 41.1 % possible iwth optimistic assumptions.


System description

350

Table 10.1 Continued


351

one is high-temperature high-pressure sealing. For lab experiments, this can be done – one can even use cold seals outside the heated zone – but on an industrial scale the sealing issue is still not solved. Connected to this is the reproducibility of the production of the membrane tubes, e.g. their circularity and roughness. Special care should be taken for the high-flux materials like SCF and BSCF that show substantial creep (Majkic et al., 2000) at high temperatures, thus making compressive seals impossible. The lower flux materials like SCFC do not suffer from this drawback. The challenge for the high-flux materials is to reduce creep and for low-flux materials exhibiting no creep to increase the flux. Furthermore extended durability tests are needed, much longer than the one shown in Fig. 10.5 which lasted only for two and a half weeks. Harsher conditions should be tested as well, for instance when (partial) oxidation in the membrane unit is foreseen. The results shown in this chapter pertain to a mildly reducing atmosphere from an SOFC off-gas stream. The previous sections discussed power plant systems with OSM units in terms of electrical efficiencies. Several schemes show relatively low efficiency penalties for CO2 capture, some even as low as two percentage points. Whether they will eventually result in full-scale power plants depends not only on electrical efficiencies but on investment cost and operational cost and revenues. Oxyfuel systems with OSMs are likely to suffer from high investment cost. This is firstly because of high cost of OSM units themselves, although cost savings for the oxygen production section of 35–48 % have been reported for membrane-based systems compared to cryogenic ASUs by Air Products (IEA, 2007b). However, in this chapter it has also become clear that these systems require extensive heat exchange, as the OSM units require high inlet temperatures of flows on both the nitrogen and carbon sides, where heat is only generated on the carbon side. The AZEP system with full CO2 capture, for example, equipped with a gas turbine combined cycle on the nitrogen side, requires in the order of 530 MW of heat exchange for a net electricity output in the order of 265 MW (i.e. around double of the electricity output needs to be exchanged). Add to this the high working temperatures of these heat exchangers (partly in excess of 1000 °C) and the low achievable heat transfer coefficients for these gas–gas heat exchangers at fairly mild pressures and it is clear that heat exchanger cost will be substantial. Norsk Hydro expects to use monoliths for heat exchange in the MCM reactor (see e.g. Selimovic, 2005; Frederiksson Möller et al., 2006; IEA, 2007b). However, Vente et al. (2006b) report that in practical membrane units, monoliths by themselves may not lead to cost reductions, partly due to complicated manifolding and sealing. Systems with partial oxidation in the membrane reactor are likely to fare better in terms of heat transfer requirements as heat can be transferred from


352


the carbon side to the nitrogen side in the form of hydrogen. However, the membrane unit itself will be more expensive due to increased complexity caused by partial oxidation, and an extra hydrogen separation step is required. Consequently, although several of the systems shown in this chapter perform well from an electric efficiency point of view, their economic viability needs to be studied in more detail. Of course this should be compared with the economic viability of other CO2 capture technologies like pre- and postcombustion capture.

10.6

Future trends

Our view on how this field will grow and change in the future is presented in the following. Air Products already has a pilot running for the production of 1–5 t oxygen per day (IEA, 2007b), which is still about four orders of magnitude lower than needed for a GW scale power plant. So for pure oxygen production, up-scaling is a major issue. A currently unresolved issue is the degradation of membrane materials by CO2, especially those containing Ba. This is an important issue for application of oxygen membranes in highefficiency oxyfuel power plants. The higher the temperature of operation the smaller the problem is, due to reduced stability of carbonates. Research should especially be devoted to developing materials and systems which are stable under reducing conditions, especially for partial oxidations in the membrane unit. Regular benchmarking with other novel developments, e.g. in solid sorption-based pre- and post-combustion capture, and CLC is necessary to assess the practical value of this research.

10.7


∑ Who’s Who in industrial gases: – The Linde Group (Munich, Germany; www.linde.com) – Air Liquide (Paris, France; www.airliquide.com) – Air Products + Consortium (Allentown, Pa. USA; www.airproducts. com) – Praxair + Consortium (Danbury, Conn. USA; www.praxair.com) – Taijo Nippon Sanso (Japan, www.tn-sanso.co.jp) – Messer Group (Sulzbach bei Frankfurt am Main, Germany; www. messergroup.com) ∑ Who’s Who in academic MIEC materials research – Balachandran (Argonne, USA) – Bouwmeester (Twente, the Netherlands) – Jacobson (Houston, USA) – Kharton (Aveiro, Portugal)



∑

353

– Caro (Hanover, Germany) – Xu (Nanjing, China) – Diniz da Costa (Queensland, Australia) Who’s Who in MIEC research in research institutes – Bredesen (SINTEF, Norway) – Haije/Vente (ECN, the Netherlands) – Meulenberg (FZJ, Germany) – Bonanos (Risø, Denmark) – Yang (Dalian, China) – Julbe, (Institut Européen des Membranes, France) – Voigt (Hermsdorfer Institut für Technische Keramik e.V., Hermsdorf, Germany)

10.8

Conclusions

Application of OSMs in oxyfuel power systems shows low efficiency penalties associated with CO2 capture, especially for cases with reducing conditions on the permeate side. Also cost benefits have been reported in literature for oxygen membranes compared to conventional cryogenic ASUs. When all options for oxygen production as described in Section 10.2 are considered with a view to energy penalty and cost reduction probably the most serious competitor for OSMs for use on a GW scale is CLC. For gas-fired systems, efficiencies up to 53 % have been calculated (Naqvi and Bolland, 2007). This technology is also still in the research phase, and likewise facing materials problems, especially concerning oxygen carrier stability under oxidation-reduction cycles. Also materials problems inherent to high-temperature heat exchanging and gas–solid separations in cyclones are clearly present. However, OSMs are also clearly not yet available for application on GW scale under the conditions needed. There are candidate materials that most of the researchers agree on, but long-term stability under reducing conditions is still a problem. Mechanical stability versus high flux is the crux for future research. The outcome may be that only lower flux materials have industrially acceptable stability, but this means that membrane area, and with that costs, will go up. Manifolding and sealing are equally important issues to be solved before this technology is ready for market introduction. This section concludes with an overview of proposed power plant cycles with CO2 capture using OSMs integrated into the power cycle. Like all oxyfuel systems, these systems have a nitrogen side where air separation takes place, and a carbon side where the fuel is oxidized. The OSM unit is the main interface between both sides. The system layout is largely shaped by the OSM requirements of high temperatures and high feed side pressures and a preference for sweep flows. The first consequence is that


354


large heat transfer is required from the carbon side to the nitrogen side. As high efficiencies demand recapturing of the heat and compression energy in the nitrogen-rich membrane retentate flow, these systems often have a gas turbine arrangement, possibly with a steam cycle, on the nitrogen side. Systems with full oxidation of natural gas inside the membrane reactor suffer from low maximum cycle temperatures reducing their electrical efficiency (the efficiency is typically 35–42 %). The main systems found falling into this category are oxyfuel boilers fed with gaseous, clean fuels. Because these fuels can be used to produce carbon-free power at higher efficiencies in other systems, these systems are more suited for heat production than power production. Numerous systems with OSMs in ATRs have been proposed for hydrogen and/or power production. A plus for these systems is that hydrogen is available that can be burnt for carbon-free heat release on the nitrogen side. A drawback is that a second separation step is required to separate this hydrogen. Together with other practical considerations (such as metal dusting), system electrical efficiencies are typically 48–52 %. Systems without oxidation in the OSM unit have better fuel flexibility than systems with oxidation in the membrane unit, although high-temperature gas cleaning is usually required for high efficiencies if coal or biomass is used as fuel. Several high-efficiency power cycles using natural gas are known (mainly the AZEP scheme with efficiencies up to 54 %). The main cycles running on coal or biomass in this category are oxyfuel boilers. The efficiencies of systems with post-oxidation of a fuel taking place in the membrane unit will to a large extent depend on the preceding fuel oxidation step. Most systems falling into this category propose post-oxidizing SOFC anode off-gases. These systems can reach efficiencies in the range 64–70 %. However, SOFCs are only available in the MW range.

10.9

References

Acharya D, Krishnamurthy K R, Leison M, MacAdam S, Sethi V K, Anheden M, Jordal K and Yan J (2006) ‘Development of a high temperature oxygen generation process and its application to oxycombustion power plants with carbon dioxide capture’, 22nd Annual Pittsburgh Coal Conference, 11–15 September, Pittsburgh, PA. Air Products and Chemicals (2002), PRISM®Oxygen VSA System, Allentown, PA, available at: http://www.airproducts.co.uk/Gas_Generation/pdf/O2_VSA.pdf (accessed January 2010). Allam R (2009), Private communication with W G Haije. Anantharaman R, Bolland O and Asen K I (2009), ‘Novel cycles for power generation with CO2 capture using OMCM technology’, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1 335–342.



355

Beggel F, Engels S, Modigell M and Nauels N (2009), ‘Oxyfuel combustion by means of high temperature membranes for air separation’, 4th International Conference on Clean Coal Technologies: CCT2009, 18–21 May, Dresden, Germany. Bredesen R, Jordal K and Bolland O (2004), ‘High-temperature membranes in power generation with CO2 capture’, Chemical Engineering and Processing 43 1129– 1158. Bücker D, Holmberg D and Griffin T (2005), ‘Techno-economic evaluation of an oxyfuel power plant using mixed conducting membranes’, in Thomas DC and Benson SM (eds), Carbon dioxide capture for storage in deep geological formations – Results from the CO2 capture project: Volume one: Capture and separation of carbon dioxide from combustion sources. elsevier, Oxford, UK, 537–559. Chen C M (2003), Engineering Development of Ceramic Membrane Reactor Systems for Converting Natural Gas to Hydrogen and Synthesis Gas for Liquid Transportation Fuels (DE-FC26-97FT96052), US Dept. of Energy Hydrogen Program Annual Review. Darde A, Prabhakar R, Tranier J P and Perrin N (2009), ‘Air separation and flue gas compression and purification units for oxy-coal combustion systems’, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1 527–534. Foy K and McGovern J (2006), ‘OFFCET – A novel power production cycle concept’, 5th Annual Conference on Carbon Capture & Sequestration, 8–11 May, Alexandria, VA. Foy K and Yantovski E (2006a), ‘Fuel-fired zero emissions power plant cycles with oxygen ion transport membranes (history and state-of-the-art)’, 5th Annual Conference on Carbon Capture & Sequestration, 8–11 May, Alexandria, VA. Foy K and Yantovski E (2006b), ‘History and state-of-the-art of fuel fired zero emission power cycles’, International Journal of Thermodynamics 9(2) 37–63. Fredriksson Möller B, Torisson T, Assadi M, Sundkvist S G, Sjödin M, Klang Å, Åsen K I, Wilhelmsen K (2006), ‘AZEP gas turbine combined cycle power plants – thermoeconomic analysis’, International Journal of Thermodynamics 9(1) 21–28. Granovskii M, Dincer I and Rosen M A (2006), ‘Application of oxygen ion-conductive membranes for simultaneous electricity and hydrogen generation’, Chemical Engineering Journal 120 193–202. IEA (2007a), CO2 Capture From Medium Scale Combustion Installations, 2007/7, IEA Greenhouse Gas R&D Programme, Cheltenham, UK. IEA (2007b), Improved Oxygen production Technologies, 2007/14, IEA Greenhouse Gas R&D Programme, Cheltenham, UK. IEA (2008), Future Developments in IGCC, CCC/143, IEA Clean Coal Centre, London, UK. IPCC (2007), Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, Pachauri R.K and Reisinger A. (eds)], IPCC, Geneva, Switzerland, available at: http://www.ipcc.ch/publications_and_data/publications_ipcc_ fourth_assessment_report_synthesis_report.htm (accessed January 2010). Jansen D, Asbroek N A M ten, Loo S van, Meuleman E E B, Dorst E van, Geuzenbroek F, Ploumen P, Kamphuis H and Rijen S van (2008), EOS-CAPTECH-Integration of CO2 capture technologies for new power plants in the Netherlands, SubProject SP5 Integration, Deliverable D1 second year (2007), KEMA report 30613029-Consulting 08-1575, July, available at: http://www.co2-captech.nl/files/File/Rapport_08-1575_ Captech_Integration_Deliverable_2007.pdf (accessed January 2010). © Woodhead Publishing Limited, 2010

356


Kvamsdal H M, Maurstad O, Jordal K and Bolland O (2005), ‘Benchmarking of gasturbine cycles with CO2 capture’, in Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 1, 233–241. Kumar R (1996), ‘Vacuum swing adsorption process for oxygen production – a historical perspective’, Separation Science and Technology 31 877–893. Leo A, Liu S and Diniz da Costa J C (2009), ‘Development of mixed conducting membranes for clean coal energy delivery’, International Journal of Greenhouse Gas Control 3(4) 357–367. Lyngfelt A, Leckner B and Mattisson T (2001), ‘A fluidized-bed combustion process with inherent CO2 separation; application of chemical-looping combustion’, Chemical Engineering Science 56 3101–3113. Lyon R K (1996), Method and Apparatus for Unmixed Combustion as an Alternative to Fire, US5509362. Majkic G, Wheeler L and Salama K (2000), ‘Creep of polycrystalline SrCo0.8Fe0.2O3–d’, Acta Materialia 48 1907. Maurstad O, Bredesen R, Bolland O, Kvamsdal H M and Schell M (2005), ‘SOFC and gas turbine power systems – evaluation of configurations for CO2 capture’, in Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 1, 273–281. McIntosh S, Vente J F, Haije W G, Blank D H A and Bouwmeester H J M (2006), ‘Phase stability and oxygen stoichiometry of SrCo0.8Fe0.2O3–d measured by in-situ neutron diffraction’, Solid State Ionics 177 833–842. Mundschau M V, Xie X, Evenson C R and Sammells A F (2004), ‘Simultaneous hydrocarbon reforming, carbon dioxide sequestration and hydrogen separation using dense inorganic membranes’, 3rd Annual Conference on Carbon Capture & Sequestration, 3–6 May, Alexandria, VA, available at: www.netl.doe.gov/publications/proceedings/04/carbonseq/047.pdf (accessed January 2010). Naqvi R and Bolland O (2007), ‘Multi-stage chemical looping combustion (CLC) for combined cycles with CO2 capture’, International Journal of Greenhouse Gas Control 1 19–30. Nsakala N, Liljedahl G N, Marion J, Levasseur A A, Turek D, Chamberland R, MacWhinnie R, Morin J and Cohen K (2004), ‘Oxygen-fired circulating fluidized bed boilers for greenhouse gas emissions control and other applications’, 3rd Annual Conference on Carbon Capture & Sequestration, 3–6 May, Alexandria, VA. Rezvani S, Bolland O, Franco F, Huang Y, Span R, Keyser J, Sander F, McIlveen-Wright D and Hewit N (2009), ‘Natural gas oxy-fuel cycles – Part 3: economic evaluation’, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1 565–572. Riensche E, Achenbach E, Froning D, Haines M R, Heidug W K, Lokurlu A and Andrian S von (2000), ‘Clean combined-cycle sofc power plant – cell modelling and process analysis’, Journal of Power Sources 86 404–410. Ryden M and Lyngfelt A (2006), ‘Using steam reforming to produce hydrogen with carbon dioxide capture by chemical-looping combustion’, International Journal of Hydrogen Energy 31 1271–1283. Selimovic F (2005) Modelling of Transport Phenomena in Monolithic Structures related to CO2-Emission Free Power Process, available at: http://130.235.81.176/~ht/documents/ lic-pres.pdf (accessed December 2009). © Woodhead Publishing Limited, 2010


357

Shah M M and Drnevich R F (2002), ‘Economic feasibility analysis of hydrogen production by integrated ceramic membrane system’, 2002 DOE Hydrogen Program Review, NREL/CP-610–32405. Shah M M, Drnevich R F, Balachandran U, Dorris S E and Lee T H (2001), Technoeconomic feasibility analysis of hydrogen production by integrated ceramic membrane system’, 2001 DOE Hydrogen Program Review, NREL/CP-570–30535. Shockling L A, Huang K, Gilboy T E, Maxwell Christie G, Raybold T M (2001), Zero Emission Power Plants Using Solid Oxide Fuel Cells and Oxygen Transport Membranes, DOE, Office of Scientific & Technical Information website, available at: http://www.osti.gov/bridge/servlets/purl/832845-EDQk0q/native/832845.pdf (accessed December, 2009). Stadler H, Beggel F, Persigehl B, Kneer R, Modigell M and Jeschke P (2009), ‘The OXYCOAL-AC process: Component behaviour and thermodynamic efficiency’, 4th International Conference on Clean Coal Technologies, 18–21 May, Dresden, Germany. Strömberg, L, Lindgren, G, Jacoby, J, Giering, R, Anheden, M, Burchhardt, U, Altmann, H, Kluger, F, Stamatelopoulos, GN (2009), ‘Update on Vattenfall’s 30 MWth oxyfuel pilot plant in Schwarze Pumpe’, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia 1 581–589. Sundkvist S G, Griffin T and Thorshaug N P (2001), ‘AZEP – development of an integrated air separation membrane – gas turbine’, 2nd Nordic Minisymposium on Carbon Dioxide Capture and Storage, 26 October, Göteborg, Sweden, at available: http://www.entek. chalmers.se/~anly/symp/01sundkvist.pdf (accessed December, 2009). Switzer L, Rosen L, Thompson D, Sirman J, Howard H, Bool L (2005), ‘Cost and feasibility study on the Praxair advanced boiler for the CO2 capture project’s refinery scenario’, in Thomas D and Benson S (eds), Carbon dioxide capture for storage in deep geological formations – Results from the CO2 Capture project: Volume one: Capture and separation of carbon dioxide from combustion sources. 561–581, Elsevier, Oxford, UK. Vente J F, Haije W G and Rak Z S (2006a), ‘Performance of functional perovskite membranes for oxygen production’, Journal of Membrane Science 276(1–2), 178–184. Vente J F, Haije W G, IJpelaan R and Rusting F T (2006b), ‘On the full-scale module design of an air separation unit using mixed ionic electronic conducting membranes’, Journal of Membrane Science 278 66–71. WRI (2007), Development of a Novel Oxygen Supply Process and its Integration with an Oxy-Fuel Coal-Fired Boiler, Final report under DE-FC26-98FT40323, Western Research Institute, Laramie, Wy, February. Yan A, Maragou V, Arico A, Cheng M and Tsiakaras P (2007), ‘Investigation of a Ba0.5Sr0.5Co0.8Fe0.2O3–d based cathode SOFC II. The effect of CO2 on the chemical stability’, Applied Catalysis B: Environment 76 320–327. Yaramchenko A A, Kharton V V, Avdeev M, Shaula A L and Marques F M B (2007), ‘Oxygen permeability, thermal expansion and stability of SrCo0.8Fe0.2O3–d-SrAl2O4 composites’, Solid State Ionics 178(19–20) 1205–1217.


11

Chemical-looping combustion systems and technology for carbon dioxide (CO2) capture in power plants E. J. A n t h o n y, CANMET Energy Technology Centre-Ottawa, Canada Abstract: Chemical-looping combustion is a completely new method of oxidizing fuels without the direct use of either air or oxygen. Using this approach, the fuel, usually natural gas in test work done to date, is oxidized by an oxygen carrier, which is then regenerated by reaction in air. For most, but not all, oxygen carriers, the fuel conversion process is endothermic while the regeneration or oxidizing step is strongly exothermic. Overall, the process is equivalent to direct burning of the fuel, but without the presence of the nitrogen in the air. The gases from the fuel reactor contain only carbon dioxide (CO2) and H2O with minimal production of nitrogen oxides (NOx), so that after condensation of the water, a pure stream of CO2 results, which is suitable for compression and sequestration in a CO2 capture and sequestration process. The research to date, which is described here, has concentrated mainly on developing suitable carriers, often based on Ni, Cu and Fe, and demonstrating them in larger and larger reactors, as a prelude to industrial demonstration of the technology. However, research is now also focussing on the use of chemical-looping combustion for use with syngas, reaction with solid fuels and use in schemes to produce highly pure hydrogen. Key words: chemical-looping combustion, hydrogen production, gasification.

11.1

Introduction

The need to reduce anthropogenic greenhouse gas emissions represents a major driving force to reconsider the technologies being used for coal combustion and gasification processes.1,2 Current estimates of CO2 production from fossil fuels3 are of the order of 5.4 Gt/a, and it is clear that if fossil fuel use is to continue, then CO2 sequestration in deep geological formations must be employed as a final sink for the CO2 produced.4 Unfortunately, alternative approaches such as mineral carbonation of naturally occurring serpentine rocks are not particularly promising as they require very fine particle sizes (5 mm), and the reaction occurs in solution at high pressures (10–15 MPa).5,6 Natural weathering of carbonate rock can also potentially represent a sink as can carbonation of artificial materials containing CaO and/or MgO, such as steel slags,7 or fluidized bed combustion (FBC) solids which can 358 © Woodhead Publishing Limited, 2010

Chemical-looping combustion systems and technology

359

be carbonated fairly easily.8 Unfortunately, none of these is in sufficient abundance to make a significant impact on CO2 emissions. At present, most CO2 reduction R&D is directed to improving boiler efficiencies via supercritical cycles, along with the possible deployment of amine scrubbing;9 for gasification technology, it is primarily directed to reducing the cost and improving reliability. Zero-emissions technologies, which essentially allow a pure stream of CO2 to be produced from a combustion or gasification process employing hydrocarbons, are now of considerable interest,10 since the bulk of the costs (80 %) of CO2 avoided is for the CO2 capture and compression stages.9 Oxyfiring combustion represents a possible new direction, allowing more or less conventional boiler technology to be employed without the penalties introduced by amine scrubbing;11 however, oxygen production is itself expensive, and methods which avoid using pure oxygen, or minimize its use, almost certainly have a window of opportunity in the next several decades. For an overview of processes for CO2 separation and capture, the interested reader is directed to a number of general reviews.12,13 It is in this context that looping cycles represent an important new class of technologies, which can be deployed for direct combustion, as well as being used with the gases produced by gasification systems.14 Here the definition of a looping cycle is one that employs a solid carrier to bring oxygen to the fuel gas, or takes CO2 out of the combustion or gasification gases so that it can be subsequently released as a pure CO2 stream suitable for use or, more likely, sequestration. This chapter will not consider the use of cycles using liquids, although such systems are being developed and amine scrubbing itself could be regarded as an example of this type of looping cycle. The use of solids also means that, in many cases, fluidized bed systems will represent an optimal method of allowing large quantities of solids to be transferred from one chemical environment to another,15–17 with the added advantage that both large (> 350 MWe) atmospheric and pressurized systems also exist.18,19 This means that fluidized bed systems could be adapted for use in a number of possible looping cycle schemes.

11.2

Basic principles

Basic principles typically involve a metal oxide that is taken from a higher to a lower state of oxidation in a reaction with a gaseous hydrocarbon, following the global reaction scheme:

(2n + m) MyOx + CnH2m = (2n + m)MyOx–1 + mH2O + nCO2 [11.1]

After the water produced from the hydrocarbon is removed, the flue gas stream contains an effectively pure stream of CO2 suitable for sequestration. The metal oxide is then regenerated in a separate reactor by reaction with


360


air (Reaction 11.2), and the regenerated solid can be transferred back for further reaction with the fuel gas (Fig. 11.1):

MyOx–1 + air = MyOx

[11.2]

Typically, the oxidation reaction is strongly exothermic and the reduction step is not20 (an exception is the CuO/Cu cycle21) so that, overall, the system yields the heating value of the fuel. Such a cycle represents an elegant way of oxidizing a fuel gas by effectively achieving air separation, without using cryogenic or membrane technology and, in addition, avoiding or minimizing the formation of fuel–NOx. In practice, these cycles typically operate at temperatures in the range 800–1200 °C, which normally ensures that the reactions occur at a sufficient rate to be compatible with a fluidized bed system, and that agglomeration and sintering are minimized; in principle, the cycles can be run either at atmospheric or high pressure, although the vast bulk of the research has been done at atmospheric pressure. It should also be noted that Reaction 11.1 is idealized since many oxides permit multiple oxidation states, and it is not necessary that complete conversion of the oxide occurs in either Reaction 11.1 or Reaction 11.2 to ensure effective use. 20 Although these cycles are usually referred to as chemical-looping combustion (CLC),10 they are also a form of chemical-looping cycles. In consequence, systems in which the solid carries oxygen to the fuel gas, and systems in which CO2 is carried away from the flue gas, have a number of common features and problems, and much of the literature on either approach is of

CO2, H2O

O2-depleted air

MyOx–1

Fuel reactor

Air reactor M yO x

Fuel

Air

11.1 Typical metal oxide looping cycle.



361

general relevance to those interested in using solids to ensure the production of pure CO2 streams at high temperatures. The CLC-looping concept was patented by Lewis and Gilliland22 in 1954 as a method of producing pure CO2, and first came to the attention of a more general research community due to the work of Richter and Knoche23 in 1983, as a potential method of improving the efficiency of the combustion process. Early work using Ni-based carriers was carried out by Ishida and co-workers (1987, 1994),24,25 with experimental studies being conducted in a thermogravimetric analyzer (TGA). Subsequently, this work was extended to fixed-bed studies, again using Ni-based carriers, which again demonstrated that the reaction was fast enough to be employed in practice, and that soot formation or carbon deposition on the particles did not appear to be a major problem if the oxidation reaction was carried out at a high enough temperature.26 Subsequent work by Ishida and his colleagues using a TGA looked at improving the binding material (in this case NiAl2O3) and determined that, with a suitable H2O/CH4 ratio, carbon deposition could be avoided altogether.27 These workers also studied other possible systems, including CoO/Co and Fe2O3/Fe3O4. In practice, there are a very large number of potential or theoretical material choices for such cycles, but important considerations are the cost of such materials, and their oxygen carrying capacity (usually defined as the mass of oxygen carried per unit mass of the carrier in its oxidized state), with values ranging from 0.03 for the Fe2O3/Fe3O4 to 0.21 for the NiO/Ni and CoO/Co systems.14 Another issue that deserves mention is that there has been some success in preparing oxygen carriers with mixed oxide systems, which together show better overall performance. For example, the addition of small amounts of Ni was shown to improve the performance of various oxygen carriers and it is suggested that this arises because Nio can catalyze methane conversion by methane pyrolysis and steam reforming.28 It is also worth noting that higher carrying capacities are possible in principle; thus for instance we have explored the following cycle for gasification:

CH4 + CaSO4 = CaS + CO2 + 2H2O

[11.3]

With subsequent regeneration of the CaS (and while this system does have the problem of ensuring that the CaS is selectively oxidized to CaSO4, rather than CaO), its carrying capacity is exceptionally high29 at 0.47 and more will be said about this later. When considering the carrying capacity, it is important to remember that normally the active agent will be distributed on some kind of porous medium, such as porous alumina particles or ZrO2, TiO2 or MgO; so the oxygen carrying capacity should not be confused with the oxygen transport capacity of the actual particles, which can be significantly lower. Thus for instance, Abad et al. (2007a)30 give figures for three CLC oxygen carriers based on Cu, Fe


362


and Ni and the actual carrying capacities are 0.02, 0.013 and 0.084. Finally, something should be said about density of these particles, which can range31 from about 1.5 to 4.5 kg/m3. The higher densities are not typical of the types used in FBC combustion. This means that standard bubbling and circulating FBC boiler designs are unlikely to be easily adapted for large-scale units with the heavier oxygen carrying particles and there will be a number of design issues, including the effects of vibration, which must first be resolved if this technology is to be adapted for use with pressurized coal syngas. However, at this moment there seems little published on such matters and we will not consider these issues much further here, other than to note that Abad et al. (2007a)30 looked at the inventories and solid circulation rates that would be required for pressurized systems, and this represents an important first step in the type of considerations that need to be addressed for application with pressurized coal syngas. There is also another new study31 suggesting that pressurized conditions enhance the undesirable effect of carbon deposition, as do high concentrations of H2S. If so, this too will demand consideration in designing a pressurized system. At the beginning of the decade, significant work on this concept began outside of Japan, most notably in Sweden with the Lyngfelt group at Chalmers University, and the interested reader is recommended to look at the link describing the work being done by this group and their collaborators, which lists more than 60 publications on CLC.32 In a personal communication to the author, Lyngfelt reported that over 500 h of operation of a dual fluidized bed looping cycle had been achieved by 2007.33 Similarly, in his 2007 thesis Johansson notes that over 600 different oxygen carriers have been tested.34 Another very prolific source of publications in this area is the Spanish Research Council.35

11.3

Technologies and potential applications

In 2001, the Chalmers group20 reported the first experiments on the iron system using methane in a small fixed bed reactor and indicated that this could be used in a practical fluidized bed dual reactor system. Subsequent work continued as described in the extensive publication list,32 but arguably the most significant development to date was the construction of a small 10 kW chemical-looping combustor (Fig. 11.2) by Chalmers University, CSIC/ ICB, the Technical University of Vienna and Alstom, which first operated in August 2003.36 Initial results from the first 100 h of operation indicated that fuel conversion efficiencies were high (99.5 %), with only 0.5 % CO, 1 % H2 and 0.1 % methane in the gas exit stream, and no leaks were detected from the reactor. A very low rate of attrition was also determined from the bed particles (0.0023 %/h), which was translated into a cost of 1 7/t of CO2 capture for an anticipated particle lifetime of 4000 h, corresponding to a



363

Flue gas

Air reactor

CO2 + H2O

Fuel reactor

Fuel

Air

11.2 Schematic of Grace 10 kW chemical-looping reactor.

loss of 10 % of the bed mass. Subsequently, a second 10 kW unit was built by CSIC/ICB, and this has been run for a number of trials, most notably including a 200 h demonstration of the CuO/Cu cycle,37,21 For this system these workers suggested a practical lifetime for the Cu-impregnated particles might be 2400 h and, overall, the results indicated excellent performance. In the last several years there have been numerous studies done, in particular with the NiO/Ni and CuO/Cu systems, looking at parameters such as the stability of the particles produced, redistribution and/or loss of the active metal component in the synthetic particles, formation of soot or carbon deposits on the particles and loss of activity due to sintering.38,39 Other issues include possible interactions between particular types of substrate and the active metal, and their implications for fall-off in activity or introduction of mechanical weakness in the carrier particles.40 There is also work done on


364


the effect of gas impurities, on what is arguably the most well studied and important CLC system, namely Ni/NiO, in a 500 W system,41 where sulphur in particular might either form low melting point sulphides (such as Ni3S2), leading to sintering or agglomeration, or poison the Ni-based carrier. These workers noted no agglomeration even after 65 h of operation, and while the presence of H2S did degrade the performance of the oxygen carrier, the oxygen carrier recovered its activity when H2S feeding was stopped. These papers indicate that there are no major technical limitations to this concept. Caveats include: the fact that the CuO/Cu system must be operated at 850 °C or below because of the low melting point of the active components; and the iron system is somewhat less effective as an oxygen carrier than expected, due to formation of a variety of compounds. However, none of these types of issues is likely to provide major limitations if the concept is to be employed with natural gas with operation at typical FBC temperatures (800–1000 °C), depending on the carrier employed. In some cases these problems are being resolved; thus for instance Chuang et al. (2008)42 reports on the preparation and performance of a carrier made from CuO and Al2O3 as a support, which showed no signs of agglomeration at temperatures up to 900 °C. For a good overview of the R&D done up to 2006, the interested reader is referred to the review article by Johansson et al. (2006b).43 For more recent work, suitable sources of such data can be found here or are suggested in the final section of this chapter.

11.4

Advantages and limitations of chemical-looping combustion (CLC) for natural gas and syngas

In this context it is interesting to look at the conclusions of Tan et al. (2006)14 in their overview paper concerning this concept. Positively, they note that the concept is highly flexible and is potentially an attractive candidate for processing gaseous fuels, suggesting that it could be adapted for a range of options including industrial boilers, gas turbines and fuel cells. However, they caution that the technology is currently only suitable for gaseous fuels, and that the majority of studies have been done using methane, with the number of studies on syngas extremely limited (although such studies carried out to date suggest syngas reacts much more readily than natural gas, and it appears there is little effect of syngas composition on the reaction rate44). They also express concern about carryover of char carbon to the air reactor, if such a concept is integrated with gasification technology, and possible interactions between ash and the oxygen carrier, and called for test work with larger (1 MWth) pilot plants to validate the work done in smaller-scale units. Another very recent evaluation by Kvamsdal et al. (2007),45 who looked at nine concepts for natural gas firing, concluded that the chemical-looping concept is promising, but requires a long timeframe for realization. In



365

addition, they noted that the concept was rate-limited by the maximum temperature in the reactors. Nonetheless, they rated the technology as third in terms of net plant efficiency, and since the other two higher-rated concepts require a solid oxide fuel cell and gas turbine, and the second additionally requires O2 separation by means of a membrane, this must be regarded as a very positive evaluation for this technology if it is used with natural gas. Naqvi et al. (2007)46 developed a very positive evaluation for the use of this technology at part load, with CO2 capture. At 100 % load, an efficiency of 52 % (including CO2 capture) was predicted and a net plant efficiency drop of only 2.6 % was projected when reducing the load to 60 %. Interestingly, in a recent study workers from the University of Vienna have carried out a concept study for a 10 MWth CLC demonstration plant, based on work with a small 120 kW CLC plant, and estimated a net efficiency of 36.3 % using a NiO-based oxygen carrier.47 For coal, on the other hand, currently such a technology can only be used after a gasification step, although as discussed subsequently there is now active research on the use of CLC technology with solid fuels. In recent work, Abad et al. (2007b)48 have noted that, in atmospheric tests with iron oxide as the oxygen carrier, better combustion efficiencies were obtained with syngas than natural gas, although combustion efficiencies of 99 % could be obtained with natural gas if higher temperatures (950 °C) were employed. They also noted that no problem is expected from the sulphur component in the syngas stream with the iron system, although they did not include sulphur compounds in the synthetic syngas they employed. These results are quite reasonable since methane should be more difficult to combust than any other hydrocarbon. It may also be possible to partially resolve such problems by using other bed materials in conjunction with the iron-based carrier. Thus, for example Pröll et al. (2009), who examined the reactions of an iron-based ore, illmenite (FeTiO3), in a 120 kW dual fluidized bed combustor, noted that while the illmenite-based carrier performed quite well for a simulated syngas (mainly CO and H2), it did not do so for methane, but the replacement of part of the bed material with olivine improved the hydrocarbon conversion moderately.49 Almost certainly the syngas to be used in such a cycle will be produced via a high-pressure gasification process such as integrated gasification combined cycle (IGCC). While the benefits of a pressurized cycle in terms of improved efficiency and reduced work to produce a pressurized gas for pipeline transportation are well understood,29 there are very few studies on how well an O2-looping cycle would perform at high pressures. Further, no test data exist from pilot-scale units as yet. However, a pressurized thermogravimetric analyzer (PTGA) study carried out with pressures up to 3 MPa showed that increasing total pressure reduces the reaction rate of the oxygen carriers examined, and these workers speculated


366


that increasing pressure might affect the internal structure of the oxygen carrier.50 A significant conclusion of their work was that solid inventories might have to be larger than those initially anticipated for pressurized oxygen looping cycles. Siriwardane et al. (2007)51 have also recently examined the behaviour of a Ni-based carrier supported on bentonite, using a high-pressure flow reactor operating from 101 kPa to 690 kPa, and showed a rather more complicated picture for this system, with improved reaction at the highest pressure. These same workers52 have also examined a Cu-based carrier using bentonite as a support on a tapered-element oscillating microbalance, at pressures up to 0.69 MPa, and concluded that reaction pressures have a negative effect on reaction rates. None of these studies has looked at the effect of sulphur compounds on the performance of the oxygen carriers. While none of the conclusions of Tan et al. (2006)14 has as yet been invalidated, the overall results of more recent studies seem to be extremely supportive of CLC for natural gas firing with CO2 capture, whether or not pressurized cycles are shown to be practical. However, it is clear that much more study is needed on the use of this technology for conversion of synthesis gas if it is to be converted at high pressures typical of IGCC operation. In this respect, it would appear that, while CLC might have great potential for natural gas, there is a need for major development and a significant amount of research to confidently deploy it in conjunction with coal gasification processes; the use of these cycles with solid fuels is discussed in detail below.

11.5

Hydrogen manufacture using chemical-looping combustion (CLC)

This chapter would not be complete without some discussion of hydrogen manufacture using CLC. The direct production of H2 via the reaction of water with iron or iron oxide is already well known. Moreover, in principle the H2 produced can be of high purity (e.g., < 50 ppm of CO) for use in polymer electrolyte fuel cells, and the primary limitation of this approach is the fact that cheaper methods of H2 production are available.53 Yang et al. (2008a)53 carried out experiments in a small atmospheric fluidized bed using 3 g batches of haematite or Fe2O3, and demonstrated: successful reduction of the Fe2O3 with potassium carbonate-promoted chars; production of H2 via the reaction of water with the iron oxide; iron produced by reaction with char; and final oxidation of the resulting Fe3O4 to Fe2O3 by reaction with air. The final step returned the reagent to its initial state and since the final oxidation is exothermic this step would have the capacity to supply steam for H2 production. Steam by itself cannot oxidize Fe3O4 to Fe2O3 as even a very low H2 concentration prevents such a transition. Thus, for example, Bohn et al. (2008)54 note that H2 concentration of 20 ppm (pH2O/pH2 < 50 000) can



367

prevent the transition of Fe3O4 to Fe2O3 at a temperature of 750 °C. The appropriate series of reactions with CO is as follows: ∑ reduction of Fe2O3

3Fe2O3 + CO = 2Fe3O4 + CO2

ΔHo298K = –43.2 kJ/mol

[11.4]

0.947Fe3O4 + 0.788CO = 3Fe0.947O + 0.788CO2 ΔHo298K = + 37.3 kJ/mol [11.5]

Fe0.947O + CO = 0.947Fe + CO2 ΔHo298K= –16.7 kJ/mol

[11.6]

∑ oxidation of Fe by steam:

0.947Fe + H2O(g) = Fe0.947O + H2 ΔHo298K = –23.8 kJ/mol

[11.7]

3Fe0.947O + 0.788H2O(g) = 0.947Fe3O4 + 0.788H2 ΔHo298K = – 69.2 kJ/mol [11.8] ∑ re-oxidization of Fe3O4 to Fe2O3

2Fe3O4 + 1/2O2 = 3Fe2O3

ΔHo298K = –239.7 kJ/mol [11.9]

The fractional numbers in Reactions 11.5–11.8 are due to the fact that wuesite has a composition which can be written as Fe(1–y)O where 0.05 < y < 0.17, such that the stoichiometric ratio of Fe/O = 1 is never achieved and typically this composition is usually written as Fe0.947O. Evidently this scheme will work as well with char carbon, and the hydrogen purity is normally controlled by the purity of the water. It is also evident that this scheme is amenable to a packed bed design,54 but fluidized bed systems operating as chemical-looping reactors would be particularly suitable for any large-scale system.

11.5.1 Autothermal reforming of natural gas and light hydrocarbons In addition, to direct reaction of iron-based materials with water, synthesis gas generation is possible by CLC, where the oxide carrier causes autothermal reforming with methane, natural gas or light hydrocarbons. Successful tests have been reported by de Diego et al. (2009),55 where over 50 h of operation with a Ni-based oxygen carrier was reported in a 900 kWth CFB reactor, and 90 h in a 120 kW reactor has been reported by workers from the Technical University of Vienna.56 The Spanish workers reported no signs of agglomeration during the 50 h of operation in the larger reactor, and high (98 %) conversion of methane, with a NiOreacted/CH4 molar ratio being calculated as sufficient to meet the requirements for a full heat balance of the necessary reforming


368


step. One concern is the formation of elemental carbon or soot and, to date, no evidence has been found of its formation when applied to autothermal reforming by either the Spanish or Viennese groups. Other studies in this area are currently fairly limited. However, Go et al. (2009)57 have carried out a study on hydrogen production from twostep steam methane reforming in a fluidized bed reactor looking at the iron oxide system. Here TGA experiments have been used to determine the best conditions for a continuous process, and a design was proposed for a double-loop solid circulation system with fuel and steam streams. Chiesa et al. (2008)58 have also looked at a CLC method for producing hydrogen using CLC, but in their case they propose a three-reactor design. In this system iron oxides in the form of Fe2O3 react with methane in the fuel reactor to produce a gaseous stream composed of water and CO2 ready for sequestration. In the steam reactor, the reduced iron oxide reacts with steam to form Fe3O4 and hydrogen, while in the final stage the Fe3O4, with traces of unreacted iron oxide, is reconverted to Fe2O3 in a highly exothermic reaction which can then be used to drive the reaction in the first stage. The authors conclude that their technology offers comparable efficiency to conventional reforming plants, but better environmental performance, i.e., complete CO2 removal, instead of 80 % or less CO2 removal for conventional technology. Interestingly, Cleeton et al. (2009)59 have explored the possibility of using the iron system for hydrogen and electricity production from steam-gasified coal, and concluded that it is possible to produce either fully heat-integrated systems or ones in which only a small amount of heat must be added for optimal hydrogen yields, suggesting that there are no theoretical reasons to prevent such systems from being applied.

11.6

The use of chemical-looping combustion (CLC) technology with solid fuels

In order to make the largest possible impact, it would ideally be possible to use CLC with solid fuels, and active research is now underway to explore this. Thus, Leion et al. (2007)60 have recently reported some success in achieving gasification of petroleum coke at atmospheric pressure, using small batches of chemical-looping cycle oxygen carriers at 950 °C with a 60/40 % by weight composition of Fe2O3/MgAl2O4 pellet in a batch fluidized bed reactor with the Fe2O3 serving as the oxidizer. In this work they estimated that the upper bounds of oxygen carrier inventory would be 2000 kg/MWfuel. A particularly positive result from this work was that the pellets were used for 100 h of testing with no obvious degradation from the combustion gases or ash. This work has now been extended using pellets with 40/60 % by weight CuO/ZrO2 produced by freeze-granulation to combust petroleum coke.61 Here



369

an important difference in this work besides the different oxygen carrier is that the reactor was fluidized with CO2 and/or H2O, rather than a H2O and N2 mixture as previously used for the fuel reaction. In these experiments, conversion of 95 % of the fuel was achieved at temperatures above 940 °C in 40 seconds or less, with the rate increasing with increasing temperature. This is over an order of magnitude faster than for the iron system, and the authors have estimated that the inventory of particles ought to be of the order of 120–200 kg/MWfuel for particles composed of 40 % CuO. Other interesting results from this study include the absence of any signs of the formation of permanent agglomerates in the bed. Another recent study on solid fuels in a small bench-scale bubbling bed reactor was performed using a Chinese bituminous coal, and the Ni/NiO system (23 % active NiO, and 77 % inert NiAl2O4).62 Here, the Ni-based particles were periodically exposed to air and the solid fuel/steam mixture to simulate up to 10 reaction cycles. Typically, in these experiments 1.5 g of coal was introduced into a bed of 105 g of the NiO carrier. These workers report that the best results were obtained at higher temperatures (900 °C), but did notice a decrease in performance with increasing numbers of cycles; thus for instance, in one set of experiments they noticed CO2 production of over 95 % in the first cycle falling to just above 85 % by the tenth cycle. Shen et al. (2009)63 have recently reported experiments in which coal was converted using a Ni-based oxygen carrier in two interconnected fluidized beds. To maximize the performance of the fuel reactor they chose to use a spouted bed (with a cross-section of 230 ¥ 400 mm, and a height of 1500 mm), while the air reactor was a conventional fast bed. The air reactor operated over a temperature range of 950–1050 °C, while the fuel reactor operated in the range 850–950 °C. One claimed advantage of the Ni/NiO system was that it helped to suppress tar formation, and the spouted bed design was seen as permitting long residence times to allow indirect gasification of the coal, via the recycled fluidizing gases. The authors note that carbon conversion efficiency reached only 92.8 % even when the fuel reactor temperature was 970 °C, and explain this by inherent carbon loss due to elutriation of fine char particles from the freeboard of the spouted bed reactor. They also note that some char was transferred to the air reactor. CO2 levels at the highest temperatures were as high as 96 %, but the authors also show that CO levels were around 4 %, which would clearly be undesirable. Although this work is remarkable, a further criticism can be made that spouted fluidized beds do not scale-up particularly well, and Bridgewater (1985) reports that the largest spouted bed in the literature had a diameter of 2.5 m, and was about 12.5 m high.15 Therefore, it is clear that the concept needs to be tested in a system more amenable to scale-up for industrial systems. Berguerand and Lyngfelt (2008)64 have also reported tests of a total duration of 22 h (with 12 h of stable operation) on a South African bituminous coal


370


using an iron-based oxygen carrier in their 10 kWth unit. Here, the feed rate of fuel was about 0.5 kg/h and the fuel reactor was operated at a temperature of about 950 °C. The air reactor was run at a nominal 1000 °C, and the solid inventory was assumed to correspond to 500 kg/MW. The authors noted that fuel conversion was relatively poor (50–80 %), but that these must be regarded as early results, and that with proper system optimization for coal, conversions achieved in these experiments would fall in the range 82–96 %. The authors’ choice of the iron system for converting coal has recently been supported by a study carried out by Rubel et al. (2009),65 who used a thermal analyzer/differential scanning calorimeter/mass spectrometer to study oxygen carriers for their potential application for solid fuels, and came to the conclusion that iron-based oxygen carriers were particularly suitable, and strongly resistant to agglomeration. A somewhat similar conclusion, namely that FeTiO3 had good attrition resistance, and that this together with its low price suggested that it had good potential for use in CLC, was made by Berguerand and Lyngfelt (2008), based on 11 hours of steady-state testing using petroleum coke in a 10 kWth CLC unit.66 In another study using a 30 mm diameter fluidized batch reactor, Leion et al. (2008)67 have looked at FeTiO3 and synthetic Fe2O3 (60 %)/MgAl2O3 reagent. Interesting conclusions from this work are that the materials are very similar in performance, and that the use of these materials might compensate for the loss of CLC particles with the ash, due to their relatively low cost. Finally, these authors have suggested a variation of CLC, namely CL with oxygen uncoupling (CLOU), which might resolve the problem of the relatively slow gasification step that must occur to convert solids into fuel-rich gases, which can oxidize the CLC solid carriers. Here, the metal oxide actually decomposes to produce oxygen which is then available to react directly with the solids, eliminating the slow gasification step with H2O or CO. This can be represented as:

MexOy = MexOy–2 + O2(g)

[11.10]

In particular, they recommended three systems, CuO/Cu2O, Mn2O3/Mn3O4 and Co3O4/CoO, as being suitable for this approach, and they carried out limited numbers of batch tests in a small (22 mm dia.) quartz reactor with the copper system, which indicated that the reaction with petroleum coke was approximately 50 times faster than seen in previous work using steam as a fluidizing medium and CLC at 950 ºC, with a Fe2O3/MgAl2O3 carrier.60 The implications of these interesting studies for coal in continuously operating systems are far from clear, especially given the ubiquitous presence of relatively high levels of ash in coal with widely varying properties, and of course the small-scale and batch nature or limited durations of the experiments reported to date. Nonetheless, these results indicate that O2-looping cycles are still offering interesting new possibilities and, should the direct use of the



371

oxygen carrier in the presence of solid fuels finally prove unpromising, then the oxygen carrier can still be used to supply heat for the gasification step, and the cleaned syngas converted by the oxygen carrier. Such an approach has, for instance, been proposed by Xiang et al. (2008),68 who suggested the use of a pipe-type gasifier, which they simulated using ASPEN®. This approach is, in their view, likely to be more feasible than direct oxidation, because they expect the direct reaction will suffer from too low a reaction rate; the coal ash will adversely affect the surface of the metallic particle serving as the oxygen carrier; and that coal and char will be entrained, affecting overall performance.

11.7

The CaS–CaSO4 system

This system, as has been noted earlier, has the highest oxygen carrying capacity of the materials normally considered for CLC processes. It is also remarkable in a number of other ways. First, it is the only non-metallic system considered for such options, and CaSO4 itself is available as the mineral gypsum (CaSO4·2H2O) or as natural anhydrite. The use of this system was proposed by Alstom Ltd (2007), in a $4 million US DoE program aimed at converting solid fuels in a CLC CFB power plant to H2. However, the system considered by Alstom involved a number of looping cycles working together, namely: a CaSO4–CaS cycle; a CaCO3–CaO cycle for CO2 removal; and a thermal loop involving bauxite. Such a system would evidently be much more complicated even if it had potentially very attractive costs for avoided CO2 ($11–13/t of CO2).69 This concept was criticized by Wang and Anthony (2008)29 on the basis that it requires the direct reduction of CaSO4 by solid fuel, and hence separation of ash and unburned carbon from the CaS. Instead they proposed an indirect scheme (Fig. 11.3) in which a syngas is oxidized to CO2 by CaSO4.29 They also proposed the use of sulphated dolomite (CaSO4 . MgO) as a possible carrier with the suggestion that this might deal with reactivity and sintering problems. However, by far the most detailed recent work on this system has been done by researchers from Southeast University in China.70–73 Initial studies were carried out using a natural anhydrite ore (94.4 % pure), with an oxygen capacity of 0.444, which was tested in a 25 mm dia., fixed bed reactor. They found that high temperature (1000 °C) favoured high conversions of natural gas, but was also associated with the release of SO2 (as much as 6 % in the flue gases), via the reactions:

CaS + 3CaSO4 = 4CaO + 4SO2

[11.11]

CaS + 3/2O2 = CaO + SO2

[11.12]

and high temperature was also associated with some agglomeration.70 In mechanistic TGA studies, Shen et al. (2008)74 have concluded that the © Woodhead Publishing Limited, 2010

372


O2 depleted air

CO2

Fuel

Gasifier 900 °C 2C + 2CO2 = 4CO Ash

CO/CO2

Reduction reactor 900 °C

CaS

4CO + CaSO4 = 4CO2 + CaS

Fresh CaSO4

Oxidation reactor 1000 °C

CaSO4

CaS + 2O2 = CaSO4 Air

Heat

11.3 Schematic of the chemical-looping process for combustion of solid fuels.

primary reaction responsible for the loss of sulphur from the process is in fact Reaction 11.11. This work also concluded that the air reactor should be operated in the temperature range 1050–1150 °C, and the fuel reactor should be operated over the temperature range 900–950 °C. Song et al. (2008b)71 then extended their work using a 25 mm dia., bubbling fluidized bed with simulated syngas, which was examined over a temperature range of 850–950 °C. They noticed significant deactivation within 15–20 cycles, which they attributed to the effects of sintering and loss of sulphur from the system in the form of SO2 and H2S. These workers also confirmed, by means of X-ray diffraction, the presence of CaO in the materials obtained after 15–20 cycles and noted that surface area changed from an initial value of 0.5266 to 0.4426 m2/g by 20 cycles, with a corresponding increase in pore volume and average pore diameter, as one would expect due to sintering and possible particle cracking. These workers also developed a 2-D Fluent® model of the fuel reactor, which can be used as a diagnostic tool to study the process.73 Finally, it should be noted that there remains the possibility that enough sulphur will escape the system to render effective use of such a cycle impossible. A possible strategy is to add CaO to the system so as to capture escaping SO2, but such a solution needs to be demonstrated. All of these observations indicate that significantly more work will need to be done on this system if it is to be deployed commercially, starting with realistic pilot-scale trials of the sort that have been carried out with the other metal oxides tested in CLC research.



11.8

373

Future trends

Although CLC can not in any way be regarded as mature technology, since there is a major need to demonstrate it at the large pilot-scale and ultimately commercial level, interest is increasingly broadening to include hydrogen manufacture and gasification processes. Proposals such as that of Wang and Anthony (2008)29 to use, in their case, the CaS/CaSO4 system or, for instance, to develop syngas chemical looping to co-produce H2 and electricity by using the iron system (Fe, FeO, Fe2O3)75 are increasingly being offered. Typically, these approaches use a simulation tool such as ASPEN® to model the overall performance, sometimes backed with small pilot-scale experiments. The interest in such approaches indicates the promise of CLC to make major strides into these important developments, with the ultimate hope that they will move society into the hydrogen economy. A final comment that can be made on such systems can be drawn from a review by Hossain and de Lasa (2008).76 These workers note that oxygen carriers ought to have the following characteristics: 1. be stable under repeated oxidation/reduction cycles at high temperatures; 2. be fluidizable; 3. be resistant to agglomeration; 4. be mechanically resistant to the friction stress associated with high circulation of particles; 5. be environmentally benign; and 6. be economically feasible. For the first three criteria the data obtained to date suggest that there should be no significant problem in meeting these requirements. For criterion 4, the information to date is also supportive, although the true test will come when large pilot-scale or, better still, industrial-scale tests are carried out for long times (preferably 10 000s of hours). The fifth criterion is more problematic, especially for the more exotic systems such as, for instance, the Co systems, and it must be noted that Ni(II) compounds are well known to be carcinogenic, albeit one for which there is significant industrial experience in roasting and other operations. However, ultimately such issues can be managed. The issue of economic and technical feasibility, however, still remains to be demonstrated by industrial-scale demonstrations, and depends critically on issues such as the necessary size of such units to convert a unit mass of fuel (i.e., kg of carrier/MWfuel) and any possible issues with vibration of large-scale units for instance. Oxygen leakage into such systems may also be an issue, but this is one that, for instance, oxyfuel combustion also faces, and this may ultimately prove more problematic for oxyfuel applications if very severe limits are put on the gas to be sequestered, especially for


374


applications like enhanced oil recovery.77 However, more challenging is the direct use of solid fossil fuels, and determining the long-term effect of ash and micropollutants such as sulphur on promising oxygen carriers remains also to be resolved at the industrial-scale as does the issue of performance at high pressure for applications involving syngas, should the efforts to employ this technology with solid fuels such as coal be unsuccessful. Nonetheless, it is clear that this technology has made remarkable strides in only a few years with pilot-scale runs now exceeding 1000 h using a 10 kWth reactor at Chalmers University, Sweden.78

11.9


In a field as active as that of CLC, it is difficult to keep track of the many publications. As an example, a search for papers for 2008 produced over 55 journal publications in 22 different journals, and many more conference papers. The top publishers of papers on CLC were the journals Fuel, Energy and Fuels and the International Journal of Greenhouse Gas Control. Of the various conferences the International Conferences on Greenhouse Gas Technologies are probably among the more important for the presentation of new CLC results. However, there are also a number of websites of interest, most notably that of Chalmers University, who are one of the most important groups pioneering CLC developments,79 and the Spanish Research Council.80 Another group doing significant R&D in this area can be found at the Technical University of Vienna.81 Outside this list, significant work is being done by a number of other groups, including the Tokyo Institute of Technology in Japan and the Korea Institute of Technology, and many university groups. These include groups in the UK at the University of Cambridge, in Canada most notably at the University of Western Ontario and the University of British Columbia, and in the USA by the US DoE, and various universities such as Ohio State University. Finally, the IEA Greenhouse Gas Program is now producing a new website for a high-temperature solid looping cycle network, and this should become of increasing value to those interested in CLC technology and science.82

11.10 References 1. Herzog H J, What future for carbon capture and sequestration? Environ. Sci. Technol. 2001, 35, 148A. 2. Jaccard M, Sustainable Fossil Fuels: The Unusual Suspect in the Quest for Clean and Enduring Energy. Cambridge University Press, Cambridge, UK, 2007. 3. Liu Z and Zhao J, Contribution of carbonate rock weathering to the atmospheric CO2 sink. Environ. Geol. 1999, 39, 1053. 4. IPCC, IPCC Special Report on Carbon Dioxide Capture and Storage, Working



375

Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, de Coninck HC, Loos M and Meyer LA (eds). Cambridge University Press, Cambridge, UK, 2005. 5. Newall P S, Clarke S J, Scholes H, Clarke N R, King P A and Barley R W, CO2 Storage as Carbonate Minerals. Report No. PH3/17, IEA Greenhouse Gas R&D Program, Cheltenham, UK, 2007. 6. Jia L, Anthony E J, Lin W G, Ruan Y H and Gora D, Carbonation of magnesium silicate minerals – an experimental study. Cdn. J. Chem. Eng. 2004, 82, 1289. 7. Teir S, Eloneva S, Fogelholm C-J and Zevenhoven R, Dissolution of steelmaking slags in acetic acid for precipitated calcium carbonate production. Energy 2007, 32, 528. 8. Rao A, Anthony E J, Jia L and Macchi A, Carbonation of FBC ash by sonochemical treatment. Fuel, 2007, 86, 2603. 9. Rao A B and Rubin E S, A technical, economic and environmental assessment of amine-based CO2 capture technology for power plant greenhouse gas control. Environ. Sci. Technol. 2002, 36, 4467. 10. McKee B (chairman), Solutions for the 21st Century, Zero Emissions Technologies for Fossil Fuels: Technology Status Report. IEA Committee on Energy Research and Technology Working Party on Fossil Fuels, International Energy Agency, Paris, France, 2002. 11. Buhre B J P, Elliot L K, Sheng C D, Gupta G P, Wall T F, Oxyfuel combustion technology for coal-fired power generation. Prog. Energy Combust. Sci. 2005, 31, 283. 12. Pennline H W, Luebke D R, Jones K L, Myers C R, Morsi B I, Heintz J and Ilconich J B, Progress in carbon dioxide capture and separation research for gasification-based power generation point sources. Fuel Process. Technol. 2008, 89, 897. 13. Yang H, Zu Z, Fan M, Gupta R, Slimane R B, Bland A E and Wright I, Progress in carbon dioxide separation and capture: A review. J. Environ. Sci. 2008, 20, 14. 14. Tan R, Santos S and Spliethoff H, Chemical looping combustion for fossil fuel utilisation with carbon sequestration. Technology Review, IFRF Doc. No. G 23/6/2, International Flame Research Foundation, Livorno, Italy, 2006. 15. Davidson J F, Clift R and Harrison D (eds), Fluidization, 2nd edn. Academic Press, London, UK, 1985. 16. Pröll T, Kolbitsch P, Bolhàr-Nordenkampf J and Hofbauer H, A dual circulating fluidized bed (DCFB) system for chemical looping processses. Proceedings of the 2008 AIChE Annual Meeting, Philadelphia, PA, 16–21 November, 2008. 17. Kolbitsch P, Bolhàr-Nordenkampf J, Pröll T and Hofbauer H, Comparison of two Ni-based oxygen carriers for chemical looping combustion of natural gas in 140 kW continuous looping cycle. Ind. Eng. Chem. Res. 2009, 48, 5542. 18. Grace J R, Avidan A and Knowlton T M (eds). Circulating Fluidized Beds. Blackie Academic & Professional, London, UK, 1997. 19. Cuenca M A and Anthony E J (eds), Pressurized Fluidized Beds. Blackie Academic & Professional, London, UK, 1995. 20. Mattisson T, Lyngfelt A and Cho P, The use of iron oxide as an oxygen carrier in chemical looping combustion of methane with inherent separation of CO2. Fuel 2001, 80, 1953. 21. De Diego L F, Garcia-Labiano F, Gayán P, Celaya J, Palacios J M and Adánez J, Operation of a 10 kWth chemical looping combustor during 200 h with a CuO–Al2O3 oxygen carrier. Fuel 2007, 86, 1036.


376


22. Lewis W and Gilliland E R, Production of pure carbon dioxide. U.S. Patent 2,665,972, 1954. 23. Richter H J and Knoche K, Reversibility of combustion processes. ACS Symp. Series 1983, 235, 71. 24. Ishida M, Zheng D and Akehata T, Evaluation of a chemical-looping-combustion power-generation system by graphic exergy analysis. Energy 1987, 12, 147. 25. Ishida M and Jin H, A novel combustor based on chemical looping reactions and its reaction kinetics. J. Chem. Eng. Jpn 1994, 27, 296. 26. Hatanaka T, Matsuda S and Hatano H, A new concept gas–solid combustion system ‘merit’ for high combustion efficiency, and low emissions. Proceedings of the 23rd Energy Conversion Engineering Conference, 2, 994–948, 1997. 27. Jin H, Okamoto T and Ishida M, Development of a novel chemical-looping combustor: synthesis of a solid looping material of NiO/NiAl2O3. Ind. Eng. Chem. Res. 1999, 38, 126. 28. Johansson M, Mattisson T and Lyngfelt A, Creating a synergy effect by using mixed oxides of iron- and nickel oxides in the combustion of methane in a chemical-looping combustor reactor. Energy Fuels 2006a, 20, 2399. 29. Wang J and Anthony E J, A process for clean combustion of solid fuels. Appl. Energy 2008, 85, 73. 30. Abad A, Adánez J, García-Labiano F, de Diego L F, Gayán P, Celaya, J, Mapping of the range of operational conditions for Cu-, Fe-, and Ni-based oxygen carriers in chemical-looping combustion. Chem. Eng. Sci. 2007, 62, 533. 31. Wang B, Yan R, Lee D H, Liang D T, Zheng Y, Zhao H and Zheng C, Thermodynamic investigation of carbon deposition and sulfur evolution in chemical looping combustion with syngas. Energy Fuels 2008, 22, 1012. 32. Lyngfelt A, Chalmers University, Sweden. Private Communication: Link to Group’s Publications (http://www.entek.chalmers.se/~anly/co2/co2publ.htm), April 2007. 33. Lyngfelt A, Private Communication. Chalmers University, Sweden, June 2007. 34. Johansson M, Screening of Oxygen-Carrier Particles based on Iron, Manganese, Copper and Nickel Oxides for use in Chemical-Looping Technologies, PhD Thesis, Chalmers University, Gothenburg, Sweden, 2007. 35. CSIC/ICB website: http://www.icb.csic.es/index.php?id=146&L=1. 36. Lyngfelt A, Kronberger B, Adanez J, Morin J-X and Hurst P, The Grace project development of oxygen carrier particles for chemical-looping combustion, design and operation of a 10 kW chemical looping combustor. in Rubin E S, Keith D W and Gilboy C F (eds) , Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 1, 115–123, 2005. 37. Adánez J, Gayán P, Celaya J, de Diego F F, García-Labiano F and Abad A, Chemical looping combustion in a 10 kWth prototype using a CuO/Al2O3 oxygen carrier: effect of operating conditions on methane combustion. Ind. Eng. Chem. Res. 2006, 45, 6075. 38. Corbella B M, de Diego L, Garcia F, Adánez J and Palacios J M, The performance in a fixed bed reactor of copper-based oxides on titania as oxygen carriers for chemical looping combustion of methane. Energy Fuels 2005, 19, 433. 39. Corbella B M and Palacios J M, Titania-supported iron oxide as oxygen carrier for chemical looping of methane. Fuel 2007, 86, 113. 40. Corbella B M, de Diego L F, Garciá-Labiano F, Adánez J and Palacios J M, Performance in a fixed bed reactor of titania-supported nickel oxide as oxygen



377

carriers for the chemical looping combustion of methane in multicycle tests. Ind. Eng. Chem. Res. 2006, 45, 157. 41. Adánez J, García-Labiano F, Gayán P, de Diego L F Abad A, Dueso C and Forero C R, Effect of gas impurities on the behavior of Ni-based oxygen carriers on chemical looping combustion, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 11–18, 2009. 42. Chuang S Y, Dennis J S, Hayhurst A N and Scott S A, Development and performance of Cu-based oxygen carriers for chemical-looping combustion. Combust. Flame 2008, 154, 109. 43. Johansson M, Mattisson T, Rydén M and Lyngfelt A, Carbon capture via chemicallooping combustion and reforming. International Seminar on Carbon Sequestration and Climate Change, Rio de Janeiro, Brazil, 24–27 October, 2006. 44. Dueso C, García-Labiano F, Adánez J, de Diego L F, Gayán P and Abad A, Syngas combustion in a chemical looping combustion system using an impregnated Ni-based oxygen carrier. Fuel 2009, 88, 2357. 45. Kvamsdal H M, Jordal K and Bolland O, A quantitative comparison of gas turbine cycles with CO2 capture. Energy 2007, 32, 10. 46. Naqvi R, Wolf J and Bolland O, Part load analysis of a chemical looping cycle combustion (CLC) combined cycle with CO2 capture. Energy 2007, 32, 360. 47. Pröll T, Private Communication. Technical University of Vienna, June 2009. 48. Abad A, Mattisson T, Lyngfelt A and Johansson M, The use of iron oxide as oxygen carrier in a chemical looping reactor. Fuel 2007, 86, 1021. 49. Pröll T, Mayer K, Bolhàr-Nordenkampf J, Kolbitsch P, Mattisson T, Lyngfelt A and Hofbauer H, Natural minerals as oxygen carriers for chemical looping combustion in a dual circulating fluidized bed system, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 27–34, 2009. 50. García-Labiano F, Adánez J, de Diego L F, Gayán P and Abad A, Effect of pressure on the behavior of copper-, iron-, and nickel-based oxygen carriers for chemical looping combustion. Energy Fuels 2006, 20, 26. 51. Siriwardane R, Poston J, Chaudhari K, Zinn A, Simonyi T and Robinson C, Chemicallooping combustion of simulated synthesis gas using nickel oxide oxygen carrier supported on bentonite. Energy Fuels 2007, 21, 1582. 52. Tian H, Chaudhari K, Simonyi T, Poston J, Liu T, Sanders T, Veser G and Siriwardane R, Chemical looping combustion of coal derived synthesis gas over copper oxide carriers. Energy Fuels 2008, 22, 3744. 53. Yang J, Cai N and Li Z, Hydrogen production from the steam-iron process with direct reduction of iron oxide by chemical looping combustion of coal char. Energy Fuels 2008a, 22, 2570. 54. Bohn C D, Müller C R, Cleeton J P, Hayhurst A N, Davidson J F, Scott S A and Dennis J S, Production of very pure hydrogen with simultaneous capture of carbon dioxide using the redox reactions of iron oxides in packed beds. Ind. Eng. Chem. Res. 2008, 47, 7623. 55. de Diego L F, Ortiz M, García-Labiano F, Adánez J, Abad A and Gayán P, Synthesis gas generation by chemical looping reforming using a Ni-based oxygen carrier, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 3–10, 2009. © Woodhead Publishing Limited, 2010

378


56. Bolhàr-Nordenkampf J, Pröll T, Kolbitsch P and Hofbauer H, Performance of a NiO-based oxygen carrier for chemical looping combustion and reforming in a 120 kW unit, in Gale J, Herzog H and Braitsch J (eds) Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 19–25, 2009. 57. Go K S, Son S R, Kim S D, Kang K S and Park C S Hydrogen production from two-step steam methane reforming in a fluidized bed reactor. Int. J. Hydrogen Energy 2009, 34, 1301. 58. Chiesa P, Lozza G, Malandrino A, Romano M and Piccolo V, Three-reactor chemical looping process for hydrogen production. Int. J. Hydrogen Energy 2008, 33, 2233. 59. Cleeton J P E, Bohn C D, Müller C R, Dennis J S and Scott S A, Clean hydrogen production and electricity from coal via chemical looping: identifying a suitable operation regime. Int. J. Hydrogen Energy 2009, 34, 1. 60. Leion H, Mattisson T and Lyngfelt A, The use of petroleum coke as fuel in chemical looping combustion. Fuel 2007, 86, 1947. 61. Mattisson T, Leion H and Lyngfelt A, Chemical-looping with oxygen uncoupling using CuO/ZrO2 with petroleum coke. Fuel 2008, 88, 683. 62. Gao Z, Shen L, Zio J, Qing C and Song Q, Use of coal as fuel for chemical looping combustion with ni based oxygen carrier. Ind. Eng., Chem. Res. 2008, 47, 9279. 63. Shen L, Wu J and Xiao J, Experiments on chemical looping combustion of coal with a NiO based oxygen carrier. Combust. Flame 2009, 156, 721. 64. Berguerand N and Lyngfelt A, Design and operation of a 10 kWth chemical-looping combustor for solid fuels – testing with South African coal. Fuel 2008, 87, 2713. 65. Rubel A, Liu K, Neathery J and Taulbee D, Oxygen carriers for chemical looping combustion of solid fuels. Fuel 2009, 88, 876. 66. Berguerand N and Lyngfelt A, The use of petroleum coke as fuel in a 10 kwth chemical looping combustor. Int. J. Greenhouse Gas Control 2008, 2, 169. 67. Leion H, Mattisson T and Lyngfelt A, Solid fuels in chemical looping combustion. Int. J. Greenhouse Gas Control 2008, 2, 180. 68. Xiang W, Wang S and Di T, Investigation of gasification chemical looping combustion combined cycle performance. Energy Fuels 2008, 22, 961. 69. Andrus H, Chemical looping combustion: R&D efforts of Alstom. Second Workshop, Oxy-Combustion Research Network, Windsor, CT, 25–27 January, 2007. 70. Song Q, Ziao R, Deng Z, Zhang H, Shen L, Xia J and Zhang M, Chemical looping combustion of methane with CaSO4 oxygen carrier in a fixed bed. Energy Convers. Manage. 2008, 49, 3178. 71. Song Q, Ziao R, Deng Z, Shen L, Ziao J and Zhang M, Multicycle study on chemical looping combustion of simulated coal gas with a CaSO4 oxygen carrier in a fluidized bed reactor. Energy Fuels 2008, 22, 3661. 72. Song Q, Ziao R, Deng Z, Shen L, Ziao J and Zhang M, Effect of temperature on reduction of CaSO4 oxygen carrier in chemical looping combustion of simulated coal gas in a fluidized bed reactor. Ind. Eng. Chem. Res. 2008, 47, 8148. 73. Deng Z, Xiao R, Jin B, Song Q and Huang H, Multiphase CFD modeling for a chemical looping combustion process (fuel reactor). Chem. Eng. Technol. 2008, 31, 1754. 74. Shen L, Zheng M, Ziao J and Ziao R, A mechanistic investigation of a calcium based oxygen carrier for chemical looping combustion. Combust. Flame 2008, 15, 489.



379

75. Fan L, Li F and Ramjumar S, Utilization of chemical looping strategy in coal gasification processes. Particuology 2008, 6, 131. 76. Hossain M and de Lasa H I, Chemical-looping combustion (CLC) for inherent CO2 separations – a review. Chem. Eng. Sci. 2008, 63, 4433. 77. Anheden M, Andersson A, Bernstone C, Eriksson S, Yan J, Lijemark S and Wall C, CO2 quality requirements for a system with CO2 capture, transport and storage, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 2, 2559–2564, 2005. 78. Linderholm C, Mattisson T and Lyngfelt A, Long-term integrity testing of spraydried particles in a 10-kW chemical-looping combustor using natural gas as fuel, Fuel 2009, 88, 2083–2096. 79. Chalmers University, http://www.entek.chalmers.se/~anly/co2/co2publ.htm. 80. CSIC, http://www.icb.csic.es/?148&L=1. 81. TUV, www.chemical-looping.at. 82. IEA, http://www.ieagreen.org.uk/


12

Gas purification, compression and liquefaction processes and technology for carbon dioxide (CO2) transport

A. A s p e l u n d, The Norwegian University of Science and Technology, Norway Abstract: This chapter describes the gas purification and conditioning of large-scale CO2 for CCS for both pipeline and ship transport. A review of the existing literature is conducted and the optimal transport conditions and specifications are discussed. Then, six transport processes are described from a set of basic building blocks, followed by a sensitivity analysis of feed gas pressure, ambient temperature and feed gas composition. It is shown that the interface between the capture and the transport processes can be used to design more energy- and cost-efficient systems and that the transport processes must be designed on a case to case basis. Key words: CO2 purification, CO2 transport, transport, pipeline, ship, impurities, CO2 conditioning.

12.1

Introduction

Carbon dioxide (CO2) capture and storage (CCS) is usually considered as a chain consisting of three separate elements; capture, transport and storage of the CO2. However, these processes are connected with the transport processes being the link between the capture process and the final storage. Therefore, the transport process configuration will, in addition to transport specifications, depend on both the capture process and the specifications from the reservoir. Also, since the elements are connected, there are interfaces which can be exploited to reduce the overall costs for the complete chain. The ambient temperature and the availability of cooling water may in some cases dictate the selected transport process. As always, the costs of electricity, the penalty for CO2 emissions and the investment costs for equipment will be of importance when designing the optimal transport process. The challenging task is to select the appropriate specifications for the capture plant, for the gas condition and transport process, as well as for the injection process and the final storage, and then design the most efficient transport process from a set of different possibilities. Choosing a gas specification that covers all possible situations will lead to unnecessarily strict requirements and thereby increase the costs. To reduce the costs, all parts of the chain must set minimum requirements and understand the impact on the other elements of imposing these restrictions. A cost saving 383 © Woodhead Publishing Limited, 2010

384


in e.g. the capture plant may increase the costs in the transport plant and vice versa. Similarly, if the specifications for the CO2 to be injected at the reservoir are too strict, this may increase the costs in the transport processes or capture plant significantly. Large-scale pipeline transport of CO2 is not a new technology, and CO2 is captured, transported in onshore pipelines and used for enhanced oil recovery (EOR) purposes in both the USA and Canada. Also, CO2 is separated from natural gas, transported in subsea or onshore pipelines and injected for final storage both at the Snøhvit LNG plant, at Sleipner and at In Salah (Michael et al., 2009). Vandenhengel and Miyagishima (1993), Skovholt (1993), Hendriks et al. (2002) and Heggum et al. (2005) have presented papers giving an overview of capture, conditioning and pipeline transport technologies, including costs and quality specifications. Food-grade CO2 is currently transported by small ships (1000 m3) at 15–18 bar (Hegerland et al., 2005). Large-scale ship-based transport of CO2 is relatively new, and several contributions were presented at the 7th international conference on greenhouse gas technologies in Vancouver in 2004 (Aspelund et al., 2005a, b; Barrio et al., 2005; Berger et al., 2005; Ozaki et al., 2005; Hegerland et al., 2005). Svensson et al. (2003) describe technical solutions and costs of ship transport of CO2. The complete ship based transport chain for CO2 is presented by Aspelund et al. (2006) in a paper that describes the technical solutions and analysis of costs, energy utilization, exergy efficiency and CO2 emissions. There is an increasing awareness that the removal of liquids (other than water) and volatile gases with a lower boiling point than CO2 must be considered in the CCS chain. In particular, gas conditioning of CO2-rich gases emerging from oxyfuel combustion of coal has emerged in recent years (Oryshchyn et al., 2007; White et al., 2007, 2009; Darde et al., 2009; Sass et al., 2009; Zanganeh et al., 2009). Li et al. (2008) published a paper that gives a theoretical view of impurity impacts on the purification process in oxyfuel combustion-based CO2 capture and storage system. Aspelund and Jordal (2007) discuss the interface between 11 different natural gas-based power plants with CO2 capture, as presented by Kvamsdal et al. (2007), and three different transport processes for both ships and pipelines. There are currently no specifications for gas transport; however, De Visser et al. (2008) recently presented some CO2 quality recommendations for pipeline transport. This work is an attempt to give an overview of the solutions for largescale gas conditioning and purification of CO2 from power plants with CO2 capture based on existing literature. The work builds to a large extent on the results from Aspelund et al. (2006) and Aspelund and Jordal (2007). The first section contains a brief discussion about the selection of transport pressure for both gas carriers and pipelines, followed by some CO2 quality


Gas purification, compression and liquefaction processes

385

recommendations. Then an overview of the gas conditioning and purification processes for large-scale CO2 transport, including a description of the basic building blocks that are available, is given. A superstructure combining these building blocks to form three pipeline processes as well as three ship processes is presented, including simple process flow diagrams of two selected processes. Next, sensitivity analysis for ambient temperature, inlet pressure and the amount of volatiles in the feed gas is presented. The succeeding part includes a discussion about the interface between capture and transport. A brief description of ship to pipeline and pipeline to ship processes is included, followed by discussions, further work, trends and, finally, conclusions.

12.2

Selection of transport pressures

In order to efficiently transport large amounts of CO2, it must be transformed into a form with high density, meaning that transport in liquid, solid or in supercritical phase may be considered. Figure 12.1 shows the CO2 density as a function of transport pressure for both pipeline and ship transport. In pipeline transport, CO2 will be transported at supercritical pressure in the range 80–150 bar and the CO2 must be dry to avoid corrosion. Impurities such as nitrogen and oxygen will not cause any harm to the pipeline; however, it will be cost-efficient to remove most of these components. The CO2 must be compressed to a pressure high enough to overcome the frictional and static pressure drops. Furthermore, the CO2 should be delivered at a pressure higher than the critical pressure to avoid two-phase flow and resulting liquid slugs 1600

1600 Solid, T = – 78 Fully refrigerated ship

1400

Density [kg/m3]

1200

TP

1200

Semi-pressurized ship

1000

Liquid Saturation line

600 400 –30 °C

0 °C

200

25 °C

itica ercr Sup CP

30.5 °C

1000

Pipeline

Fully pressurized ship

800

0

1400

800

l

600

40 °C

400 200

Vapour 0

20

40

60 80 Pressure [bar]

100

120

140

0

12.1 CO2 density as a function of pressure (adapted from Aspelund et al., 2006).


386


in the pipeline, as well as operational problems for the injection compressor (pump). In semi-pressurized ships, the CO2 to be transported is kept in the liquid phase on the saturation line by a pressure higher than the atmospheric pressure and a temperature lower than the ambient temperature. CO2 exists in liquid form at pressures between 5.2 bar and –56.5 °C at the triple point (TP) and 73 bar and 30.9 °C at the critical point (CP). Aspelund et al. (2006) and Hegerland et al. (2005) claim that CO2 is most efficiently transported at 6.5 bar and –51.2 °C, due to the higher specific weight, as well as a reduction in the costs for the vessels both at the intermediate storage and the gas carrier. The 1.3 bar margin from the TP should be enough to avoid operational problems. Furthermore, it should be noted that the energy requirements for liquefaction of CO2 at 6.5 bar is somewhat higher than for liquefaction at 15–18 bar.

12.3

Carbon dioxide (CO2) quality recommendations for transport in pipelines and by ship

The gas quality (temperature, pressure, composition) from the power plant or industrial source will vary to a large extent, depending both on the selected capture process and the feedstock, being coal, natural gas or biomass. Kvamsdal et al. (2007) have presented an overview of the CO2 specifications from various power plants with CO2 capture based on natural gas. On a dry basis, the gas quality ranges from almost pure (amine absorption) to less than 95 % for some oxyfuel processes. Coal or biomass power plants will in general have a higher concentration of impurities, as the coal may contain sulphur that will react to H2S in the combustion chamber. Also, CO2 from coal-fired power plants is expected to have a higher concentration of volatiles such as nitrogen, and gas qualities down to 75 % CO2 purity are reported in the literature (Li et al., 2008). In pipeline transport, the main considerations are corrosion and hydrate formation, which will cause operational challenges if free water is present in the CO2. Furthermore, there will be additional costs for transporting gases other than CO2, due to the increase in compressor power and higher investment costs for the pipeline. Also, since there might be leakage from a pipeline, the content of H2S and CO may have to be below 200 ppm and 2000 ppm, respectively, due to health and safety considerations (De Visser et al., 2008). These amounts will be below the threshold for the short-term exposure limits (STEL) that gives the maximum amount of a compound that one can be exposed to for 15 minutes without adverse health effects. Ship transport will have stricter specifications than pipeline transport. The water content should be low due to possible freeze-out of water during the liquefaction process. The main difference is that liquid CO2 (LCO2)



387

cannot contain more than roughly 0.3 mole% of volatiles such as nitrogen, as a higher content will increase the liquefaction costs and decrease the temperature for the same pressure, thereby increasing the risk for dry-ice formation. Since the LCO2 will be transported in type C tanks (independent tanks with no secondary barrier), there is a very little chance for leakages; hence the H2S and CO specifications given for pipeline transport may possibly be increased. The type of reservoir (e.g. depleted oil reservoir, saline aquifer) may add restrictions to gas quality as well as temperature and pressure, and it is expected that there will be stricter requirements if the CO2 is to be used for EOR (De Visser et al., 2008). The reason is that immiscible components may increase the miscible pressure in the reservoir and thereby decrease the efficiency of the CO2. Furthermore, oxygen may cause precipitation reactions and thereby reduce the permeability of the reservoir. Also, oxygen reacting exothermally with oil may lead to overheating at the injection point. As a consequence, specifications of 300 ppm for nitrogen and 50 ppm for oxygen may be required (Steenevelt et al., 2006). An important aspect of this is that oxygen cannot easily be removed to such stringent specifications using conventional separation without increasing the reboiler and condenser duty significantly. Hence, the complexity and costs for conditioning of all oxyfuel concepts will increase. Knauss et al. (2005) published a paper presenting the evaluation of the impact of CO2, co-contaminant gas, aqueous fluid and reservoir rock interactions on the geologic sequestration of CO2. They investigate the potential for co-injecting SOX, NOX and H2S from a coalfired power plant with CO2 capture along with the CO2 to possibly reduce the overall costs. It is therefore of vital importance that the specifications are not determined only from a conservative reservoir point of view, but also take into account the additional costs associated with demanding stringent specifications. Finally, the gas composition must be within the rules and regulations set by governments and organizations; however, such regulations are not yet developed. Tables 12.1 and 12.2 on page 388 show the recommended CO2 quality for pipeline and ship transport, respectively, based on the work of De Visser et al. (2008).

12.4

Overview and basic building blocks in carbon dioxide (CO2) transport processes

An overview of the transport processes is presented in Fig. 12.2. The CO2 from the power plant or industrial source is normally saturated with water and will contain some nitrogen. The first step for all transport processes is to cool the CO2 to ambient conditions and remove the water in gravity-based separators. The CO2 is then compressed to 20–30 bar in two stages with intermediate


388


Table 12.1 CO2 quality recommendations for pipeline transport Component Concentration

Limitation

Reason

H2O 500 ppm Design and operational considerations H2S 200 ppm Health and safety considerations CO 2000 ppm Health and safety considerations CH4 Aquifer < 4 vol%, As proposed in ENCAP project EOR < 2 vol.% < 4 vol.% (all non- As proposed in ENCAP project N 2 condensable gases) O 2 Unknown Literature not consistent Ar < 4 vol.% (all non- As proposed in ENCAP project condensable gases) H 2 < 4 vol.% (all non- Reduction of H2 is recommended condensable gases) due to its energy content CO2 > 95.5 % Balanced with other compounds

Corrosion, hydrates STEL STEL Costs, energy Costs Challenges in the reservoir Costs Costs, energy Economy

Table 12.2 CO2 quality recommendations for ship transport Component

Concentration

Limitation

H2O 50 ppm Design and operational considerations H2S 200 ppm Health and safety considerations CO 2000 ppm Health and safety considerations CH4 < 0.3 vol.% (all non- Design and operational condensable gases) considerations N2 < 0.3 vol.% (all non- Design and operational condensable gases) considerations O 2 Unknown Literature not consistent Ar < 0.3 vol.% (all non- Design and operational condensable gases) considerations H 2 < 0.3 vol.% (all non- Design and operational condensable gases) considerations CO2 > 99.7% Balanced with other compounds

Reason Freeze-out in heat exchangers STEL STEL Dry ice formation, costs for liquefaction Dry ice formation, costs for liquefaction Challenges in the reservoir Dry ice formation, costs for liquefaction Dry ice formation, costs for liquefaction Dry ice formation

cooling. Most of the water will be removed; however, it is still necessary to remove the remaining water using regenerative adsorption columns. If the gas is to be transported in a pipeline, there are three main choices. The CO2 can be compressed directly to the transport pressure (P3); however, it is then not possible to remove volatiles such as nitrogen and oxygen. The


Pipeline

(P3)

(P)

Recompression

(P1)


Water removal Water removal by separation by adsorption

Condensation Removal of volatiles

Removal of SO2 and H2S

Water

Expansion (liquefaction)

CO2 flash gas

Volatiles and CO2 CO2-rich gas

Power plant with CO2 capture Purge

Coal

Gas

Bio

12.2 Transport processes in a CCS chain.

(P)

Pumping

Final storage (S) (S)

LCO2 storage (S)

Ship

Pumping (S) and heating


Compression and cooling

389

390


alternatives are to compress the CO2 to 60–65 bar and condense it, preferably by seawater (P1), alternatively by a refrigeration unit (P2). The volatiles can then be removed in a distillation column or a flash tank if very small amounts of volatiles are present. The liquid, purified CO2 is then pumped to injection pressure and sent to the pipeline. For long pipelines, or if the reservoir has a high pressure, recompression is required. To avoid operational problems, the pressure should be maintained above the critical point pressure (alternatively the cricondenbar pressure for CO2 rich mixtures). This is beneficial to avoid two-phase flow in the pipeline and operational challenges in the injection compressor. If the CO2 is to be transported by ship, several options are available. As for (P2), it can be compressed to 60–65 bar and condensed by seawater, before most of the volatiles are removed in a distillation column. The CO2 is then expanded to transport pressure, the liquid CO2 is sent to the storage tank, whereas the flash gas is sent to the compressors for recompression (S1). Alternatively, it can be liquefied by an ammonia refrigeration unit at 20–30 bar, before the volatiles are removed and the liquid CO2 is expanded (S2). Finally, the CO2 can be compressed to 100 bar and cooled in the dense phase by ambient cooling water or ambient air (S3). In order to transport CO2 at 6.5 bar, the CO2 can only contain 0.2–0.3 % of volatiles to ensure that dry ice is not formed. The CO2 is then loaded onto the ship and transported to the injection site, where it is pumped and heated to injection specifications. As can be seen from Fig. 12.2, the compression and liquefaction processes have many similarities, both in process equipment, minimum energy requirement and costs. The composition of the CO2 will not change during the transport or in the offshore installation process, provided that the CO2 chains are designed without leakage. Hence, the CO2 specifications have to be met by the CO2 capture and conditioning process. The building blocks for the transport processes are: ∑ compression and cooling; ∑ removal of water and other liquids in vapour–liquid separator drums; ∑ removal of water by adsorption; ∑ removal of unwanted components by chemical or physical treatment or advanced separation processes; ∑ condensation; ∑ pumping; ∑ expansion to transport pressure (liquefaction); ∑ removal of volatile gases by distillation.

12.4.1 Compression and cooling Centrifugal compressors are the natural choice for compression of large amounts of CO2. The compressors will have a polytropic efficiency of 80–85 %. One © Woodhead Publishing Limited, 2010


391

casing (stage) can increase the pressure ratio up to five fold, e.g. from 1 to 5 bar; however, 3–4 is more energy-efficient. Two casings can be included in one compressor. Intermediate cooling can be done by seawater, freshwater or ambient air heat exchangers. The first two stages must be designed to withstand the corrosion that is expected due to the acid water. It is possible to use the heat from the compressor coolers to pre-heat the feed water to the heat recovery and steam generator (HTSG) in the power process; however, the increased cost and possible reduction in operability will not necessarily justify the small increase in efficiency and power production.

12.4.2 Removal of water and other liquids in the vapour–liquid separator drums Vapour–liquid separator drums are needed to ensure no liquid entrainment to the CO2 compressors. Separation by gravity using liquid–vapour separator drums is the simplest and most cost- and energy-effective way to remove the bulk of components with higher density than gaseous CO2. Components with high solubility in water or components with significantly higher boiling points than CO2 will be removed together with the water in the separator drums. The solubility of water in CO2 gas decreases with higher pressure and lower temperatures (Diamond and Akinfiev, 2003). Most of the water is removed in the first vapour–liquid separator drums before compression and cooling. As the gas is compressed and cooled, some additional water condenses and is removed in the separator drums prior to the compressor. The last free water should be removed at a pressure between 20 and 30 bar and at a temperature as close to the hydrate formation curve as possible. With proper design and cold seawater, the vapour–liquid separator drums can remove water down to approximately 400–500 ppm (Austegard et al., 2006). Therefore, by using the specifications given by De Visser et al. (2008) drying may not be necessary; however, it is recommended, as small deviations from the ideal separator may cause free water to form in the pipeline. Some CO2 will dissolve in water and there will be a loss of CO2 to the ambient. For power cycles that contain large amounts of water in contact with the CO2, it might be of interest to design the capture and gas processing so that the CO2 emissions with water are kept low. The solubility of CO2 in water will decrease with higher temperature and increase with higher pressure. The water from intermediate pressure separator drums should therefore be sent to a final drum at ambient pressure. The solubility of CO2 in water at 1 bar and 20 °C is approximately 0.1 mol% (Diamond and Akinfiev, 2003).

12.4.3 Removal of water by adsorption After the last separator drum, the CO2 gas can be dried to single digit ppm level by regenerative adsorption columns (molecular sieves or silica) which can © Woodhead Publishing Limited, 2010

392


be regenerated with heated CO2. The purge gas from the adsorption column should be cooled and recycled into one of the first vapour–liquid separator drums to avoid purging of CO2. Water drying in adsorption columns will not impact the efficiency or the CAPEX and OPEX to any great extent and should therefore be included to avoid operational problems. Furthermore, since the gas is dried anyway, there is not much to gain from increasing the water content from e.g. 50 ppm to 200 ppm.

12.4.4 Removal of unwanted components by chemical or physical treatment or advanced separation processes Unwanted components that cannot be a part of the CO2 product and cannot be removed in the separator drums or volatile column should be avoided in the capture process or be removed as early as possible in the capture or transport process to avoid unnecessary handling. Such components can be removed by absorption or adsorption, rigorous distillation or membranes, or burnt in catalytic processes. One example is removal of H2S, which could be removed by the Rectisol process or by molecular sieves; however, it is more cost-efficient to remove the H2S in the capture process, as discussed in a later section. White et al. (2007, 2009) argues that components like H2S will react with water in the separator drums and form sulphuric acid. Further work should be conducted to prove this theory, and to find the conversion rate.

12.4.5 Condensation If volatiles are to be removed, the CO2 must be partly or fully condensed. Furthermore, the most energy-efficient pipeline process and all of the liquefaction processes require the CO2 to be condensed. For low seawater temperatures, the CO2 can be condensed at 60–65 bar by seawater; if not, an external refrigeration unit, with e.g. ammonia as the working fluid, must be applied. For pipeline transport, the CO2 should then be condensed at 50–65 bar and for ship transport at 30 bar.

12.4.6 Pumping In the most energy-efficient pipeline process, the CO2 is pumped in membrane or centrifugal pumps to achieve transport pressure. Pumping of CO2 from 65 bar to 150 bar is about 10–15 % more energy-efficient than compression; this will, however, depend on the cooling water temperature.



393

12.4.7 Expansion of carbon dioxide (CO2) (liquefaction) For ship transport, the CO2 that is condensed needs to be expanded to the transport pressure of 6.5 bar. The most energy-efficient way is to subcool the CO2 by cold flash gas at each stage. Furthermore, an expander will be more efficient (and costly) than a JT-valve, where the gas is allowed to expand adiabatically, resulting in lowering of its temperature. After the expansion, some flash gas will form. The flash gas needs to be recompressed and is therefore sent to the compressor train at the appropriate pressure level. After heating, the flash gas could be used to regenerate the adsorption dryers.

12.4.8 Removal of volatile gases (distillation) Volatiles, like N2 or Ar, will usually not account for any safety or operational problems for the pipeline transport process or the final storage. However, as transport is both energy- and cost-intensive, it makes little sense to process and transport the volatiles. Hence, the purity in pipeline transport is determined by technical and economical evaluations as there is also an energy and capital penalty for removing unwanted components. In ship transport near the TP, however, most volatiles must be removed to avoid dry ice formation during liquefaction or transport, meaning that there is less room for economic evaluations here when optimizing the CO2 processing. The volatiles should be removed in a column instead of a flash separator to reduce the amount of CO2 in the volatile purge stream. The quality of the purge stream depends on the feed gas composition and the column condenser and reboiler duty. A high condenser duty will reduce the content of CO 2 in the purge stream. A high reboiler duty will reduce the content of volatiles in the CO2 product. If deemed necessary, a two-column system can be used to enhance the purity of the CO2 product and reduce the CO2 in the purge stream. As a rule of thumb, in a one-column system, the loss of CO2 to the purge stream in moles will be equal to the amount of volatiles in moles.

12.4.9 The carbon dioxide (CO2) product stream Components with boiling points and densities similar to CO2 will be accumulated in the CO2 if not removed by the means of more advanced separation techniques. If it is technically possible to inject the component in the reservoir together with the CO2, this is a cost-efficient solution, as the presence of these components is merely a design, safety and operational challenge. If these components cannot be a part of the product stream, they must be removed, at a potentially high cost. Components that cannot easily be separated from CO2 by flashing or simple distillation are propane, ethane, H2S and SO2. It is very unlikely that propane and ethane should remain after


394


combustion of natural gas and, in any case, small quantities of propane and ethane will not cause any operational problems. H2S, on the other hand, might be a safety hazard and should be removed in the capture process. Furthermore, it is not clear if H2S will be regarded as industrial waste and therefore cannot be injected in reservoirs or aquifers due to legal restrictions. A superstructure for CO2 transport processes is presented in Fig. 12.3. Purge CCS power plant Water removal in separator drum, compression and cooling

Water removal in separator drum, compression and cooling

Water removal by adsorption

Yes

Unwanted components

Yes

Allowed in CO2 product No

No

Remove components

No

Yes

Seawater @ < 15 °C No

Ship

S3

S2


Condensation by external refrigeration

Compression and cooling 100 bar Expansion to 60 bar

Flash gas

Yes

Volatiles < 5 %

Pipeline

Pipeline or ship

P1 and S1

Volatiles in product

30 bar No


Yes

P2

P3




Condensation by external refrigeration

Compression

Condensation by seawater 60–65 bar No

Volatiles > 0.2 % Yes Removal of volatiles

Pipeline

Pipeline or ship

Pumping

Ship

Expansion to 6.5 bar Storage

150 bar

Pipeline

12.3 A superstructure for CO2 transport processes.


Ship

6.5 bar


395

As can be seen, the first three building blocks, water removal by separation, compression and cooling and gas drying in adsorption column, are common to all the processes. If unwanted components such as sulphur (H2S and SO2) or other contaminants that cannot be removed by distillation are present, they should be removed at this point. If seawater is available at a temperature lower than 15 °C and volatile content is less than 5–10 mole%, the CO2-rich gas should be compressed and cooled in two stages and condensed. If more than 0.2 mole% of volatiles are present, they should be removed in a column with reboiler and condenser. For pipeline transport, the CO2 should be pumped to transport pressure and sent to the pipeline (P1). For ship transport, the CO2 should be expanded to transport pressure and the flash gas recompressed in the compressor train (S1). If ambient air heat exchangers or cooling water above 15 °C is used, or the amount of volatiles is above 5 mole%, the pipeline processes should be compressed to 50–65 bar and cooled by an external refrigeration unit before it is sent to the volatile removal column (P2). Alternatively, the gas can be compressed directly to transport pressure (P3); however, this requires more energy and the volatiles cannot be removed. For ship transport, the CO 2 should either be liquefied at 30 bar by an ammonia refrigeration unit (S2) or, alternatively, be compressed to a pressure higher than the CP, e.g. 100 bar, and cooled by the ambient before it is expanded to 60–65 bars (S3). In both cases, the CO2 is sent to the volatile removal column. It should be noted that the choice of method depends on both the cooling water temperature and amount of volatiles, and that the numbers given above are not absolute, but are rough guidelines. The energy requirements for the six processes for a seawater temperature of 10 °C, a feed gas pressure of 1.1 bar and a volatile content of 0.5 mole% are shown in Table 12.3. Simplified flow-sheets of process P1 and process S1 are shown in Figs 12.4 and 12.5, respectively.

12.5

Sensitivity analysis

In this section, the energy requirement for the presented processes will be investigated with respect to ambient temperature, CO2 inlet pressure and volatiles (nitrogen) in the feed gas. Li et al. (2008) published a paper that Table 12.3 Energy requirements for the transport processes Process

P1

P2

P3

S1

S2

S3

Energy requirement [kWh/t] Removal of volatiles possible Removal of volatiles necessary

96 Yes No

102 Yes No

109 No No

110 Yes Yes

113 Yes Yes

123 Yes Yes


Compressor


Cooler Feed gas

15

Separator drum

Cooler 15

1

Compressor

3.5

Separator drum

Compressor

Cooler 15

15

12

Separator drum

35

Condenser Condenser 15

Separator drum

Water to water treatment

60 Volatile removal column

CO2 to pipeline

Regenerative adsorption dryer CO2 pump Heater II

12.4 Pipeline transport process P1.

Heater I

Reboiler


Volatiles

396

Compressor

Electric motor

Volatile components


Feed gas

15

13

14

5

1

3.5

12

35

52

Water to water treatment

-52

6.5


14

Liquefied CO2 to tank

397

12.5 Ship process S1.

398


gives a theoretical view of impurity impacts on the purification process in oxyfuel combustion based CO2 capture and storage system. They show the impacts of the vapour liquid equilibrium, enthalpy, compression and non-condensable gas separation for impurities such as O2, SO2, Ar and N2. Häring (2008) shows the energy requirements for liquefaction at 18 bar for small-scale liquefaction of CO2 as a function of volatiles present in the feed gas. Darde et al. (2009) shows the impact of removal of the volatiles from oxyfuel. The results presented in this section are primarily found in a paper including sensitivity analysis for ship-based transport processes (Aspelund et al., 2006) and pipeline- and ship-based processes (Aspelund and Jordal, 2007). A description of the processes including background data can be found in these papers. The CO2 purity is set to be 99.7 % for the P1, P2, S1, S2 and S3 processes. The purge gas has a CO2 concentration of roughly 40 mole%, and the calculations are performed on a feed gas basis, including the volatiles; thus the energy requirements for the transported CO2 will be lower when volatiles are present in the stream. The results in Figs 12.6–12.8 are in reasonable agreement with the other publications. It should be noted that the energy requirements may be reduced by allowing for higher contents of volatiles in the product stream. The energy requirements can also be increased by decreasing the CO2 content in the purge stream. Finally, more sophisticated column systems may be used. The results shown in Figs 12.6–12.8 should therefore be used only as a general guideline. By investigation of Figs 12.6–12.8, it can be concluded that increasing the pressure of the CO2 feed gas will greatly reduce the energy requirements S1

S2

S3

P1

P2

P3

Energy requirements (kWh/t)

115

95

75

55

35 15 0.0

10.0 Inlet pressure (Bara)

20.0

12.6 Energy requirements as a function of feed gas pressure for six transport processes.



399

170 Requireed energy (kWh/t)

160 150 140 130 120 110 100 90 80

5

10

15 20 25 30 35 40 Temperature for heat rejection to the ambient (°C) S1

S2

S3

P1

P2

45

P3

12.7 Energy requirements as a function of temperature for heat rejection for six transport processes.

Energy requirements (kWh/t)

165 155 145 135 125 115 105 95 85 0.0

S1 1.0

2.0

S2

S3

P1

3.0 4.0 5.0 6.0 7.0 8.0 Nitrogen content in feed (Mole %)

P2 9.0

P3 10.0

11.0

12.8 Energy requirements as a function of feed gas volatile content (nitrogen) for six transport processes.

both for pipeline and ship transport processes. In contrast, increasing the amount of volatiles or increasing the ambient temperatures will increase the energy requirements. It can also be seen that some processes will not work if the amount of volatiles or the ambient temperature is too high. It is worth noticing that the difference in energy requirements between the pipeline processes and the ship processes is roughly 20 %. The processes consist to a large extent of the same building blocks and the difference in CAPEX and OPEX will be in the same order of magnitude. A rule of thumb could be that


400


liquefaction processes are 30 % more expensive than compression processes. Furthermore, the feed gas quality will have a large impact on the energy requirements as well as the costs, and the difference in energy requirement and cost may be higher than between ship and pipeline processes.

12.6

The interface between capture and transport

In gas-fired power plants, there will usually be small amounts of sulphur components; thus, in most cases, sulphur recovery will not be needed. In coal-fired power plants, the H2S will be removed in the capture process to be within the specifications; hence further treatment for H2S in the transport process is not required. In a standard IGCC process, the H2S can be removed by the Selexol process. In membrane-based integrated coal gasification combined cycles, a Rectisol process can be used. For post-combustion processes, the H2S can be removed by absorption prior to the CO2 removal, if necessary. The gas purification processes are described by Kohl and Nielsen (1997). In most cases, and definitely for large fractions of H2S, it is more cost- and energy-efficient to remove the H2S in the capture process than in the transport process. Most of the components with low boiling points, such as N2, O2, NO, CO, H2, CH4 and Ar, will be found in the volatile purge stream. The purge stream flow will be much smaller than the total flue gas flow, and parts of it could typically be recycled back to the power process with CO2 capture; however, a purge is necessary to avoid accumulation of N2, Ar and other contaminants. Such a recycling would mean that the heating value of components such as CO, CH4 and H2 can be made use of if they are recycled to a point prior to the combustion chamber. H2 and Ar are valuable substances that may be recovered as saleable products. The CO2-rich streams can roughly be divided in three categories, depending on their composition. These three categories and the different options for gas conditioning and components recycle are schematically illustrated in Fig. 12.9. The simplest group is Category 1 consisting of only CO2 and water, possibly with traces of oxygen and of solvents that were employed to separate the CO2 from the exhaust or syngas. Processes with minor fractions of N2 could also be referred to as Category 1. The process integration options depend on possible heat integration options and cleaning and recovery of the condensed water. The relative simplicity of the CO2 processing compared to the other two groups may be a very attractive feature for this category of capture processes. Recalling that the composition in the pre-combustion autothermal reforming (ATR) case in practice will contain some H2, this signifies that post-combustion capture should have an advantage over precombustion from a gas conditioning point of view. In the case of Category 2, typically consisting of CO2-rich streams from oxyfuel power processes with cryogenic air separation unit (ASU), the


CO2, water

Gas conditioning

CO2 for transport

Water, amines


Non-combustibles and CO2 Purge Oxyfuel capture processes with cryogenic ASU

Advanced separation

Non-combustible volatiles

CO2 CO2, water, Non-combustible volatiles (Ar, N2, O2)

Gas conditioning

CO2 for transport

Water Combustibles Non-combustibles, combustibles and CO2 Purge Capture processes with ‘innovative reactors’ and incomplete fuel conversion

CO2, water, non-combustible volatiles (Ar, N2, O2), combustible volatiles (CO, CH4, H2)

Advanced separation

Volatiles

CO2 Gas conditioning

CO2 for transport


Pre- and post-combustion capture processes with solvents

Water

401

12.9 The interface between capture and transport (adapted from Aspelund and Jordal, 2007).

402


processes with volatile purge removal (P1, P2, S1, S2 and S3) will generate a purge gas consisting of non-combustible volatiles and CO2. The amount of CO2 in the purge stream will vary depending on the amount of volatiles and the column configuration. A large fraction of this purge stream may be envisaged to be recycled back to the power cycle in order to avoid CO2 emissions. Alternatively, the stream may be fed to an advanced separation system, after which the CO2 may be recycled at an appropriate pressure in the gas conditioning and the volatiles may be vented to the atmosphere, or be recovered as saleable products. For Category 3 of CO2-rich streams, there are also combustible gases in the purge stream, which offers some interesting prospects. It may very well be that the group of reactors that may be referred to as ‘innovative’ (membrane reactors, sorption enhanced reactors, fuel cells and chemical looping reactors) that are sometimes studied in the context of CO2 capture should not be pushed towards maximum yield. Since full fuel conversion is virtually impossible in these reactors, it should be further investigated whether it is more beneficial to optimize the CO2 capture process at a lower fuel conversion rate than the highest possible one, and integrate the capture process with the gas conditioning process with recirculation of the combustible volatiles, keeping in mind the operational and safety challenges that this may imply. In Fig. 12.9, two possible ways of recycling the combustibles back into the capture process are sketched. The purge gas, containing both combustibles and non-combustibles, can be recycled directly to the capture process. Alternatively, it can be sent to an advanced separation, which separates the combustibles from the CO2 and other volatiles, before the combustibles are recycled. The most effective way to treat the purge gas, and where in the process the recycle should occur, are topics for future work. This includes the evaluation that must be made for different cycles, whether the combustibles should be recycled in such a way that the capture process can be a virtually zero-emission cycle, or in such a way that the combustibles are burned. If the combustibles are burned and they contain CO or CH4, this will lead to CO2 emissions to the atmosphere. However, in the pre-combustion ATR case, in practice, the combustibles consist mainly of hydrogen and, from an emissions point of view and in order to maximise process efficiency, recirculation of this hydrogen to the gas turbine combustion chamber appears to deserve further investigation, although operational and security issues must also be taken into account.

12.7

Ship to pipeline and pipeline to ship processes

Before injecting the CO2 to a reservoir, the LCO2 must be pumped and heated to transport specifications before the CO2 can be transported in a pipeline. The process will require roughly 4–5 kWh/t CO2 due to pumping



403

from 6.5 bar to 150 bar, and large amounts of heating. If waste heat or warm seawater (> 15°) is available, such a process can be designed relatively easily. However, in areas with cold seawater or cold ambient air, fuel oil may be required to heat the CO2, increasing the energy demand significantly and causing some CO2 emissions. Furthermore, although the process is relatively simple, the required equipment is relatively large and further work should be conducted to find suitable equipment for a reasonable cost. There is a shortage of literature about this topic; however, Aspelund et al. (2007) describes a process for pumping and heating the CO2 from a gas carrier to be injected in an offshore field. CO2 at 100–150 bar can be liquefied by expansion of the gas to transport pressure, as shown for process S3 in Fig. 12.3. The flash gas that forms needs to be recompressed and liquefied. To reduce the flash gas, the high-pressure gas can be cooled by an external cooling unit. The energy requirement for this process is approximately 30 kWh/t CO2.

12.8

Discussion

Although the basic building blocks for the transport processes are simple and based on existing technology, it is still a challenging task to design the most cost- and energy-efficient transport process. The energy requirements and the costs depend to a large degree on the feed gas pressure and composition, as well as the transport and reservoir specifications. In the simplest pipeline transport, the volatiles cannot be removed. In contrast, for ship transport, a purity of 99.7 % is required to avoid problems with the dry-ice formation. For long distances, it will most likely be sensible to remove most of the volatiles for pipeline transport as well, as the costs for building a larger pipeline and transporting these volatiles may be higher than the energy and capital costs for removal. This can be done following a path similar to that used for ship transport processes, the exception that the gas is pumped to high pressure instead of being expanded to low pressure for liquefaction at 6.5 bar. Therefore, the costs for pipeline and ship transport processes will be quite similar, the latter demanding 20 % more energy and having in the order of 30 % higher investment costs. By investigating the impact of feed gas pressure, feed gas composition and ambient temperature, it can be concluded that the difference in costs and energy requirements caused by these factors is larger than the difference in transport method. The purge stream from the volatile removal column will contain a relatively large fraction of CO2, and should therefore not be emitted directly to the atmosphere. If possible, as much of the purge stream as possible should be recycled to the power plant. If this is not possible, the purge stream should be further processed and the CO2 sent back to the compressor train. There is a lack of thermodynamic data as well as models for CO2-rich


404


gas mixtures. The thermodynamic models in commercially-available process simulators should be used with care, especially for mixtures including water, CO2 and hydrocarbons. Although pipeline transport of CO2 is done both onshore and offshore and ship transport is done for small food-grade quality CO2 (almost 100 % pure and at a pressure of 15–18 bar), it is important to build demonstration plants in order to gain knowledge about operational procedures as well as investment and operational costs. One particular point is that CO2 has other qualities than natural gas, and that cost data from liquefied natural gas (LNG) cannot be used for LCO2.

12.9

Future trends and future work

Due to increasing electricity prices, the efficiency of the transport processes is becoming more and more important. Furthermore, it is expected that more research will be conducted on the integration with the CCS plant on a case-by-case basis in order to find the most cost-efficient power plant with CO2 capture, conditioning and purification. One specific field of interest is the combined ship transport of LNG and LCO2. The concept has two main advantages. The gas carrier can be used in both directions and the cold exergy in the LNG and LCO2 can be recycled to increase the efficiency significantly. One concept of particular interest is the Liquefied Energy Chain presented by Aspelund and Gundersen (2009a,b,c,d). In this concept, nitrogen is used as a cold carrier together with CO2. This will reduce the energy requirements by 50 % and may reduce the CO2 losses by up to 10 percentage points, compared to conventional ship transport. Although some initial work has been done on the gas conditioning and purification of large-scale transport of CO2, there are still some challenges to meet in order to find the most-efficient transport processes. More work needs to be done to achieve realistic minimum specifications for CO2 transport both by pipelines and ships. The main questions are whether gas drying is required for pipeline transport of CO2 in warm areas, and how much H2S and CO can be allowed for both pipeline and ship transport. It is also unclear from the literature whether the H2S will react with water to form sulphuric acid or not; furthermore, if H2S can be removed this way, how should the dilute sulphuric acid be treated? Finally, to answer these questions and design as efficient processes as possible with a high level of confidence, better thermodynamic data and equations are required; it is known that small amounts of certain components will change the thermodynamic properties significantly, one example being the solubility of water in pure CO2 versus CO2 with some CH4.



405

12.10 Conclusions The optimal CO2 transport process depends on the feed gas specification from the capture plant, the transport specifications, the reservoir specifications and legislation, as well as the ambient conditions and cooling water availability. It is important to acknowledge the interface between the capture and transport process and not impose unnecessarily strict specifications for the reservoir, so that the most cost- and energy-efficient system can be designed. The pipeline and ship transport processes are similar and consist of the same basic building blocks; however, the pipeline transport process is roughly 20 % more energy-efficient and is expected to have 30 % less investment costs than a liquefaction process. Feed gas pressure, the amount of volatiles and ambient temperature may increase the energy requirements from 90 and 110 kWh/t CO2 for pipeline and ship transport, respectively, to up to 200 kWh/t. The specifications for ship transport are stricter than for pipeline transport. It is possible to transform pressurized CO2 at e.g. 100 bar in a pipeline to liquefied CO2 at 6.5 bar in a semi-pressurized ship and vice versa for 4 and 30 kWh/t, respectively. In order to design the most efficient processes possible with a high level of confidence, better thermodynamic data and equations are required. Although transport processes are generally thought of as being the least technologically challenging part of the CCS chain, it is important to build demonstration plants to get a better understanding of the costs and energy requirements involved in the gas conditioning and processing of CO 2 from large-scale capture plants.

12.11 Acknowledgements This work is to a large extent based on results from studies from SINTEF and NTNU. I would like to thank the researchers involved in these studies, especially Kristin Jordal from SINTEF for the valuable discussions.

12.12 References Aspelund A and Gundersen T (2009a) A liquefied energy chain for transport and utilization of natural gas for power production with CO2 capture and storage – Part 1, Appl Energy, 86 (6), 781–792. Aspelund A and Gundersen T (2009b) A liquefied energy chain for transport and utilization of natural gas for power production with CO2 capture and storage – Part 2: The offshore and the onshore processes, Appl. Energy, 86 (6), 793–804. Aspelund A and Gundersen T (2009c) A liquefied energy chain for transport and utilization of natural gas for power production with CO2 capture and storage – Part 3: The combined carrier and onshore storage, Appl Energy, 86 (6), 805–814. Aspelund A and Gundersen T (2009d) A liquefied energy chain for transport and utilization of natural gas for power production with CO2 capture and storage – Part 4: Sensitivity analysis of transport pressures and benchmarking with conventional technology for gas transport, Appl Energy, 86 (6), 815–825. © Woodhead Publishing Limited, 2010

406


Aspelund A and Jordal K (2007) Gas conditioning – the interface between CO 2 capture and transport, J Greenhouse Gas Control, 1 (3), 343–354. Aspelund A, Mølnvik M J and De Koeijer G (2006) Ship transport of CO2 – technical solutions and analysis of costs, energy utilization, exergy efficiency and CO2 emissions, Chem Eng Res Des, 84-A9, 847–855. Aspelund A, Sandvik T E, Krogstad H and de Koeijer G (2005a) Liquefaction of captured CO2, for ship-based transport, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Vol. 2, 2245–2549. Aspelund A, Weydahl T, Sandvik T E, Krogstad H, Wongraven L R, Henningsen R, Fivelstad J, Oma n and Hilden T (2005b) Offshore unloading of semi-pressurised CO2 to an oilfield, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Vol. 2, 2552–2554. Austegard A, Solbraa E, De Koeijer G and Mølnvik M J (2006) Thermodynamic models for calculating mutual solubilities in a H2O–CO2–CH4 mixture, Chem Eng Res Des, 84–A9, 781–794. Barrio M, Aspelund A, Weydahl T, Sandvik T E, Wongraven L R, Krogstad H, Henningsen R, Mølnvik M J and Eide S I (2005) Ship-based transport of CO2, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Vol. 2, 1655–1660. Berger B, Kaarstad O and Haugen H A (2005) Creating a large-scale infrastructure for enhanced oil recovery, in Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Vol. 1, 321–328. Darde A, Prabhakar R, Tranier J P and Perrin N (2009) Air separation and flue gas compression and purification units for oxy-coal combustion systems, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 527–534. De Visser E, Hendriks C, Barrio M, Mølnvik M J, De Koeijer G, Liljemark S and Le Gallo Y (2008) Dynamic CO2 quality recommendations, J Greenhouse Gas Control, 2, 478–484. Diamond L W and Akinfiev N A (2003) Solubility of CO2 in water from –1,5 °C to 100 °C and from 0,1 to 100 MPa: evaluation of literature data and thermodynamic modelling, Fluid Phase Equilib, 208, 265–290. Häring H W (2008) Industrial Gases Processing, Wiley-VCH, Weinheim, Germany. Hegerland G, Jørgensen T and Pande J O (2005) Liquefaction and handling of large amounts of CO2 for EOR, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Vol. 2, 2541–2544. Heggum G, Weydahl T, Roald W, Mølnvik M J and Austegard A (2005) CO2 conditioning and transportation, in Thomas DC and Benson SM (eds), Carbon Dioxide Capture for Storage in Deep Geologic Formations, 2, Elsevier, Oxford, UK. 925–936. Hendriks C et al. (2002) Building the cost curves for CO2 storage, Part 1: Sources of CO2, PH4/9, IEA Greenhouse Gas R&D Programme, Cheltenham, UK. Knauss K G, Johnson J W and Steefel C I (2005) Evaluation of the impact of CO 2 co-contaminant gas, aqueous fluid and reservoir rock interactions on the geologic sequestration of CO2, Chem Geol, 207, 339–350. © Woodhead Publishing Limited, 2010


407

Kohl A and Nielsen R (1997) Gas Purification, Gulf Publishing Company, Houston, TX. Kvamsdal HM, Jordal K and Bolland O (2007) A quantitative comparison of gas turbine cycles with CO2 capture, Energy 32, 10–24. Li H, Yan J and Anheden M (2008) Impurity impacts on the purification process in oxy-fuel combustion based CO2 capture and storage system, J Appl Energy, 86, 202–213. Michael K, Allison G, Golab A, Sharma S and Shulakova V (2009) CO2 storage in saline aquifers II–Experience from existing storage operations, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 1973–1980. Oryshchyn D, Ochs T, Gerdemann S, Summers C and Patrick B (2007) Developments in integrated pollutant removal for low-emission oxy-fuel combustion, in Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK, Elsevier, CD-ROM. Ozaki M, Davison J and Minamimura (2005), Marine transportation of CO2, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Vol. 2, 2535–2539. Sass B M, Farxan H, Prabhakar R, Gerst J, Sminchak J, Bhargava M, Nestleroth B and Figueroa J (2009) Considerations for treating impureties in Oxy-Combustion flue gas prior to sequestration, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 535–542. Skovholt O (1993) CO2 transportation system, Energy Conver Manage, 34 (9–11), 1095–1103. Steeneveldt R, Berger B and Torp T A (2006) CO2 capture and storage: closing the knowing–doing gap, Chem Eng Res Des, 84–A9, 739–763. Svensson R, Odenberger M, Johnsson F and Strömberg L (2003) Transportation systems for CO2–application to carbon capture and storage, Energy Convers Manage, 45, (15–16), 2343–2353. Vandenhengel W and Miyagishima W (1993) CO 2 capture and use for EOR in Western Canada 2. CO2 extraction facilities, Energy Convers Manage, 34 (9–11), 1151–1156. White V, Allam R and Miller E (2007) Purification of oxyfuel-derived CO2 for sequestration or EOR, in Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK, Elsevier, CD-ROM. White V, Torrente-Murciano L, Sturgeon D and Chadwick D (2009) Purification of Oxyfuel-derived CO2, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 399–406. Zanganeh K, Shafeen A and Salvador C (2009) CO2 capture and development of an advanced pilot-scale cryogenic separation and compression unit, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 247–252.


13

Infrastructure and pipeline technology for carbon dioxide (CO2) transport

P. N. S e e v a m, J. M. R a c e, and M. J. D o w n i e, Newcastle University, UK Abstract: This chapter focuses on technical issues related to the transport of carbon dioxide (CO2) in a transport infrastructure for CO2 capture and storage designed for climate change mitigation. Naturally occurring CO2 has been transported for enhanced oil recovery in the USA and Canada for many years now and the practice is well understood. However, there are significant differences between established practice and the requirements of anthropogenic CO2 transport systems to be instituted in countries in other parts of the world in response to climate change. This chapter concentrates on issues arising from those differences with particular reference to CO2 transport in the UK. Some of the issues, such as those relating to the specification of the CO2 quality for transport, have been introduced in Chapter 12 and will be developed further in the context of their implications for the design and operation of a CO2 transport infrastructure. The chapter starts by identifying some of the differences between existing and future systems and goes on to describe the basic properties of CO2 relevant to the issues arising from them. It proceeds to discuss some of the technical aspects of the transport of CO2 by pipeline and ship followed by the development of a wider transport infrastructure. It concentrates on pipelines because in the early days of the development of large-scale CO2 capture and storage, it is anticipated that this will be the dominant mode of CO2 transport. Key words: CO2 transport, transport, pipeline, ship, impurities, hydraulics, materials, fracture, corrosion, regulation.

13.1

Introduction

The transport of CO2 is routinely practised in the USA and Canada and also exists in other parts of the world, as mentioned in Chapter 12. The activity largely involves transporting CO2 by pipeline for enhanced oil recovery (EOR) and food-grade CO2 by ship, and the technology for these purposes is well established. The transport of CO2 for capture and storage (CCS) aimed at climate change mitigation will build on existing technologies, but it will also have new challenges with which to contend. In the USA, for EOR, naturally occurring CO2 is transported considerable distances overland through mostly sparsely populated regions. There is also some limited transport of anthropogenic CO2 (e.g. Gill, 1985). CCS for climate change mitigation will involve capturing CO2 from power plants (or other large stationary industrial sources), transporting it to one or more geological 408 © Woodhead Publishing Limited, 2010

Infrastructure and pipeline technology for CO2 transport

409

storage sites (such as depleted oil or gas wells, saline aquifers or disused coalmines) and injecting it into the rock formation. The storage sites have to be selected so that they are secure for geologically significant timescales, if not permanently. The countries that are likely to adopt CCS, almost by definition, will be heavily industrialised and densely populated. Some of them may intend to store the CO2 in subsea locations. The UK, for example, has a number of suitable offshore CO2 sinks in the North and Irish Seas. It has been commonly assumed that the transport of CO2 from UK sources to offshore sinks is straightforward, and may even be able to make widespread use of existing infrastructure. However, there are significant differences between the US experience and the UK transport requirements. The UK will be dealing with anthropogenic CO2, mostly from power plant, which will impose constraints on the hydraulics that have not yet been fully explored. Considerable proportions of the transport system will be subsea, of which there is as yet virtually no experience. There are questions as to the suitability of much of the existing infrastructure and the desirability of using it. There is little experience with multisource transport systems through densely populated regions. This chapter will address some of the technical issues that need to be considered for the development of a CO2 transport infrastructure, in the postdemonstration phase of CCS, capable of mitigating the emission of greenhouse gases whilst allowing the continued use of fossil fuels. It is based on work looking at the CO2 transport needs for the UK (Downie et al., 2007; Race et al., 2007; Seevam et al., 2008) carried out under a research programme run by the UKCCSC (UK Carbon Capture and Storage Consortium). A number of the issues, such as the factors influencing the specification of the fluid to be transported, have already been introduced in Chapter 12.

13.2

Carbon dioxide (CO2) phase properties

To date, there is no accepted specification for the CO2 stream to be transported in a CCS infrastructure designed to mitigate climate change. The specification for storage only may be less stringent than for EOR, and it is possible that the composition of the fluid will be different from any that has been transported hitherto. The phase properties of the CO2 stream are sensitive to its composition and they, in turn, determine its hydraulic characteristics in pipeline transport. The successful design of a safe, economic and efficient transport infrastructure must be informed by these relationships.

13.2.1 Properties of pure CO2 When considering the hydraulic properties of CO2 it is useful to go back and recast the phase diagram presented in Chapter 12 Fig. 12.1 as a


410


pressure–temperature plot as shown in Fig. 13.1. The phase diagram for pure CO2 contains two distinct features, the triple point (5.2 bar, –56 °C) and the critical point (74 bar, 31 °C) marked TP and CP, respectively, in the figure. In the vicinity of the triple point, CO2 can exist as one of three phases; solid, liquid or gas. The curve connecting the two points is the vapour–liquid line separating the gaseous and liquid phases. At pressures and temperatures above critical, CO2 no longer exists in distinct gaseous and liquid phases, but as a supercritical phase with the density of a liquid and the viscosity of a gas. Increases in pressure no longer produce liquids at temperatures exceeding the critical temperature. At pressures above but temperatures below critical, the CO2 exists as a liquid whose density increases with decreasing temperature. For these reasons, the most efficient way of transporting CO2 by pipeline is in its supercritical or dense phases in the vicinity of the critical point. Similarly, when transporting it by ship it is desirable to carry it in semipressurised ships at pressures and temperatures nearer to the triple point.

13.2.2 Composition of CO2 transport stream

120 100

Supercritical CO2

140

Pressure [bar]

As mentioned in Chapter 12, the gas composition from the CO2 source will depend on the capture process and its feedstock, and there are a number of reviews on the subject. The IPCC special report (IPCC, 2005) identified the composition ranges for CO2 captured from coal-fired or gas-fired power stations. As part of the EU Framework 6 ENCAP (European Enhanced Capture of CO2) project, Vattenfall have analysed the potential compositions of the

Dense phase CO2 80

CP

60

Solid CO2

40 Liquid CO2 20

Vapour CO2

TP 0 –100

–80

–60

–40 –20 Temperature (°C)

0

20

40

60

13.1 Phase diagram for pure CO2 (adapted from Seevam et al., 2008).



411

CO2 stream that could result from post-combustion, pre-combustion and oxyfuel capture (Anheden et al., 2005). The Oosterkamp and Ramsen (O&R) quality specification (Oosterkamp and Ramsen, 2008) is based on compiled ranges of impurities published by IEA (2004a) and White et al. (2007) for post- and pre-combustion capture. The composition for the oxyfuel stream is derived from both IEA (2004a) and White et al. (2007), with additional analysis conducted using the NIST thermophysical properties of fluid systems database. Table 13.1 shows a set of CO2 stream compositions that have been compiled by Seevam (2010) to investigate the effect of composition on their phase diagrams. The compositions do not include water, but are otherwise derived from those presented in the studies above and have been selected to cover the range of compositions that occur in the literature. Specifications 1, 2 and 3 in Table 13.1 are based on the IPCC, ENCAP and O&R oxyfuel specifications, respectively. The CO2 concentration has been determined by assuming that all of the other impurities are present at the maximum limit specified and calculating the percentage CO2 as the remainder, ensuring that the CO2 minimum specification is still met. The maximum ranges of impurities that could be present on the basis of this analysis are shown in Table 13.1. The CO2-rich gas from the source must be processed for transportation as described in Chapter 12 which also presents recommendations for the quality of the product streams for pipelines and ships. As described, the exact composition will be a tradeoff, relating not only to the capture technology but Table 13.1a Possible CO2-rich gas composition from post-combustion capture

Component

Post-combustion

Spec. 1

Spec. 2

Spec. 3

Unit

Coal

Gas

Coal

Coal & Gas

CO2 CH4 N2 H2S C2+ CO O2 NOX SOX H2 Ar HCN COS

vol% vol% vol% vol% vol% vol% vol% vol% vol% vol% vol% vol% vol%

99.97 – 0.0033* – – – 0.0033* 0.01 0.01 – 0.0033* – –

99.97 – 0.0033* – – – 0.0033* 0.01 0.01 – 0.0033* – –

99.95 – 0.021 – 0.003 0.001 0.003 0.002 0.001 – 0.021 – –

99.79 0.01 0.17 Trace 0.01 0.001 0.01 0.005 0.001 Trace Trace – –

* The limit for N2, O2 and Ar in the IPCC specification was quoted as a total for all three impurities in combination. For the purposes of comparison, it is assumed that the impurities are present in equal proportions.


412


Table 13.1b Possible CO2-rich gas composition from pre-combustion capture Component

Pre-combustion

Unit

Spec. 1 Coal

Gas

Spec. 2

Spec. 3

Coal

Coal & Gas

Max purity Min purity

CO2/H2S

CO2 + H2S


99.14 0.01 0.0033* 0.01 – 0.03 0.0033* – – 0.8 0.0033* – –

98.00 0.035 0.03 0.01 – 0.17 – – – 1.7 0.05 0.0005 0.0005

95.72 0.035 0.03 2.3 – 0.17 – – – 1.7 0.049 0.0005 0.0005


96.39 0.01 0.2* 0.6 – 0.4 0.2* – – 2.0 0.2* – –

95.66 2.0 0.43 0.01 – 0.04 0.43 – – 1.0 0.43 – –

95.6 0.021 0.35 1.99 0.059 0.23 Trace – – 1.8 0.03 N/A –

Table 13.1c Possible CO2-rich gas composition from oxyfuel capture Component

Oxyfuel

Unit

Spec. 1 Coal

Gas

Spec. 2

Spec. 3

Coal

Coal & gas

CO2/SO2

CO2 + SO2


90.76 – 0.61 – – – 1.6 – 0.076 0.25 5.7 – –

90.46 – 0.60 – – – 1.6 – 1.5 0.24 5.6 – –


95.80 – 1.23* – – – 1.23* 0.01 0.5 – 1.23* – –

95.87 – 1.37* – – – 1.37* 0.01 0.01 – 1.37* – –

90.00 – 3.94 Trace – Trace 1.69 0.14 1.41 Trace 2.82 – –

also economics (i.e. the increased capture cost associated with the removal of impurities to low levels), and legislative and regulatory requirements, specifications and safety considerations applying at all stages of the chain. All other factors being equal, it is advantageous to retain as many components of the CO2-rich gases as possible, although their effect on the properties of the CO2 stream is not completely understood or reliably predictable.



413

13.2.3 Equations of state Modelling of phase behaviour requires the use of an equation of state (EOS) to provide a relationship between the thermodynamic variables of the system (e.g. temperature, pressure and volume) and to describe the state of the system under a given set of conditions. However, there is no consensus in the literature regarding the EOS that should be used for the design of CO2 pipelines. Commonly used EOSs include: the Benedict–Webb–Rubin–Starling (BWRS) equation (e.g. McCollough, 1986); the Peng–Robinson (PR) equation (e.g. Zhang et al., 2006); and the Soave–Redlich–Kwong (SRK) equation (e.g. Hein 1985, 1986). Li and Yan (2007) have conducted a comparative study of all of these EOSs and concluded that the selection of the EOS may have a significant impact on the pipeline design, although without experimental data it was not possible to identify the most accurate EOS to use. The captured CO2 streams in Table 13.1 can contain up to 12 different impurities, and experimental validation for these types and levels of impurities does not exist. The phase diagrams for the different compositions have been computed using proprietary software whose different EOSs were first assessed using experimental data for the binary acid gas system CO2–H2S (Bierlein and Kay, 1953), as this contains at least one of the expected impurities in the captured CO2 stream. Although both the PR and SRK EOSs provided a good fit to the experimental data in the preliminary assessment, the PR EOS was found to be slightly more accurate than the BWRS and the SRK EOSs. Figure 13.2 shows the phase envelopes for the three CO2 compositions presented in Table 13.1 (c) for the oxyfuel capture technologies reviewed. The PR EOS was used to generate the phase diagrams. The figure illustrates the two characteristic effects of the addition of other components to the pure CO2 stream, one effect being to change the critical pressure and temperature, and the other to open up the liquid–vapour line into a two-phase region. The degree to which these changes occur depends on the range of components involved. The critical temperature decreases for all the capture technology combinations of impurity in Table 13.1 and the critical pressure shows an increase, when compared to pure CO2 (Seevam et al., 2008). The postcombustion CO2 stream shows the least change in the critical temperature and pressure. It also shows similar phase properties to the pure CO2 shown in Fig. 13.1. The oxyfuel impurity combination had the greatest effect on the area of the two-phase region having the highest decrease in critical temperature and the maximum increase in critical pressure when compared to pure CO2. The phase diagrams for the oxyfuel specifications presented in Table 13.1 are shown Fig. 13.2. These effects have direct implications on the hydraulic properties of CO2 which in turn have a significant effect on CO2 pipeline design and operation as will be seen in Section 13.3.2.


414


100

P (Bar)

80

60

40

Spec. 2 Coal CO2 + SO2 Spec. 1 Coal Spec. 3 Coal and gas

20

Spec. 1 Gas Spec. 2 Coal CO2/SO2

0 0

5

10

15

T (°C)

20

25

30

35

13.2 Phase diagrams for different oxyfuel specifications.

13.3

Transport of carbon dioxide (CO2) by pipeline

For transport in pipelines, as mentioned in Chapter 12, it is generally desirable to transport CO2 in supercritical or dense phase because there is a combination of a relatively high density and a relatively low viscosity in this state. For this reason, virtually all existing CO2 pipelines are highpressure lines. Current CO2 pipelines operate from 86 bar to about 200 bar (Farris, 1983; Seevam, et al., 2008) with ambient temperatures ranging from 4 °C to 38 °C (McCollough, 1986; Mohitpour et al., 2003). There are also circumstances in which it might be desirable to transport CO2 in gaseous phase, such as when using existing infrastructure or collecting CO2 at the outer fringes of a network. Whether at high or low pressure, operators will transport CO2 in a single phase for the entire length of the pipeline from source to collection or injection facility, and avoid mixed phase flow which could result in liquid slugs in the pipeline and introduce a number of operational problems, particularly in relation to pumping/compression. This requires maintaining a buffer zone between the operating region and the envelope of two-phase flow. Prior to transport by high-pressure pipeline, captured CO2 must be transformed to dense phase by compression above its critical pressure, which itself will be determined by the impurities present in the flow. They influence the critical temperature and pressure, as well as the area of the phase envelope that dictates the region of two-phase flow, as shown in Fig. 13.2. This, in turn, may change the operating region of the pipeline, which may have to be operated at a higher pressure to maintain it as single-phase supercritical or dense phase. At the same time, the CO2 must be compressed to a pressure high enough to overcome the frictional and static pressure



415

drops along the pipeline. It may be necessary to introduce booster stations at intermediate stations along the pipeline to achieve this. If the pipeline is subsea, it is not viable to introduce intermediate booster stations and so the pressure at the departure point onshore must be sufficiently high to maintain the CO2 in dense phase to the sink location. The maximum allowable operating pressure (MAOP), minimum operating pressure, pipeline length and operating temperatures are required in the initial design stages of a pipeline. The maximum pressure relates to pipeline integrity and is usually set by the design codes, regulations and economic considerations. The minimum pressure should be set to avoid two-phase flow. The pipeline length is governed by the route chosen between the CO 2 source and the CO2 sink, which is normally dictated not only by topology and economics but also by regulations based on considerations of health and safety. Operating temperatures (other than the compressor discharge temperature) are usually determined by the ambient temperature. This is an important operating variable as it markedly affects the transport properties of CO2 in terms of density, compressibility and static head losses (i.e. increasing discharge temperature lowers static head loss) (Farris, 1983). These properties and parameters change with the presence of impurities.

13.3.1 Flow equations The design of any pipeline requires consideration of the relationship between the flow of the fluid in the pipeline and the properties of both the pipe and the fluid under the given conditions of temperature and pressure. This enables the designer to establish the capacity of the pipeline. The flow equations that are used in the design of oil and gas pipelines are derived from the equation for one-dimensional fluid flow in a horizontal pipe (as cited by Schroeder, 2001): Q=C

Tb 2.5 Ê P12 – P22 – H c ˆ D eÁ Pb Ë LGTa Z a fdw ˜¯

0.5

[13.1]

where C is a constant, D is the pipe internal diameter (mm), e is the pipe efficiency, fdw is the Darcy–Weisbach friction factor, G is the gas specific gravity, L is the pipe length (km), Pb is the pressure base (kPa), P1 is the inlet pressure (kPa), P2 is the outlet pressure (kPa), Hc is the static head, Q is the flow rate (scm/d), Ta is the average temperature (°R), Tb is the temperature base (°K) and Za is the compressibility factor. Equation 13.1 indicates the behaviour of the flow of a fluid in a pipeline with respect to the critical parameters, and it can be appreciated that there are key properties of the fluid that have to be considered in addition to pressure, temperature and elevation. Due to the unique properties of CO2, it is


416


essential to understand how these variables inter-relate and how they change along the pipe length. For pipeline design, a number of different variations on the basic flow equation have been used, but literature suggests that two flow equations in particular have been used for the hydraulic modelling of CO2 pipelines; the AGA (American Gas Association) equation (13.2) and the Beggs and Brill equation (13.3).

Q = C1

1 f

Pu2

–

Pd2

2 ¸ Ô ÔÏG D hPavg – C2 Ì ˝ ÔÓ Z avgTavg Ô˛

0.5

2.5

[13.2]

avg avg

where Q is the gas flow È rate, Pu the upstream line˘ pressure, Pd the downstream line pressure, G theÍ specific gravity compared ˙ to air, Tavg the average Ô˙ temperature, Zavg theÍ fluid compressibility at average pipeline pressure, f the E ˙ d pipeline GTpipeline, Z L d the internal friction factor, L theÍlength of diameter, Dh ˙ Í the change in elevation, E the pipeline efficiency and C are constants. i ˙˚ ÍÎ fm rn vm2 dP = 2D + rs g sin q [13.3a] dL 1 – Ek

Ek =

vm vsg rn P

rs = rL HL(q) + rg Î1 – HL(q)˚

[13.3b] [13.3c]

where fm is the Moody friction factor, g the acceleration due to gravity, HL(q) the liquid hold up fraction as a function of inclination angle q, L the pipe length, P the pressure, rg the gas density, rL the liquid density, rn the no-slip density, vm the mixture velocity, and vsg is the superficial gas velocity. Farris (1983) has shown that the AGA equation for gas flow gives good results with CO2. One of the most widely utilised AGA equations for metering purposes is that defined in AGA Report No.3 (AGA, 1990). The report is now in its third edition, but was utilised by the Cortez pipeline (Marsden and Wolter, 1986) and the Central Basin Pipeline (McCollough, 1986) for mass custody transfer measurements. Recht (1987) states that pipeline hydraulic calculations can also be carried out with the AGA flow equation but cautions that changes in pipeline elevation need to be considered carefully. The pressure drop along the pipeline can also be calculated using the Beggs and Brill correlation (Beggs and Brill, 1973) with the Moody friction factor (Moody, 1944; Brill and Mukherjee, 1999). The Beggs and Brill equation was used



417

by Hein (1986) to simulate CO2 pipeline flows as it allows prediction in the two-phase region, although it reduces to the single-phase gas flow equation if no liquid is present. Seevam et al. (2008) used the Peng–Robinson EOSs together with the Beggs and Brill flow equation to investigate the hydraulics of CO2 pipeline flow with impurities. For the scenarios modelled, a flow equation was required that could be used for gas, dense phase liquid and two-phase mixtures. The Beggs and Brill flow equation is both a single and multiphase fluid flow equation whereas the AGA equation is purely a single-phase gas equation and, therefore, the Beggs and Brill equation with the Moody friction factor was used.

13.3.2 Pipeline hydraulics The presence of impurities in the CO2 stream can have a significant effect not only on hydraulic parameters such as pressure and temperature drops but also the density and viscosity. Its extent depends on the type, quantity and combination of impurities present. An indication of the influence of the composition of the CO2 stream was demonstrated by Seevam et al. (2008) who modelled the flow through a horizontal 406 mm diameter pipeline carrying pure CO2, CO2 with 5 % of different binary impurity combinations and the potential CO2 stream of the three capture technologies presented in Table 13.1. The inlet pressure, PI, was raised to 110 bar to ensure that the CO2–impurity combination was in the dense phase. The inlet temperature, TI, was set at 50 °C with an ambient temperature of 5 °C. The average steady-state flow rate was taken to be 72 kg/s, based on figures for pre-combustion capture of CO2 from an 860 MW combined cycle power plant (Hanne and Kvamsdal, 2005). A pipe roughness of 0.0457 mm was taken. Pumping is necessary in CO2 pipelines to ensure the CO2 remains in the dense or supercritical phase. The distance that the CO2 stream can be transported before repressurisation becomes necessary, XR, when the pressure in the pipeline approaches a 10 bar buffer above the critical pressure of the transported fluid, depends on its composition. The results for this analysis are shown in Fig. 13.3 where the CR is equal to XR for the CO2 combination normalised by XR for pure CO2, and CP and CT are the average pressure and temperature gradients over XR, also normalised with respect to the gradients for pure CO2, i.e.

CR = XR /XRCO2 ; CPG =

(PI – PXR ) XRC 2 (PI – PXRCO2 ) XCO R

[13.4]

In Fig. 13.3, coefficients for pure CO2 take a value of one. Values of CR lower than one indicate that XR is a shorter recompression distance for the combined stream than for pure CO2. Values of CPG and CTG greater than


418


3.0

CR, recompression distance coefficient CPG, pressure gradient coefficient CTG, temperature gradient coefficient

2.5

CR, CPG, CTG

2.0 1.5 1.0 0.5

Oxyfuel

Pre/comb

Post/comb

CH4

CO

SO2

O2

H 2S

Ar

N2

NO2

H2

Pure CO2

0.0

Composition

13.3 Influence of CO2 composition on recompression distance, pressure gradient and temperature gradient along the pipe.

one show larger pressure and temperature drops over the recompression distance. Some combinations cause higher pressure and temperature drops for a given pipeline length than others, especially if hydrogen or nitrogen is present. This in turn has implications for the distance between booster stations along the pipeline required to keep the pressure sufficiently high to maintain dense phase. The pipeline cost increases with the incidence of intermediate booster stations which, in any event, are not viable for subsea pipelines. A pipeline transporting CO2 streams from capture plants as specified in Table 13.1 will require more booster stations for an oxyfuel capture plant than a precombustion plant, which in turn will require more than for a post-combustion plant. Sudden temperature drops can have potential material implications such as embrittlement, and can also cause hydrate formation, which could damage the pipeline. Under certain conditions the presence of about 5 % hydrogen (Seevam et al., 2007) has been shown to exhibit the Joule Thomson effect in which, due to a sudden pressure drop (expansion), the temperature of the fluid drops below the ambient temperature, in this case due to adiabatic expansion at constant enthalpy (Oldenburg, 2006). The density is greatly affected by such pressure and temperature interactions, with temperature having a more significant effect than pressure. Density is an important parameter, especially in terms of product metering, flow assurance and the formation of undesired liquid slugs in the pipeline. It is therefore critical to specify the quality of the CO2 by being able to predict



419

how these impurities are going to affect the pipeline hydraulics (Race et al., 2007; Seevam et al., 2007).

13.3.3 Materials In dry CO2, carbon steel can be specified for compressor piping, valves, fittings, flanges and pressure vessels. In critical locations where the CO 2 cannot be dried, stainless steel is used either for the complete component, or as cladding or overlay on carbon steel (Decker et al., 1985; Recht, 1987). Cladding or overlaying significantly reduces the costs. Decker et al. (1985) indicate that some areas in the Sheep Mountain Pipeline system are coated with electroless nickel to avoid corrosion. Pipeline operators in the USA have used both stainless steel and mild steel with a high-quality plastic liner in CO2 gathering systems as these systems contain large amounts of water. The supersolvent properties of supercritical CO2 are known to be detrimental to elastomers used in valves, coatings and the gaskets and O-rings used for sealing purposes and coatings (Schremp and Roberson, 1975). At high pressures, the supercritical CO2 diffuses into the elastomers and then, when the pressure is reduced, blistering and even explosions can occur as the material decompresses. Decompression damage is the cause of most failures in dry supercritical CO2 pipelines (Najera, 1986). Many of the materials used in the oil and gas service are not suitable for CO2 service (Najera, 1986). Pipeline inspection of CO2 pipelines is rather difficult as the supercritical CO2 dissolves the non-metallic components of the in-line inspection and cleaning tools. High durometer elastomers can be used to reduce the problem but not eliminate it totally. The non-metallic material should also be able to withstand the extremely low temperatures and the potential embrittlement of the material that can occur, especially during blowdown, decompression and potential upset conditions in the pipeline.

13.3.4 Fracture Pipelines conveying gas or liquids with high vapour pressures may be vulnerable to crack propagation. The problem was recognised in the gas industry over 40 years ago. Fractures can propagate in either the fully brittle or fully ductile modes for long distances and, in theory, could propagate almost indefinitely. In the literature on CO2 pipelines, many authors have indicated that ductile fracture propagation may be an issue (King, 1982a,b; Decker et al., 1985; Maxey, 1986; Marsili and Stevick, 1990) and, indeed, the requirement to consider fracture propagation in CO2 pipelines is included in the federal regulations in the USA (49CFR195, 2010). The problem has prompted extensive research leading to the establishment


420


of a number of models that describe fracture propagation behaviour for gas pipeline systems. These models have been very successful in defining toughness requirements for pipe material, which ensure fracture arrest. To appreciate the models, it is first necessary to understand the process of fracture propagation in the ductile mode. Once a fracture is initiated in the pipeline, the fluid starts to decompress and a pressure relief front propagates in both directions from the site of the fracture. The velocity of the decompression wave is determined by thermodynamic properties and the initial pressure and temperature. Whether the fracture will propagate once initiated is dependent on whether there is sufficient driving force for propagation, i.e. whether the initial pressure is high enough to sustain a fracture. If the fracture does start to propagate along the pipeline, it will do so at a velocity that is dependent on the material properties (strength and toughness) and pipeline geometry (diameter and wall thickness). If the fracture propagation velocity is slower than the decompression velocity, and the fluid is decompressing in front of the crack, there is no driving force for propagation and the crack will arrest. This phenomenon can be modelled by various approaches, but it is instructive to consider the Battelle Two-Curve Model (TCM). The approach was developed to determine the conditions required to prevent ductile fracture propagation in natural gas pipelines, and is illustrated for a methane pipeline in Fig. 13.4. The ‘two-curves’ in the model are the gas decompression curve, representing the crack driving force, and the fracture velocity curve, representing the material resistance to propagation. The model assumes that the two curves are independent and can be considered separately. If the fracture curve and the gas decompression curve intersect or are tangent, then there exists a pressure where the pressure wave and the fracture are travelling at the same speed and the crack can propagate indefinitely under those conditions. Therefore, in Fig. 13.4, in order to prevent ductile fracture propagation, the material toughness has to be greater than that represented by the lowest curve, Cv1, which is just tangential to the decompression curve. The gas decompression curve is dependent on the thermodynamic properties of the fluid and therefore the decompression characteristics are extremely important. The decompression curve for a dense phase CO2 from a capture plant will differ in nature from that of the methane. As the CO2 decompresses from a dense phase liquid, it crosses a phase boundary after which it decompresses as a two-phase fluid. The result is a plateau in the decompression curve where the pressure remains constant but the velocity decreases. This constant pressure is termed the ‘saturation pressure’ and is defined as the pressure at which the fluid crosses the phase boundary. The fracture velocity curve tends towards a pressure below which a propagating ductile fracture cannot be sustained as the velocity decreases.



421

15 C v3 C v2

C v1

Increasing toughness

p (MPa)

10 Fracture

Gas decompression curve

5

Gas decompression 0

0

50

100

150 v (m/s)

200

250

300

13.4 Battelle TCM for methane illustrating the effect of toughness.

This pressure is termed the arrest pressure which can be estimated at high levels of toughness from Equation 13.5 (Maxey, 1986),

Pa =

st 3.33 R

[13.5]

where s is the flow stress, t is the pipe wall thickness and R is the nominal outside radius, all in SI units. It was suggested by Maxey (1986) and Cosham and Eiber (2008) that an estimate of the toughness requirement for arresting a ductile fracture in a CO2 pipeline can be obtained from the arrest pressure and the saturation pressure (calculated from phase diagram information). In order to arrest a ductile fracture, therefore, the arrest pressure must be greater than the saturation pressure, i.e. either the arrest pressure must be raised or the saturation pressure must be lowered. Maxey observed that the saturation pressure can be lowered by lowering the operating temperature or by removing impurities with lower critical temperatures than CO2. In particular, hydrogen has been shown to have a detrimental effect in terms of fracture propagation (King, 1982a). Conversely, it has been shown that the arrest pressure can be raised by increasing the wall thickness, increasing the toughness, decreasing the pipe diameter or increasing the yield strength (King, 1982b). In the gas industry, there has been less work conducted on fracture propagation in offshore pipelines than for onshore pipelines. The modelling is more complicated in this case because of the interaction of the escaping fluid with the water, which reduces the hoop stress in the pipe wall. It is


422


considered that, for gas pipelines, the methods for calculating fracture arrest conditions are conservative when applied to offshore pipelines (Maxey, 1986). However, this has not been confirmed for CO2 pipelines offshore.

13.3.5 Corrosion and hydrate formation The presence of water in CO2 may cause corrosion of the pipeline steels, lead to hydrate formation and/or result in two-phase flow in the pipeline. It is, therefore, important to understand the conditions under which these phenomena could occur and to control the level of water, temperature and pressure in the pipeline to prevent damage. Long-distance, high-pressure CO2 pipelines are constructed from pipeline-grade carbon steel as this is, realistically, the only economical choice. Due to the high operating pressures CO2 pipelines are typically constructed with API 5L Grades X65 or X70 (Mohitpour et al., 2003) which are essentially non-corrosive in pure CO2. However, in the presence of water, highly corrosive carbonic acid is formed, and it has been reported that carbon steel can corrode at rates of more than 10 mm/year in wet pure CO2 (Seiersten and Kongshaug, 2005). The level to which the CO2 stream has to be dried to avoid this is dependent on the water solubility limit which, in turn, is dependent on the pressure, temperature and the level of impurities. At a constant pressure, the solubility of water in both the liquid and gas phase increases with increasing temperature for pure CO2 (King, 1982a; Austegard et al., 2006). The significance of this behaviour of CO2 is that for pipelines operating at low temperatures and pressures, the water content would have to be more tightly controlled. Consequently, pipelines operating in the gas phase require more stringent allowable water content specifications than those operating in the liquid or dense phase. The solubility of water in pure CO2 is well understood and documented as a function of temperature and pressure. However, the solubility of water in CO2 containing impurities is less well researched, although there are some limited and sometimes conflicting data available. A recent experimental study by Seiersten and Kongshaug (2005) on the CO2–H2O–CH4 system concluded that the water solubility in the liquid and supercritical phase decreases with increasing CH4 content and therefore the risk of forming free water increases. This conclusion is in agreement with the results and calculations quoted by Heggum et al. (2005) and Austegaard et al. (2006). In contrast, a study conducted as part of the Dynamis project (de Visser et al., 2008) showed that for the CO2–H2O–H2S system at 4 °C and 100 bar, adding H2S increased the solubility of water. However, it was shown that at the low levels of H2S expected in the CO2 stream, around 200 ppm, this effect was negligible. It should be appreciated that this study was based on calculations and no experimental data were generated. The implication



423

of these observations is that the drying specification for CO2 containing impurities will be dependent on the type and quantity of impurities present. Therefore experiments are required to determine the solubility of water in CO2 containing the expected impurities from power station capture. A hydrate is a solid crystalline compound formed by the addition of water, the ‘host’ molecule, into a hydrate former, the guest molecule such as methane, CO2 or H2S. The hydrate is not a chemical compound as no chemical bond is formed between the host and the guest molecule. Hydrate formation in pipelines causes problems because hydrates are solid compounds with similar properties to ice; consequently, they can block the pipeline and plug or foul other equipment such as heat exchangers. In order for hydrates to form in a CO2 pipeline, there must be the required combination of pressure and temperature and a sufficient amount of water. In CO2 pipelines, it would be possible for hydrates to form at around 10–11 °C (Wallace, 1985; Fradet et al., 2007). Although Wallace (1985) states that hydrates will not form in the absence of free water, Hendriks et al. (2007) and Carroll (1999) assert that free water is not necessarily required for hydrate formation and that hydrates can form with dissolved water. However, the quantity of dissolved water in the CO2 will be determined by the requirement to prevent corrosion, and it is considered that the amount of hydrate that could be formed with these levels of dissolved water will not be sufficient to cause pipeline operational problems (de Visser and Hendriks, 2007). It can certainly be concluded that the risk of hydrate formation increases if free water is present.

13.4

Transport of carbon dioxide (CO2) by ship

The issues relating to the transport of CO2 by ship have been introduced in Chapter 12, and a good overview of the topic may be gained from references cited there, such as Barrio et al. (2005) and Aspelund et al. (2005, 2006). The most efficient way to transport CO2 is in supercritical or dense phase because it has a higher density in these states. The practice of transporting liquefied gas, such as LPG and LNG, is well established and they can be carried in fully refrigerated, semi-pressurised or fully pressurised ships. In fully refrigerated LPG and LNG ships, the cargo is maintained in liquid phase at atmospheric pressure solely by refrigeration. However, CO2 does not exist as a liquid under these conditions and so it is transported in semi-pressurised ships. Semi-pressurised ships operate under conditions approaching the triple point, where the density is highest, while the fully pressurised ships operate closer to the critical point. For a fixed volume vessel, almost twice as much CO2 can be transported at low pressures near the triple point than at high pressures near the critical point (Aspelund et al., 2005). There are a number of small-scale semi-pressurised CO2 carriers in operation


424


shipping liquid CO2 between distribution terminals in countries bordering the North Sea. LPG is also carried by semi-pressurised ships, but generally at slightly lower pressures, lower densities and higher temperatures than CO2 carriers. The latest LNG carriers reach more than 200 000 m3, a capacity that could accommodate 230 kt of liquid CO2 (IEA, 2004b; IPCC, 2005). The design and construction of large-scale semi-pressurised CO2 carriers operating at pressures near the triple point could follow well-established criteria for the commercial construction of LPG carriers (Barrio et al., 2005). Transportation of CO2 by ship requires a supporting infrastructure with a number of components. CO2 delivered from source to an onshore terminal must undergo a liquefaction process including the removal of water and volatile gases as described in Chapter 12. It is then stored in semi-pressurised tanks in close proximity to the quay where it will be loaded into the ship using specialised facilities. The CO2 may then be transferred to another collection terminal, offloaded and then transported by pipeline to a sink offshore, or it may be unloaded directly at sea prior to injection. The costs of CO2 transport by ship comprise many elements. As well as the investment in the ships themselves, they include investment in the loading and unloading facilities and in intermediate storage and liquefaction plant. Also to be taken into account are operational costs such as fuel costs, harbour fees and manning. For these reasons, the costs are scenario dependent and difficult to generalise. Typically, ship transport only becomes competitive with pipeline transport over relatively large distances. The IPCC Special Report on CCS quotes a breakeven distance of 1000 km for transporting 6 Mt of CO2 per year by pipeline and by ship. However, it is emphasised that the crossover is not simply a matter of distance alone but involves many other factors in addition to those already mentioned. Actual costs quickly become outdated, but Barrio et al. estimated a total cost, including all the costs in the transport chain, of 150–200 NOK/t for volumes larger than 2 Mt/year and distances limited to the North Sea (Barrio et al., 2005). The subject of the cost of operating transport systems using CO2 carriers of up to 50 000 t has been reported by the IEAGHG R&D programme (IEA, 2004b). The great advantage of ships over pipelines is their flexibility, and it is possible to envisage circumstances in which they would be the preferred mode of transport. Examples include: the early stages of the infrastructure development before pipelines are in place; transport to sinks receiving CO2 only over a short period where pipelines are economically and/or technically not viable; collection from isolated sources. In the longer term, integrated infrastructures serving a number of countries, such as those bordering the North Sea, may be developed and transport by ship may well have a role to play in such systems. In the first phase of CCS development in the CO2 transport, infrastructure will be dominated by pipelines.



13.5

425

Transport economics

The subject of transport economics has already been covered Chapter 2, the principal points of which may be summarised as follows. Pipeline costs are dependent on the route taken (from a demographic as well as a geographical/topographical perspective) and its functionality (flowrate, the need for booster stations and physical characteristics, such as length, diameter and roughness). There is a trade-off between lower pipeline costs due to reduced pipe sizes achieved through using booster stations and the resulting added costs of compression. Booster stations are not viable offshore and so the pipelines may have to be larger than onshore to maintain the required pressures. This may be offset, to a certain degree, because of a gain in gravity head due to the drop in elevation from the shore to the pipe outlet. Pipeline costs include capital costs (engineering, installation and materials) and operations and maintenance costs (monitoring, inspection and repair). These may be increased significantly by added use of booster stations. The capital costs per metre of pipelines are weakly dependent on pipeline length and increase approximately linearly with diameter, as expressed by Equation 2.4 in Chapter 2. The operation and maintenance costs are of the order of up to 4 % of the capital costs onshore. Offshore, the costs may be expected to be of the order of two and a half times those of pipelines onshore. The discussion of the economics of CO2 transport by ship in Section 13.4 above and a comparison of ship and pipeline transport is dealt with in more detail in Chapter 2.

13.6

Large-scale transport infrastructure

There is over 30 years experience with extensive infrastructure used for transporting natural and anthropogenic sources of CO2 for EOR, mainly in the USA. In fact, it is estimated that there are over 2500 km of CO2 pipelines in the world. In order to learn and benefit from existing experience, it is important to understand the key differences between the design, operation and location of currently operating pipelines and networks and those proposed for CCS for climate change mitigation in the UK and other countries. The fundamental differences between the two are that, for the latter, CO2 will be collected from a relatively large number of anthropogenic sources over a potentially wide geographical distribution to be transported to sinks that may well be subsea. For example, in the UK the major sources are mostly coal/gas/oil-fired power stations and some large industrial concerns such as steel and cement works. There are over 250 sources of this type emitting individually in excess of 1 Mt/year each (BGS, 2004). The 20 largest of these represent around 8 % of the total number emitting around 50 % of the cumulative emissions, as shown in Fig. 13.5.


426


Over 60 (24 %) of the sources are required to achieve 80 % of the cumulative emissions after which the law of diminishing returns applies. Ultimately, a large-scale CO2 infrastructure must match a sufficient number of sources to sinks to satisfy the requirements of the UK emissions reduction policies. The development of the CO2 transport infrastructure in the UK will depend on government and EU policy and the financial, social and political drivers. It will involve incremental development and is likely to take place in three phases. The first phase of the infrastructure development will revolve around demonstration projects (Downie et al., 2007). The second phase of development will be a limited commercial enterprise, providing the necessary economic incentives have been set in place, in which the lessons learned from the demonstration plants are put into practice on a wider scale. The infrastructure is likely to be more complex and allow the beginnings of the main spurs of future networks to come into being. It is logical to expect this phase to involve the largest emitters connected to the most economic and technically viable sinks. The actual manner in which this development takes place depends on a variety of complex factors relating to the economic climate, the supply of CO2 and the storage available, and the strategies put in place to deal with them. The third phase of the development of the CO2 transport infrastructure will involve the development of increasingly complicated networks as the CO2 is collected from an increasing number of sources.

100 200

90

100

80 50 40

Emissions (%)

70

30

60 50

20

40 10

30 20 10 0 0

20

40 60 Sources (%)

80

13.5 Cumulative plot of emissions with sources.


100


427

The earlier phases of the development will consist of point-to-point pipelines linking sources to sinks; in the later stages, larger integrated transport systems will evolve, possibly involving a number of countries bordering on the North Sea. This final phase will ultimately be limited by the decline in use of fossil fuels for power generation. The rates at which these pipeline networks develop have a significant impact on the reduction of CO2 emissions to meet climate change targets. Other potential issues include logistics such as the construction time, labour requirements and obtaining pipeline material.

13.6.1 Pipelines offshore In the USA, all the CO2 pipeline networks are onshore and their efficient and economic operation is simplified since it is relatively straightforward to recompress the CO2 if the pressure drops below the critical pressure. Obviously, offshore repressurisation along the pipeline is to be avoided and therefore pipelines in the UK will have to be designed such that the inlet pressure is high enough to prevent the pressure from decreasing below the critical pressure before the delivery point. This requirement may place some constraints on the ability to reuse existing infrastructure. Typical maximum design pressures for long-distance gas pipelines range from 50–150 bar. The critical pressure and temperature for methane, the main component of natural gas, are 46.2 bar and –83 oC, respectively, whereas those CO2 of are 73.8 bar and 31 °C. Injection pressures may be in the region of 85–200 bar depending on circumstances. Pressures well in excess of the critical pressure, depending on the distances involved, are required as inlet conditions if the CO2 is to be transported to the delivery point without recourse to intermediate repressurisation. In addition, as discussed previously, impurities in the CO2 can raise or lower the critical pressure. It is, therefore, crucial to be able to predict the phase behaviour for CO2 containing impurities to ensure that offshore repressurisation and two-phase flow in the pipeline is avoided. It is emphasised that the currently operating pipelines in the USA may well not contain the same combinations of impurities as will be present in the CO2 captured from power plant, and the benefit derived from operational experience in this area could, therefore, be limited.

13.6.2 Regulations There is no experience with pipeline transportation of dense phase CO2 in the UK. Consequently, the design codes (BSI, 2004) and Pipeline Safety Regulation (PSR, 1996) do not contain any classification for dense phase CO2 that would allow pipelines onshore or offshore to be designed and routed. In the USA, dense phase CO2 pipelines have been classified as hazardous liquid pipelines, which are regulated under the Department of Transportation Code


428


of Federal Regulations 49CFR195. Most operators appear from the literature to have designed their pipelines conservatively using the ASME B31.8 code (ASME, 2004) for gas pipelines (Kantar and Connor, 1995; Decker et al., 1985; McCullough, 1986) as this code is more restrictive than the hazardous liquid design code and also takes into account population density in the determination of the maximum allowable stress in the pipeline. With the exception of the Snövhit pipeline, all of the current operating CO2 pipelines are located onshore and most are routed through sparsely populated areas. Therefore, the consequence of an accident is likely to be relatively small because the CO2 will dissipate before affecting the local population. However, if networks are to be developed onshore in the UK for the collection of CO2 from disparate sources, this will require the routing of pipelines close to population centres. In the UK pipeline design codes (BSI, 2004), the minimum distance between the pipeline and occupied buildings (building proximity distance – BPD) is defined by the substance being transported. The preferred method for calculating BPDs is to use risk contours and to route the pipeline so that the BPD is no less than the distance to the 10 chance per million contour (i.e. a person at this location has a less than 10 chances in a million of receiving a dangerous dose of CO2 in the event of a pipeline failure). It should be noted that risk contours are not available for supercritical CO2 as the dispersion characteristics have not been determined for a release from a dense phase CO2 pipeline nor, as mentioned above, has supercritical CO2 been classified. Therefore, in order for pipeline networks to be designed in the UK, regulations need to be put in place to classify supercritical CO2 and so to allow safe pipeline design both onshore and offshore.

13.7

Discussion

Onshore source to sink transport of CO2 for EOR is well understood and the knowledge gaps relating to CCS for climate change mitigation are becoming increasingly defined. Less effort has been expended on the problems that may be encountered in the development of a large-scale infrastructure incorporating multiple anthropogenic sources using a variety of capture technologies. Providing health and safety and environmental criteria are met, it is likely to be advantageous to accept CO2 quality specifications for transportation with the widest scope possible. The transport media, pipelines or ships, are not the limiting constraint on the CO2 quality specification other than with respect to the water content. The environmental impact of the CO2 stream on the storage medium may be a limiting factor, but the transport medium should not be a constraint if there is the requisite knowledge to enable its design to maintain its structural integrity. In this respect, further work is required to establish the thermodynamic properties of CO2 containing combinations



429

of components that are output from the various capture technologies. In addition, this knowledge needs to be incorporated into numerical models that can be used for design calculations. The models currently in use are often modified from those used for the design of gas pipelines and are mostly set up to deal with pure CO2. Related knowledge gaps exist in the context of material developments required for pipelines to be designed with sufficient toughness to resist fracture when carrying dense phase CO2 with impurities. In the event of a catastrophic release of CO2 from such a pipeline, the means of modelling its dispersion with a high degree of certainty are not currently available and so it is not possible to undertake a rigorous risk analysis since the consequences have not been reliably predicted. The problems arising from these knowledge gaps are compounded when moving from simple source to sink pipelines to networks linking industrial sources subject to operational imperatives. In these circumstances, the design of the infrastructure should encompass planned operational procedures and unintended incidences, requiring the use of reliable transient computational models. In addition to the technical engineering requirements, there is a need for resources to allow rational strategic planning and, as far as possible, optimal solutions for safe, economic networks with minimal environmental impact. To enable its implementation, regulatory frameworks and financial incentive schemes need to be developed.

13.8

Future trends and future work

Future trends in the transport of CO2 for CCS will be determined by the need to develop large-scale transport infrastructures starting locally with national initiatives and expanding from nothing to integrated systems that may well be transnational in their scope. The future work will continue to focus on pipelines and ships, and will entail the development of enabling technologies that fill the knowledge gaps already identified and those to be discovered as the process unfolds.

13.9

Conclusions

The transport of CO2 is an area that has not received the same amount of attention as other aspects of CO2 capture and storage in the UK and beyond. This is because it has been successfully undertaken in the USA for many years, and in the UK there appears to be an extensive existing infrastructure that could be readily adapted for re-use. However, in the USA the pipelines have been custom built and designed for purpose, largely to transport natural (relatively pure) CO2 overland through mostly sparsely populated areas. In the UK, there is little experience in designing CO2 pipelines, and the design criteria and constraints are not just significantly different from those of the


430


USA but some of them are actually quite new. In the UK, the pipelines will be required to carry anthropogenic CO2 both overland and subsea through areas that are often in close proximity to population centres. The impurities in the CO2 stream are different, and will occur in different combinations to those that that have been transported previously, and their properties and their effect on hydraulic performance have yet to be determined. There is virtually no experience of subsea CO2 pipelines or of the re-use of existing pipelines for its transport. There needs to be a detailed rigorous technical audit of existing infrastructure to determine how much of the available system is actually viable for CO2 transport. There are a number of technical issues to be resolved in simply transporting anthropogenic CO2 in subsea pipelines and injecting it into geological formations for storage. There are further complications to be tackled when developing a more general infrastructure collecting CO2 from a wide variety of different types of source. The task of implementing a complete CO2 transport infrastructure to collect enough of the emissions, along with other mitigation measures, for the UK to meet its emissions target is a challenging one. These issues must be addressed if CO2 transportation is to be implemented safely, economically and efficiently and in time for CO2 capture and storage to perform its function as a bridging technology to a carbon neutral future.

13.10 Sources of further information and advice The following two publications provide good general background to CO2 transport. ∑

Doctor, R., Palmer, A., Coleman, C., Davison, J., Hendriks, C., Kaarstad, O. and Ozaki, M. (2005) Transport of CO2, Chapter 4, IPCC Special Report on Carbon Dioxide Capture and Storage, prepared by Working Group III of the Intergovernmental Panel on Climate Change, edited by Metz, B., Davidson, O., de Coninck, H.C., Loos, M. and Meyer, L. A., Cambridge University Press, Cambridge, UK and New York. This reference discusses all aspects of CO2 transport including a section on the economics of different transport modes. ∑ Mohitpour, M., Golshan, H. and Murray, A (2003) Pipeline Design and Construction: A Practical Approach, American Society of Mechanical Engineers, New York. This is a general pipeline design reference book, but it also has a dedicated chapter on CO2 pipelines.

13.11 Acknowledgements Much of the work conducted by the authors and included in this chapter has been conducted as part of the UK Carbon Capture and Storage Consortium © Woodhead Publishing Limited, 2010


431

(UKCCSC) sponsored by the Natural and Environment Research Council (NERC), whose support is acknowledged with thanks. Thanks are also due to the British Geological Survey for providing a web enabled GIS of sources and sinks along with existing infrastructure information used originally in Downie et al. (2007) and referred to in this chapter.

13.12 References 49CFR195 (2010) Transportation of Hazardous Liquids by Pipeline, Revision 02/10, Current through amendment 195–93c, Department of Transportation, 1 February 2010. AGA (1990), AGA 3.1: Orifice Metering of Natural Gas and Other Related Hydrocarbon Fluids, Part 1: General Equations and Uncertainty Guidelines, 3rd edn, American Gas Association and American Petroleum Institute, Washington DC. Anheden M, Andersson A, Bernstone C, Eriksson S, Yan Liljemark S and Wall C (2005) CO2 Quality requirement for a system with CO2 capture, transport and storage, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 2, 2559–2563. ASME (2004) B31.8 Gas Transmission and Distribution Piping Systems, American Society of Mechanical Engineers, New York. Aspelund A, Sandvik TE, Krogstad H and De Koeijer G (2005) Liquefaction of captured CO2 for ship-based transport, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 2, 2545–2549. Aspelund A, Mølnvik MJ and De Koeijer G (2006) Ship transport of CO2 – Technical solutions and analysis of costs, energy utilization, exergy efficiency and CO2 emissions, Journal of Chemical Engineering Research and Design, 84–A9, 847–855. Austegard AE, Solbraa G, de Koeijer MJ and Mølnvik (2006) Thermodynamic models for calculating mutual solubilities in H2O–CO2–CH4 mixtures, Chemical Engineering Research and Design (ChERD), Part A, 2005, Special Issue: Carbon Capture and Storage, 84 (A9) 781–794. Barrio M, Aspelund A, Weydahl T, Sandvik TE, Wongraven LR, Krogstad H, Henningsen R, Molnvik M and Eide SI (2005) Ship-based transport of CO2, in Wilson M, Morris T, Gale J and Thambimuthu K (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, IEA GHG, Cheltenham, UK, Vol. 2, 1655–1660. Beggs HD and Brill JP (1973) A study of two phase flow in inclined pipes, Journal of Petroleum Technology, 25, 607–617. BGS (2004) GIS file provided courtesy of the British Geological Survey, UK. document website: http://www.bgs.ac.uk/science/co2/ukco2.html (accessed February 2010). Brill JP and Mukherjee H (1999) Multiphase Flow in Wells, SPE Monograph, Society of Petroleum Engineers, Richardson, TX. Bierlein JA and Kay WB (1953) Phase equilibrium properties of system carbon dioxide – hydrogen sulphide, Journal of Industrial and Engineering Chemistry, 45(3), 618–624. BSI (2004) PD 8010-1 Code of Practice for Pipelines, Part 1 – Steel pipelines on land, British Standards Institute, London UK.


432


Carroll JJ (1999) An Introduction to Gas Hydrates, available at: http://www.telusplanet. net/public/jcarroll/HYDR.HTM (accessed December 2009). Cosham A and Eiber RJ (2008) Fracture control in carbon dioxide pipelines – the effect of impurities, Proceedings 7th International Pipeline Conference, 29 September – October 3, Calgary, Canada, IPC 2008-64346. Decker LC, Checak JR, Heath WH and Warren RN (1985) CO2 pipeline design, construction and operation, in Brooks GF and Martinez G (eds) Facilities, Pipelines, and Measurements: A Workbook for Engineers, Proceedings 41st Petroleum Mechanical Engineering Workshop and Conference, American Society of Mechanical Engineering New York. de Visser E and Hendriks C (2007) Towards Hydrogen and Electricity Production with Carbon Dioxide Capture and Storage – D3.1.3 DYNAMIS CO2 quality recommendations, DYNAMIS, Project No 019672. de Visser E, Hendriks C, Barrio M, Mølnvik MJ, de Koeijer G, Lijemark S and Le Gallo Y (2008) Dynamis CO2 quality recommendations, International Journal of Greenhouse Gas Control, 2, 478–484. Doctor R, Palmer A, Coleman D, Davison J, Hendriks C, Kaarstad O and Ozaki M (2005) Transport of CO2, in Metz B, Davidson O, de Coninck H, Loos M and Meyer L (eds), IPCC Special Report on Carbon Dioxide Capture and Storage, Cambridge University Press, Cambridge, UK, 173–194. Downie MJ, Race JM and Seevam PN (2007) Transport of CO2 for carbon capture and storage in the UK, Proceedings Offshore Europe 2007, 4–7 September, Aberdeen, UK, SPE 109060. Farris C (1983) Unusual design factors for supercritical CO2 pipeline, Energy Progress, 3, 150–158. Fradet A, Saysset S, Odru P, Broutin P, Ruer J and Bonnissel M (2007) Technical and economic assessment of CO2 transportation for CCS purposes, Journal of Pipeline Engineering, 6 (3), 173–180. Gill TE (1985) Canyon Reef Carriers, Inc. CO2 pipeline: description and 12 years of operation, 8th Annual Pipeline Engineering Symposium, Dallas, TX, ASME. Hanne M and Kvamsdal TM (2005) Tjeldbergodden power/methanol – CO2 reduction efforts. SP2: CO2 Capture and Transport, Technical Report TRF6062 SINTEF Energy Research, Trondheim, Norway. Heggum G, Weydahl T, Mo R, Mølnvik M and Austergaard A (2005) CO2 conditioning and transportation, in Thomas DC and Benson SM (eds), Carbon Dioxide Capture for Storage in Deep Geologic Formations, Elsevier, Oxford, UK, 925–936. Hein MA (1985) Rigorous and approximate methods for CO2 pipeline analysis, in Brooks GF and Martinez G (eds), Facilities, Pipelines and Measurements: A Workbook for Engineers, Proceedings 41st Petroleum Mechanical Engineering Workshop and Conference, American Society of Mechanical Engineers, New York, 71–75. Hein MA (1986) Pipeline design model addresses CO2’s challenging behaviour, Oil and Gas Journal, 71–75. Hendriks C, Hagedoorn S and Warmenhoven H (2007) Transportation of Carbon Dioxide and Organisational Issues of CCS in the Netherlands, Report prepared by Ecofys for EnergieNed, the Ministry of Economic Affairs and the Ministry of Housing, Spatial Planning and the Environment, March 2007. IEA (2004a) Impact of impurities on CO2 capture, transport and storage, Report no. PH4/32, Study conducted by SNC Lavalin, August 2004, IEA Greenhouses Gas RRD Programme, Cheltenham, UK.



433

IEA (2004b) Ship transport of CO2, Report Number PH4/30, IEA Greenhouse Gas RRD Programme, Cheltenham, UK. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz B, Davidson O, de Coninck HC, Loos M and Meyer LA (eds), Cambridge University Press, Cambridge, UK. Kantar K and Connor TE (1995) Turkey’s largest oil field poised for CO2 immiscible EOR project, Oil and Gas Journal, 82 (48), 95. King GG (1982a) Here are key design considerations for CO2 pipelines, Oil and Gas Journal, 80(39), 219. King GG (1982b) CO2 pipeline design – fractures, pipe sizing probed for high operating pressure CO2 pipelines, Oil and Gas Journal, 80(40), 100. Li H and Yan J (2007) Comparative study of equations of state for predicting phase equilibrium and volume properties of CO2 and CO2 mixtures in Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Elsevier, Oxford, UK, CD-ROM. Marsden GW and Wolter D (1986) Mass custody transfer measurement technique, Oil and Gas Journal, 84, 74–78. Marsili DL and Stevick GR (1990) Reducing the risk of ductile fracture on the canyon reef carriers CO2 pipeline, Proceedings SPE Annual Technical Conference and Exhibition, 23–26 September, New Orleans, LA, 20646 – MS. Maxey WA (1986) Long shear fractures in CO2 lines controlled by regulating saturation arrest pressures, Oil and Gas Journal, 84 (31), 44. McCollough DE (1986) The Central Basin Pipeline: a CO2 system in West Texas, Energy Progress, 6 (4), 230–234. Mohitpour M, Golshan H and Murray A (2003) Pipeline Design and Construction: A Practical Approach, American Society of Mechanical Engineers, New York. Moody LF (1944) Friction Factors for Pipe Flow, Trans. ASME, 66, 671. Najera G (1986) Maintainenance techniques proven on CO2 line, Oil and Gas Journal, 84, 55–57. NIST database, Reference Fluid Thermodynamic and Transport Properties Database (REFPROP), NIST, Gaithersburg, MD, available from: http://www.nist.gov/data/ nist23.htm (accessed January 2010). Oldenburg CM (2006) Joule Thompson cooling due to CO2 injection into natural gas reservoirs, TOUGH Symposium, 15–17 May, Lawrence Berkeley Laboratory, Berkeley, CA (CD-ROM). Oosterkamp T and Ramsen J (2008) State of the Art Review of CO2 Pipeline Transportation with Relevance to Offshore Pipelines, Report No. POL-0-2007-138-A, Polytec, Haugesund, Norway. PSR (1996) Pipeline Safety Regulations, Statutory Instrument 1996 No. 825, The Stationary Office, London, UK. Race JM Seevam PN and Downie MJ (2007) Challenges for offshore transport of anthropogenic carbon dioxide, 2007 Proceedings of the International Conference on Offshore Mechanics and Arctic Engineering, American Society of Mechanical Engineers, New York. Recht DL (1987) Carbon dioxide pipeline design considerations, in Seidens, EJ (ed.), ASME Pipeline Engineering Symposium, American Society of Mechanical Engineers, New York, 155–159. Schremp FW and Roberson GR (1975) Effect of supercritical carbon dioxide (CO2) on construction materials, Society of Petroleum Engineers Journal, 15, 227–233. © Woodhead Publishing Limited, 2010

434


Schroeder DW (2001) A Tutorial on Pipe Flow Equations, Pipeline Simulation Interest Group (PSIG) Report 2001–12, available at http://www.psig.org/papers/2000/0112. pdy (accessed January 2010). Seevam P (2010) PhD Thesis, School of Marine Science and Technology, Newcastle University. Seevam PN et al. (2007) Carbon dioxide pipelines for sequestration in the UK, Transmission of CO2, H2 and Biogas: Exploring New Uses for Natural Gas Pipelines Conference, 30–31st May, Amsterdam, the Netherlands. Seevam PN, Race JM, Downie MJ and Hopkins P (2008) Transporting the next generation of CO2 for carbon, capture and storage: the impact of impurities on supercritical CO2 pipelines, Proceedings of the IPC2008 7th International Pipeline Conference, September 29–October 3, Calgary, Canada, IPC 2008–64063. Seiersten M and Kongshaug KO (2005) Materials selection for capture, compression, transport and injection of CO2, in Thomas DC and Benson SM (eds), Carbon Dioxide Capture for Storage in Deep Geologic Formations, Elsevier, Oxford, UK, 2, 937–953. Wallace CB (1985) Drying supercritical CO2 demands care, Oil and Gas Journal, 83, 98–104. White V, Allam R and Miller E (2007) Purification of oxyfuel derived CO2 for sequestration or EOR, in Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Elsevier, Oxford, UK, CD-ROM. Zhang ZX, Wang GX, Massarotto P and Rudolph V (2006) Optimization of pipeline transport for CO2 sequestration, Energy Conversion and Management, 47(6), 702–715.


14


S. S o l o m o n and T. F l a c h, DNV – Research and Innovation, Norway Abstract: This chapter discusses carbon dioxide (CO2) injection technology in underground geological formations. Underground fluid injection activities across a range of industrial applications are briefly presented with emphasis on natural gas storage and deep injection of acid gas as background information. A review of the technology surrounding the injection of fluids in general is illustrated together with the major reservoir parameters which control the injectivity of CO2. The injection of CO2 in different storage formations is also presented, including the hydro-mechanical impact of CO2 injection. Other key topics addressed include CO2 injection field operations, well integrity and monitoring technologies for injection well integrity. Key words: underground fluid injection, injection well technologies, CO 2 injectivity, injection well integrity, monitoring technologies.

14.1

Introduction

This chapter comprises nine sections. After a brief introduction in this section, underground fluid injection activities across a range of industrial activities are presented in Section 14.2. Section 14.3 consists of knowledge and experience of the underground injection of other fluids such as natural gas storage and deep injection of acid gas and the lessons to be learned for CO2 injection. A review of the technology surrounding the injection of fluids in general is given in Section 14.4. The major reservoir parameters which control the injectivity of CO2, formation permeability, injectivity loss, wellbore design including injection well pressure and reservoir constraints, are presented in Section 14.5. These parameters are essential for understanding geological injection science and technology. Geological storage is presented in greater detail in Chapters 2–5 in Volume 2. Section 14.6 of this chapter focuses on the injection of CO2 in different storage formations (e.g. saline formations, oil and gas fields and coal beds). The effects of CO2 injection are reviewed in the context of the coupled hydro-mechanical process in this section. Section 14.7 addresses CO2 injection field operations in both onshore and/ or offshore conditions. Section 14.8 covers a diverse range of other issues concerning the injection of CO2, including well integrity together with induced seismicity/hydro-fracturing, cement degradation, debonding and corrosion and their impacts on injection well integrity. Technologies for 435 © Woodhead Publishing Limited, 2010

436


monitoring injection well integrity, including standard detection techniques in the oil and gas industries to assess well integrity issues, are dealt with in Section 14.9. In the last section, the chapter focuses on future trends of underground CO2 injection, including the main challenges with the emphasis on well integrity. Sources of further information and references are presented in the last sections.

14.2

Underground fluid injection

Disposal of liquids (industrial waste) into underground formations through injection wells was started as early as the 1930s, e.g. by the US petroleum industry (Clark et al., 2005). Injection wells have been used for fluids associated with industrial facilities. Disposal of liquid radioactive waste started in the early 1960s in Russia in experimental reservoirs at shallow depths (270–386 m). After this experimental shallow-depth injection project, deep-well injection was started and further developed. The experience gained from the deep-well injection of radioactive waste was then used at an injection site for non-radioactive waste at the Chepetsk Mechanical Plant (Glazov) in 1992 at a depth of 1435–1600 m in a limestone formation (Rybalchenko et al., 2005). Positive results from these experimental activities together with additional investigations led to the creation of a deep-well injection facility at industrial scale. Typical examples in Russia include the Kirovo-Chepetsk Chemical combine at 1260–1440 m depth of injection in limestone formations and the Kalinin Atomic Power Plant at 1200–1400 m depth of injection in sand formations (Rybalchenko et al., 2005). The primary objective of deep-well disposal is to permanently isolate injected fluids from the biosphere. For the purpose of regulation, different categories are used to classify injection wells. For instance, in the USA Class I wells are used to inject hazardous and non-hazardous fluids below any underground sources of drinking water (USDW). Class II wells inject brine fluids associated with oil and gas production. Class III wells pertain to in situ mining wells and Class IV wells (banned except for remediation) handle disposal of hazardous liquids into or above USDWs. Class V wells relate to geothermal and other wells that do not fall into the previous categories. Similarly, in the case of deep radioactive injection wells, three categories are used to classify the wells according to the type of waste: low level, medium level and high level. The construction and operating requirements for each category are different depending on the purpose, for instance whether the well is an injection well or a monitoring well.



14.3

437

Analogues for carbon dioxide (CO2) storage and best practices from other sectors

Typical industrial analogues for CO2 storage include natural gas storage, acid gas injection and liquid waste injection. Underground natural gas storage projects that offer experience relevant to CO2 storage have operated successfully for almost 100 years and in many parts of the world (Lippmann and Benson, 2003; Perry, 2005). These projects provide for peak loads and balance seasonal fluctuations in gas supply and demand. The Berlin Natural Gas Storage Project is an example of this. The facility lies to the east of the North German Basin, which is part of a complex basin structure extending from the Netherlands to Poland with complicated tectonic setting and heterogeneous reservoir lithologies. The sandstone storage horizons are at approximately 800 m below sea level. The gas storage layers are sealed by claystone, anhydrite and halite, approximately 200 m thick. The facility has a capacity of 1085 million m³. A storage production rate of 450 000 m³ h–1 can be achieved with the existing storage wells and surface facilities. Twelve wells drilled at three sites are available for natural gas storage operation with varying well completions and well designs to allow for wireline logging. The geological and engineering aspects and scale of the facility make it a useful analogue for a small CO2 storage project. The majority of gas storage projects are in depleted oil and gas reservoirs and saline formations, although salt caverns have also been used extensively. A number of factors are critical to the success of these projects, including a suitable and adequately characterized site (permeability, thickness and extent of storage reservoir, tightness of caprock, geological structure, lithology, etc.). Injection wells must be properly designed, installed, monitored and maintained, and abandoned wells in and near the project must be located and plugged. Proper operation procedures require avoiding over-pressurizing the storage reservoir (injecting gas at a pressure that is well in excess of the formation fracture pressure), such that the stored natural gas is not lost due to leakage out of the target storage reservoir. Acid gas injection operations represent a commercial analogue for some aspects of geological CO2 storage. Acid gas is a mixture of H2S (hydrogen sulphide) and CO2, with minor amounts of hydrocarbon gases; it is a byproduct from petroleum production or processing. Although the purpose of the acid gas injection operations is to dispose of H2S because of its toxic nature, significant quantities of CO2 are injected at the same time because it is uneconomic to separate the two gases. Acid gas injection in Western Canada occurs over a wide range of formation and reservoir types, acid gas compositions and operating conditions. Injection takes place in deep saline formations at 27 sites, into depleted oil and/or gas reservoirs at 19 sites and into the underlying water leg of depleted oil and gas reservoirs at four sites.


438


Carbonates form the reservoir at 29 sites and quartz-rich sandstones dominate at the remaining 21. In most cases, shale constitutes the overlying confining unit (cap rock), with the remainder of the injection zones being confined by tight limestones, evaporites and anhydrites (IPCC, 2005). Carbon dioxide often represents the largest component of the injected acid gas stream, in many cases, 14–98 % of the total volume. A total of 2.5 Mega tonne (Mt) CO2 and 2 Mt H2S had been injected in Western Canada by the end of 2003, at rates of 840–500 720 m3/day per site, with an aggregate injection rate in 2003 of 0.45 Mt CO2 yr–1 and 0.55 Mt H2S yr–1, with no detectable leakage. This demonstrates that acid gas injection is a good analogue to CO2 injection and provides additional confidence in operational safety issues. In many parts of the world, large volumes of liquid waste are injected into the deep subsurface every day as a disposal option. For example, for the past 60 years, approximately 34.1 million m3 of hazardous waste has been injected into saline formations in the USA from about 500 wells each year. In addition, more than 2843 million m3 of oil field brines are injected from 150 000 wells each year. This combined annual US injectate volume of about 3000 million m3, when converted to volume equivalent, corresponds to the volume of approximately 2 Gt (Giga tonne) CO2 at a depth of 1 km, which is of a similar magnitude to that which may be required for geological storage of CO2 (IPCC, 2005). Furthermore, evidence from oil and gas fields indicates that hydrocarbons and other gases and fluids, including CO2, can remain trapped for millions of years (Magoon and Dow, 1994; Bradshaw et al., 2003). Carbon dioxide has a tendency to remain in the subsurface (relative to hydrocarbons) via its many physicochemical immobilization mechanisms. World-class petroleum provinces (country/state- or sedimentary basin-scale petroleum occurrences/ accumulations) have storage times for oil and gas of 5–100 million years, others for 350 million years, while some minor petroleum accumulations have been stored for up to 1400 million years. However, some natural traps do leak, which reinforces the need for careful site selection, characterization and injection practices (IPCC, 2005). Therefore, the experience gained from existing deep-fluid-injection projects is relevant in terms of operation and provides best practices for injection and storage of CO2 to be feasible and safely managed.

14.4

Injection well technologies

Many of the technologies required for large-scale geological storage of CO2 already exist. Drilling and completion technology for injection wells in the oil and gas industry has evolved to a highly sophisticated state, such that it is now possible to drill and complete vertical and extended reach



439

wells (including horizontal wells) in deep formations, wells with multiple completions and wells able to handle corrosive fluids. On the basis of extensive oil industry experience, the technologies for drilling, injection, stimulations and completions for CO2 injection wells exist and are being practised with some adaptations in current CO2 storage projects (e.g. In Salah–Algeria and Sleipner–Snøhvit Norway projects). In a CO2 injection well, the principal well design considerations include pressure, corrosion-resistant materials and injection rates. The design of a CO2 injection well is very similar to that of a gas injection well in an oil field or natural gas storage project. Most downhole components need to be upgraded for higher pressure ratings and corrosion resistance. The technology for handling CO2 has already been developed for enhanced oil recovery (EOR) operations and for the disposal of acid gas. A simplified vertical CO2 injection well is illustrated in Fig. 14.1. Injection wells are commonly equipped with two valves for well control, one for regular use and one reserved for safety shutoff. The valves may lie externally as master valve above the wellhead or along the CO2 supply pipeline or Injection pressure gauge

CO2 pipeline from source Annulus pressure gauge Surface casing Long string casing Annular space Injection tubing

Double-grip packer Perforations

14.1 Illustration of a simplified vertical CO2 injection well.


440


one external (as mentioned) and the other internally in the injection tubing above the double-grip packer. The main function of the double-grip packer is to serve as a mechanical hold-down system, which enables differential pressures to be held securely from both above and below the packer in the annulus between the casing and the injection tubing or the pressure in the annulus to be maintained. In acid gas injection wells, a downhole safety valve is incorporated in the tubing, so that if equipment fails at the surface, the well is automatically shut down to prevent back flow. Jarrell et al. (2002), recommend an automatic shutoff valve on all CO2 wells to ensure that no release occurs and to prevent CO2 from inadvertently flowing back into the injection system. A typical downhole configuration for an injection well includes a double-grip packer, an on–off tool and a downhole shutoff valve. Annular pressure monitors help detect leaks in packers and tubing, which is important for taking rapid corrective action. To prevent dangerous highpressure buildup on surface equipment and CO2 releases into the atmosphere, CO2 injection must be stopped as soon as a leak occurs. Rupture disks and safety valves can be used to relieve built-up pressure. Adequate plans need to be in place for dealing with excess CO2 if the injection well needs to be shut in. Options include having a backup injection well or methods to safely vent CO2 to the atmosphere. The number of wells required for a storage project will depend on a number of factors, including total injection rate, permeability and thickness of the formation, maximum injection pressures, wellbore design and availability of land-surface area for the injection wells. In general, fewer wells will be needed for high-permeability sediments in thick storage formations and for those projects with horizontal wells for injection. For example, the Sleipner Project, which injects CO2 into a high-permeability, 200 m thick formation uses only one well to inject 1 Mt CO2 yr–1 (Korbol and Kaddour, 1994). Proper maintenance of CO2 injection wells is necessary to avoid leakage and well failures. Several practical procedures can be used to reduce the probabilities of CO2 blowout (uncontrolled flow) and mitigate the adverse effects if one should occur. These include periodic wellbore integrity surveys on drilled injection wells, real-time monitoring of wellbore conditions, regular inspection and testing of downhole safety valves, improved crew awareness, contingency planning and emergency response training (Skinner, 2003). For CO2 injection through modified, re-used old wells, key factors include the mechanical condition of the well and quality of the cement and well maintenance. A leaking wellbore annulus can be a pathway for CO2 migration. Detailed logging programmes for checking wellbore integrity can be conducted by the operator to protect formations and prevent reservoir cross-flow. A well used for injection must be equipped with a packer to isolate pressure to the injection interval (Fig. 14.1). All materials used in injection wells should be designed to anticipate peak injection rate, pressure and the full



441

range of temperatures in all modes of operation. Use of corrosion-resistant materials is essential in all CO2 injectors, such as the use of chromium steel (typically 13 % chromium).

14.5

Controlling parameters for carbon dioxide (CO2) injectivity

Injectivity is the rate at which CO2 can be injected into a given reservoir interval (a volume of CO2 per unit of time per unit pressure drop in the reservoir) and the ability of the subsequent CO2 plume to migrate away from the injection well (Kaldi and Gibson-Poole, 2008). During CO2 injection into a reservoir, the injectivity and nature of plume migration will depend on parameters such as the viscosity ratio between the injecting fluid and the in situ fluid, injection rate, differential injection pressure, fracture gradient, well/completion design, relative permeability effects as a result of two-phase or multiphase flow and stress dependency of permeability. Most of the parameters will in turn depend on variables such as depositional environment and reservoir heterogeneity, stratigraphic architecture, postdepositional diagenetic alteration, structural dip, fault distribution and fault seal capacity and pressure distribution, and on the nature of the formation fluids. Injectivity issues that require special attention include: the geometry and connectivity of individual flow units, the nature of the heterogeneity within those units (i.e., the likely distribution and impact of baffles such as interbedded siltstones and shales) and the physical quality of the reservoir in terms of porosity and permeability characteristics (Kaldi and Gibson-Poole, 2008). Here focus is given to the key factors that affect injectivity, including formation permeability, injectivity loss, wellbore design and injection well pressure and reservoir constraints.

14.5.1 Permeability Numerical investigations show that formation permeability is one of the main controlling parameters of CO2 storage in geological formations (van der Meer, 1995; Law and Bachu, 1996; Ennis-King and Paterson, 2001). For low-permeability formations, there will be large pressure gradients near the wellbore, which will restrict the injectivity considerably. Thus, high permeability formations would be desirable for injection. However, high permeability formations would allow relatively fast migration of CO2, lowering the proportion of CO2 trapped behind the main plume by residual trapping, (e.g. Solomon et al., 2008). Slower movement provides greater residence time and thus several advantages such as enhancement of the dissolution potential of CO2 in formation water and high volumes of CO2 trapped by capillarity (Flett et al., 2005). Bachu et al. (1994) suggest that © Woodhead Publishing Limited, 2010

442


while higher permeability may be required near the wellbore to increase injectivity, lower permeability is desirable outside the radius of influence of the wellbore to increase residence times and maximize the rate of residual trapping, dissolution and mineral trapping. The Utsira sand at Sleipner in Norway is a typical example of CO2 injection in a high-permeability formation which is about 1000 milli-Darcy (mD). On the other hand, the CO2 injection at In Salah in Algeria may be typical of low-permeability formation which is between 5 and 15 mD. Formation permeability controls not only CO2 injectivity but also whether CO2 storage is commercially viable or not, because the overall injectivity of the selected injection location and the pressure and fluid conductivity of the total affected storage space can control the total storage space and capacity. If the fluid conductivity (directly proportional to absolute reservoir permeability) is too low, unacceptable pressure gradients could render the potential location unsuitable for large storage volumes (van der Meer and Egberts, 2008). This could, however, potentially be overcome by drilling more injectors or improving their injectivity through design and stimulation. Figure 14.2 shows pressure development at the end of injection cycle for a hypothetical saline aquifer CO2 injection modelling case, for low and high 100

K = 200 mD K = 1000 mD

90 80

Pressure (Bar)

70 60 50 40 30 20 10 0 0

20

40

60 80 Distance (km)

100

120

14.2 Pore pressure variations and effect of formation permeability on CO2 injectivity data from generic reservoir simulations in saline aquifer.



443

permeabilities. For the high-permeability (1000 mD) case, high injection rates can be sustained with minimum probability of exceeding maximum pressure thresholds set in order to avoid fracturing the reservoir or re-activating existing fractures. For the case of low-permeability formation, the injection rate is the same as for the high-permeability formation injection, but the low-permeability formation experiences much higher injection pressure. The probability of exceeding maximum pressure thresholds is significantly higher for the low-permeability case, and the result may be that injection rate should be reduced.

14.5.2 Injectivity loss Several different processes can cause loss of injectivity. The most relevant for CO2 injection are halite precipitation, mobilized fine particles, entrained fine particles and geochemical reactions. Each process is discussed below. Halite precipitation The injection of dry supercritical CO2 into brine aquifers has the potential to dry formation waters, due to evaporation effects (Spycher and Pruess, 2005). Dry supercritical CO2 has the ability to ‘evaporate’ (or dissolve) small amounts of water. This could significantly impair injection rates, as has been documented in natural gas storage reservoirs (Kleinitz et al., 2001). During injection of dry CO2, the formation water is ‘evaporated’ by dissolution into the dry CO2 and since deep formation waters have dissolved salts, the water dissolution into the flowing CO2 stream will lead to increase in salt concentrations that eventually reach the solubility limit, giving rise to precipitation of halite (solid NaCl). Close to the borehole, where the dry CO2 enters the rock formation, pore throats can be blocked by halite precipitation that changes pore geometry. The absolute permeability can be reduced significantly with the effect of dramatic decrease in injectivity. Simulation results (Müller, 2007) show up to 50–100 % decrease in permeability and thus the injection capacity becomes limited by the permeability reduction in the formation adjacent to the injection well. Thus halite precipitation could have a severe impact on the injectivity of CO2. There are simple reservoir engineering techniques available that could prevent or remediate halite precipitation. One solution could be pre-flushing the formation close to the borehole with a carefully chosen water-based solution (often simply freshwater with an additive, e.g. potassium chloride, in order to avoid mobilising discrete clay particles in the pore network) and inject CO2 after the pre-flush. The high salt concentration front would be at further distance from the injection point. Halite precipitation would still


444


occur, but less spatially concentrated. Thus, the permeability impairment at the injection point would be less severe. Mobilized fine particles Loss in injectivity in reservoir rocks can occur when dispersed in situ fine particles are mobilized and re-deposited in pore throats and narrower passages, thereby lowering local permeability and, in extreme cases, completely plugging the formation. Although this problem is best known in production wells in hydrocarbon reservoirs (Cailly et al., 2005), it cannot be ruled out for CO2 injection wells. Loss of well injectivity may also arise due to presence of entrained particles (e.g. discrete particles resulting from corrosion in the pipe system). Geochemical reaction products and effects Reactive dissolution of certain reservoir matrix minerals in contact with injected CO2 and CO2-saturated brine can cause increased porosity and permeability in the initial reaction front, but insoluble reaction products will be precipitated downstream. This can potentially plug pore throats. For flow systems in series, the permeability-reduced zone will dominate the permeability-enhanced zone. For flow systems in parallel, the effect is much more subtle. Local flow paths in the reservoir are a very complex combination of series and parallel, especially for the simultaneous flow of CO2 and brine because it limits the access of the reactive brine to a limited fraction of the pore space due to the non-wettability of the CO2 phase (Cailly et al., 2005). The non-wettability of the CO2 phase can be explained in terms of preference of the reservoir mineral grain surfaces for the fluid brine instead of the CO2 gas phase. Usually both fluids (CO2 and brine), compete to occupy the available pore space in the reservoir rock. Therefore, the distribution of minerals within individual pore throats may constrain the way the pore structure is modified by dissolution or precipitation and the associated permeability changes. Effects of impurities in the injection stream Loss of well injectivity may also arise due to gaseous impurities in the CO2 stream that will compete for the available storage space to be occupied by the CO2. The potential for geochemical reactions to plug reservoir pore throats may also be increased, depending on the types of impurities in the injection stream and the reactive nature of the reservoir mineralogy. Important factors to be considered include:



∑ ∑ ∑ ∑ ∑ ∑ ∑

445

the concentration of impurities contained in the CO2 stream (in the case of some capture technologies this can be up to 10 % of the injected stream); the pore-size distribution, porosity and permeability of the porous medium; the nature of the interaction of the injected CO2 with the rock materials and pore fluids; the flow velocity of the injected CO2 stream; the geometry of the injection system (horizontal completion or vertical completion); the multiphase flow effects; the thermodynamic conditions.

Although there is little field experience with CO2 injection into brine reservoirs, there is an extensive literature resource when brine is injected for recovery purposes in the oil and gas industry (e.g. Al-Homadhi, 2001). This could help to understand the injection process in general and the controlling parameters in particular. Most of the above factors have been studied by many researchers in the context of rock matrix plugging in hydrocarbon reservoirs when brine is injected, applying both laboratory techniques on small rock samples from reservoirs (cores) as well as numerical simulations. One of the common, general conclusions is that, as is the case when injecting CO2, initial matrix permeability is important, and it is therefore given wide attention in the literature (Al-Homadhi, 2001). Experience has shown, however, that experimental core flow tests (small-scale and short-duration) are often inconsistent with field tests, where injectivity losses, when observed, occur over periods of weeks to months (Al-Homadhi, 2001). Simpson and Paige (1991) stated that no decline in well injectivity was observed throughout a six weeks’ field injection trial. This was explained by fracturing and other forms of deformation caused by the brine injection process. Fracturing increases injectivity, and it creates a large sandface area, which takes a much longer time to plug. In some circumstances, fractures are capable of extension to create a new sandface and maintain injectivity in response to any plugging which does occur (Al-Homadhi, 2001). Thus injectivity loss in the context of CO2 injection may not be a critical issue, but it requires further study as well as good risk management. This implies that site-specific evaluation of potential injectivity losses is necessary, applying the extensive experience from oil and gas field operations and applicable regulations for underground injection.

14.5.3 Wellbore design An appropriately designed and constructed well would help to enhance CO2 injectivity in geological formations, prevent endangerment to the environment © Woodhead Publishing Limited, 2010

446


and would maintain integrity throughout the lifetime of the project, from the injection operation period through and beyond the post-injection site care period once the well is permanently plugged (EPA, 2008). Current drilling and well construction practices for CO2 injection wells are based on existing knowledge and practices from the oil and gas industry. The wellbore design aspect can be seen in terms of whether a vertical well completion, horizontal well completion or an inclined well completion is more appropriate for CO2 injection, or a combination of any of the individual wellbore designs. Horizontal wells are constructed using a directional drilling system. This drilling technique achieves the transition from vertical hole to horizontal hole by employing what is known as a lateral drilling assembly which forces the drillbit to deflect from vertical over a curvature radius that is compatible with the rest of the drilling operations, i.e. allows safe passage of the straight sections of pipe and assemblies through the curved section of hole. A straight, inclined section is usually drilled before a final curved section achieves horizontal orientation. The horizontal section can be lined using sandscreens, slotted liners or a liner cemented in place (which would then need to be perforated) or remain as an open hole. Importantly, several horizontal sections can be completed, stemming from a single vertical completion. While horizontal well construction is not typical in deep injection wells in the UIC (Underground Injection Control) Program (a division under the United States Environmetal Protection Agency reponsible for regulating the construction, operation, permitting and closure of injection wells that place fluids underground for storage or disposal), there are examples of horizontal well completions being used with success to improve the production of EOR and enhanced coal bed methane recovery (ECBM) operations (EPA, 2008). Horizontal and extended reach wells can be good options for improving the rate of CO2 injection from individual wells. The Weyburn field in Canada is a good example in which the use of horizontal injection wells contributes to improving injectivity, overall displacement (sweep) efficiency and oil recovery. At the In Salah project in Algeria, horizontal wells are also constructed for CO2 geological storage purposes. CO2 is injected into a 20 m thick formation with much lower permeability (Riddiford et al., 2003). Here, three long-reach horizontal wells with slotted intervals over 1 km long in the reservoir are used to inject 1 Mt CO2 yr–1 (Fig. 14.3). The use of horizontal wells for a CO2 geological storage project could provide several benefits over vertical or inclined wells. First, the horizontal injectors reduce the number of injection wells required for field development. Second, horizontal wells provide enhanced connectivity with permeable sections of the formation, increasing injectivity. In addition to these, the use of horizontal wells can improve the sweep, or formation contact area, of the injected CO2 plume, as vertical channeling through high permeability © Woodhead Publishing Limited, 2010


447

Processing facilities

Sandstone and mudstone (900 m thick)

Mudstone (950 m thick) CO2 injection wells

Production wells Gas

Water

14.3 Schematic of the In Salah Gas Project, Algeria where 1 Mt CO2 is stored annually in the gas reservoir.

regions is reduced. Increasing the sweep results in enhancing residual CO2 trapping and dissolution, which promotes permanent storage in the reservoir. Horizontal wells also reduce the pressures needed to inject any given volume of fluid, which has the additional benefit of reducing the load on the CO2 compressors and their subsequent energy requirements. In addition, fewer vertical completions are required with the use of horizontal wells, which reduces the number of artificial penetrations in the formation through which fluid could migrate, and can potentially reduce overall costs (EPA, 2008). A horizontal injection well has the added advantage that it can create injection profiles that can potentially avoid the adverse effects of preferential flow of injected fluid through high-permeability zones (which is often referred to as poor sweep).

14.5.4 Injection well pressure and reservoir constraints The initial pre-injection pore pressure in the reservoir formation also plays a role in designing the injection rate together with the other parameters. Usually, the injection pressure should be kept lower than the minimum in situ stress (fracture pressure) of the reservoir. The pressure of fluids within the pores of a reservoir is usually referred to as the hydrostatic pressure, or the pressure exerted by a column of water from the formation’s depth to sea level. When impermeable rocks such as shales formed as sediments are compacted, their pore fluids cannot always escape and must then support


448


the total overlying rock column, leading to anomalously high formation pressures. The pressure above which injection of fluids will cause the rock formation to break or fracture hydraulically is called the fracture pressure. At CO2–EOR projects, e.g. Weyburn, current practice is to operate at initial reservoir pressure of 14.2 MPa, maximum injection pressure (90 % of fracture pressure) is in the range 25–27 MPa and fracture pressure is in the range 29–31 MPa (IPCC, 2005). Injectivity characterizes the ease with which a fluid can flow from the wellbore into the targeted geological formation. This depends on the injection rate as well as the pressure difference between the injection point inside the well and the formation. Although CO2 injectivity should be significantly greater than brine injectivity (because CO2 has a much lower viscosity than brine), this is not always the case. Grigg (2005) analyzed the performance of CO2 floods in west Texas and concluded that, in more than half of the projects, injectivity was lower than expected or decreased over time. Christman and Gorell (1990) showed that unexpected CO2 injectivity behaviour in EOR operations is caused primarily by differences in flow geometry, probably due to formation heterogeneity and fluid properties of the oil. Injectivity changes can also be related to unanticipated relative permeability effects. To introduce CO2 into the storage formation, the downhole injection pressure must be higher than the reservoir fluid pressure. On the other hand, increasing reservoir pressure may induce fractures in the formation. Regulatory agencies normally limit the maximum downhole pressure to avoid fracturing the injection formation. Measurements of in situ formation stresses and pore fluid pressure are needed for establishing safe injection pressures. Safe injection pressures can vary widely, depending on the state of stress and tectonic history of a basin. Regulatory agencies have determined safe injection pressures from experience in specific oil and gas provinces. Van der meer (1996) has derived a relationship for the maximum safe injection pressure. This relationship indicated that for a depth down to 1000 m, the maximum injection pressure is estimated to be 1.35 times the hydrostatic pressure – and this increased to 2.4 for depths of 1–5 km. The maximum pressure gradient allowed for natural gas stored in an aquifer in Germany is 16.8 kPa m–1 (Sedlacek, 1999). This value exceeds the natural pressure gradients of formation waters in north-eastern Germany, which are on the order of 10.5–13.1 kPa m–1. According to IPCC (2005), in Denmark or the UK, the maximum pressure gradients for aquifer storage of natural gas do not exceed hydrostatic gradients. In the USA, for industrial waste water injection wells, injection pressure must not exceed fracture initiation or propagation pressures in the injection formation (USEPA, 1994). For oil and gas field injection wells, injection pressures must not exceed those that would initiate or propagate fractures in the confining units. In the USA, each state has been delegated authority to establish maximum injection pressures.



449

Until the 1990s, many states set state-wide standards for maximum injection pressures; values ranged from 13–18 kPa m–1. More recently, regulations have changed to require site-specific tests to establish maximum injection pressure gradients. Practical experience in the USEPA’s UIC Program has shown that fracture pressures range from 11–21 kPa m–1.

14.6

Carbon dioxide (CO2) Injection in different storage formations

Chapters 2–5 in Volume 2 cover the geological aspects of storage in more detail. Here, the injection characteristics in each storage option will be presented briefly.

14.6.1 Saline formations Saline formations are deep sedimentary rocks saturated with formation waters or brines containing high concentrations of dissolved salts. These formations are widespread and contain enormous quantities of water. CO2 storage in saline formations has several advantages over other geological storage options (i.e. depleted oil and gas reservoirs, coal seams), such as greater storage volume potential and less risk of compromising existing resources (Kaldi and Gibson-Poole, 2008). However, many deep saline formations exhibit low permeability, due either to depositional processes (fine grained sediments or poorly sorted sediments and corresponding small pore throat sizes) or diagenetic processes (post-depositional mineralogical modification of pores and pore throats). The Sleipner Project in the North Sea is the best available example of a CO2 storage project in a saline formation. It was the first commercial-scale project dedicated to geological CO2 storage. Approximately 1 Mt CO2 is removed annually from the produced natural gas and injected underground at the Utsira sand formation. The operation started in October 1996, and over the lifetime of the project a total of 20 Mt CO2 is expected to be stored. The CO2 is injected into uncemented (unconsolidated), clean, well-sorted sands about 800–1000 m below the sea floor. The sandstone contains secondary thin shale or clay layers, which influence the internal movement of injected CO2. The overlying primary seal is an extensive thick shale or clay layer of Pliocene age. The saline formation into which CO2 is injected has a very large connected pore volume and corresponding storage capacity. The pore pressure is ca 8.75 MPa while the injection pressure is slightly above the pore pressure, about 9 MPa, due to its high permeability and hence high injectivity.


450


14.6.2 Depleted oil and gas fields Although depleted oil and gas reservoirs are prime candidates for CO2 storage for several reasons, injectivity is limited by the need to avoid using pressures that damage the caprock. Oil and gas fields have demonstrated their integrity and safety; however, Jimenez and Chalaturnyk (2003) suggest that geomechanical processes, during depletion and subsequent CO2 injection, may affect the hydraulic integrity of the seal rock in hydrocarbon fields. Movement along faults can be produced in a hydrocarbon field by induced changes in the pre-production stress regime. This can happen when fluid pressures are substantially depleted during hydrocarbon production (Streit and Hillis, 2003). Determining whether the induced stress changes result in compaction or pore collapse is critical in assessment of CO2 injectivity potential in a depleted field. If pore collapse occurs, then it might not be possible to return a pressure-depleted field to its original pore pressure without the risk of induced failure. By having a reduced maximum pore fluid pressure, the total volume of CO2 that can be stored in a depleted field could be substantially less than otherwise estimated. Depletion of fluid pressure during production can affect the state of stress in the reservoir. Analysis of some depleted oil and gas reservoirs indicated a decrease of the horizontal rock stress by 50–80 % of the pore pressure, which implies an increased possibility of fracturing the reservoir if used for CO2 storage (Streit and Hillis, 2003).

14.6.3 Injection in coal seams Coal contains fractures (cleats) that impart some permeability to the system. Between cleats, solid coal has a very large number of micropores into which gas molecules from the cleats can diffuse and be tightly adsorbed. Gaseous CO2 injected through wells will flow through the cleat system of the coal, diffuse into the coal matrix and be adsorbed onto the coal micropore surfaces, displacing and freeing up gases with lower affinity to coal (e.g. methane). If CO2 is injected into coal seams, it can displace methane, thereby enhancing coal bed methane recovery (ECBM). However, coal plasticization or softening may adversely affect the permeability that would allow CO2 injection. Furthermore, coal swells as CO2 is adsorbed and/or absorbed, which reduces permeability and injectivity by orders of magnitude or even more (Shi and Durucan, 2005). Some studies suggest that the injected CO2 may react with coal, further highlighting the difficulty in injecting CO2 into low-permeability coal (IPCC, 2005). Although coal plasticization due to CO2 injection may also cause reduction of permeability in some coals, in a series of recent publications no evidence of irreversible behaviour of the coals on exposure to CO2, even up to 20



451

MPa exposure, has been found (Kaldi and Gibson-Poole, 2008). Although this does not completely exclude the possibility that some coals can be plasticized by exposure to CO2, the phenomenon may be less common than previously believed. Other parameters which control injectivity of CO2 in coals are seam thickness and adsorption capacity (Kaldi and Gibson-Poole, 2008).

14.6.4 Hydro-mechanical impact of carbon dioxide (CO2) injection Regardless of the type of storage formation, injection of CO2 will result in formation fluid pressure increase, especially around the injection source. Such a fluid pressure increase will cause local changes in the stress fields due to continuous modification of the effective stress. This will induce mechanical deformations and possible irreversible mechanical failure in the cap rock. This mechanical failure may involve shear along many of the pre-existing fractures (if any) or creation of new fractures that reduce the sealing capacity of the cap rock. In addition to this, replacing the native formation fluid with CO2 may cause changes in the rock mechanical properties through chemomechanical interactions between the CO2 and the host rock, or through desiccation of fractures (Rutqvist and Tsang, 2005). The overall result of the hydro-mechanical impact of CO2 injection is that new leakage pathways through a cap rock or existing faults and fractures might be initiated. Thus the evaluation of the risk of breach in a cap rock/reservoir system is required and involves prediction of complex coupled thermal–hydrological–geomechanical–geochemical processes over a long period of time (Rutqvist and Tsang, 2005). Such predictions can be performed through appropriate modelling tools in combination with input data from laboratory and field experiments and measurements. Such evaluations should, as far as possible be combined with in situ monitoring of system mechanical response to the CO2 injection. Several geophysical techniques (e.g. time-lapse seismic monitoring and acoustic tomography) are available to accomplish the task. These techniques can monitor fluid-induced changes and can help to detect leakage pathways through a cap rock. Details on the approach for analysis of coupled hydro-mechanical effects on cap rock integrity and potential reservoir leakage during CO2 injection can be obtained from Rutqvist and Tsang (2005).

14.7

Carbon dioxide (CO2) injection field operations

Whether the CO2 injection operation is offshore or onshore, the CO2 is typically transported from its source to the sink or site of storage through a pipeline and is then injected into the reservoir through an injection well,


452


usually after compression. Field operations of CO2 injection can vary from site to site. For instance, at Sleipner the CO2 is separated from the natural gas during production and the captured CO2 is then injected back to a saline aquifer at 800–1000 m depth (Utsira formation) overlying the natural gas producing reservoir (Sleipner East Field or Heimdal formation, Fig. 14.4). In EOR operations, CO2 produced from the production wells along with oil and water is separated and then injected back through the injection well (Fig. 14.5). The field application of CO2–ECBM technology is broadly similar to that of EOR operations. Carbon dioxide is transported to the CBM field and injected in the coal seam through dedicated injection wells (IPCC, 2005). At the production well, coal-seam gas and formation water is lifted to the surface by pumps. Jarrell et al. (2002) pointed out that CO2 facilities are similar to those used in conventional facilities such as for water floods, and they have outlined typical surface facilities for CO2–EOR projects. Differences result from the effects of multiphase flow, selection of different materials and the higher pressure that must be handled. It is common to use existing facilities for new CO2 projects to reduce capital costs, although physical restrictions are always present. Starting a CO2 flood in an old oil field can affect almost every process and facility (Jarrell et al., 2002); for example, (i) the presence of CO2 makes the produced water much more corrosive; (ii) makeup water from new sources may interact with formation water to create new problems with Platform Sea Gas from sleipner west

Utsira formation (800–1000 m depth)

Caprock/shale

CO2 injection well Sleipner East production well

Sleipner east field (Heimdal formation)

14.4 Simplified illustration of the Sleipner CO2 storage site in Norway.



453

Production well

CO2 injection well CO2 Recycled CO2

CO2

Miscible zone

Oil recovery

Oil reservoir

14.5 Injection of CO2 for enhanced oil recovery (EOR) and the CO2 that is produced with the oil is separated and re-injected back into the formation.

scale or corrosion; (iii) a CO2 flood may cause paraffins and asphaltenes to precipitate out of the oil, which can cause plugging and emulsion problems; and (iv) the potentially dramatic increase in production in association with the flood could cause more formation fines to be mobilized, potentially causing plugging, erosion and processing problems.

14.8

Injection of carbon dioxide (CO2) and well integrity

14.8.1 Induced seismic events Local stress accumulation in reservoirs could cause sudden failure and therefore a micro-earthquake or induced earthquakes. Numerous mechanisms generate microseismicity, and these may include: gas storage, hydraulic fracturing, fluid injection, production-related reservoir compaction, presence of lowporosity reservoirs, high injection pressures, the use of fluids/proppants, the stress regimes and the presence of natural fractures and faults. Hydraulic


454


fracturing is a form of tensile failure that occurs when the fluid pressure/ injection pressure exceeds the local least principal stress. When previously undetected structures such as faults and fractures exist, the pre-existing fractures become reactivated even at lower injection pressure and contribute to the propagation of the fractures which could potentially cause casing failure or damage the cement sheath, hence compromising the wellbore integrity. The natural state of stress due to the geological setting, the mechanical properties of the reservoir rocks and host rocks and the changes in pore pressure due to fluid withdrawal or injection are the principal causes of induced seismicity in a reservoir in which there is fluid movement (Fabriol, 2002). Deep-well injection of waste fluids may induce earthquakes with moderate local magnitudes (ML), as evidenced by the 1967 Denver earthquakes, e.g. ML of 5.3 (Healy et al., 1968) and the 1986–1987 Ohio earthquakes, e.g. ML of 4.9 (Ahmad and Smith, 1988) in the USA. Data on the induced microseismicity due to CO2 injection is not available in the public domain. However, recent data from seismicity induced by gas exploitation is reported from the north of the Netherlands (van Eck et al., 2006), which may give an indication to the magnitude of these induced earthquakes, assuming that the rate of gas production is equivalent to gas injection. Data obtained from three gas reservoirs show that the seismicity that was recorded so far has been associated with the reactivation of major faults in each gas field. It is conceivable that reactivation of major faults may be the cause of induced seismicity in the case of CO2 injection due to increased pore pressure in the reservoir. Most events have been located around a depth of 2.5 km ± 0.5 km, corresponding to the average depth of the reservoirs with local magnitudes ranging from 1–3.5. The first events recorded vary from field to field but range between 12 and 28 years after gas production. These data may provide a rough estimation of the expected magnitudes as well as expected durations for events related to CO2 injection.

14.8.2 Cement degradation The cement used to line and/or plug the well may be vulnerable to acid attack. Injected CO2 will dissolve in reservoir brines, forming carbonic acid, which can readily react with calcium hydroxide [Ca(OH)2] and calcium silicate hydrate, key components in hardened cement, and thus can compromise the integrity of the injection wellbore. Laboratory experimental results conducted under simulated storage reservoir conditions show that upon exposure to aqueous CO2, hardened cement formed well-defined reaction zones by a two-step process (Kutchko et al., 2007). The first step is the dissolution of Ca(OH)2 (s) and subsequent precipitation of CaCO3 (s). The formation of CaCO3 (s) has been reported to decrease cement permeability and increase its compressive strength. The second step is the dissolution of CaCO3 (s), © Woodhead Publishing Limited, 2010


455

resulting in a leaching of calcium from the cement matrix. The resulting cement paste has a significant increase in porosity, is primarily composed of amorphous silica gel and lacks structural integrity. Although it is clear that cement is degraded, the results of this study suggest that the reactions involved are slow (Kutchko et al., 2007). In fact, long-term experiments show that the rate of degradation decreases over time, probability due to the precipitation of CaCO3 (s) within the pore space of the cement. This phenomenon should limit the negative impact that chemical degradation will have on well bores. Supercritical CO2 exposure (saturated with water vapour) led to a very different process by which CaCO3 (s) was deposited throughout the matrix and on the surface, rather than within an isolated reaction zone. Over the one-year time period of the experiments, this condition led to a smaller amount of total degradation than in the aqueous phase. However, in this case, there was no deceleration of the reaction observed. It is unlikely that the diffusion-controlled degradation process observed in these experiments would lead to wellbore failure in well completions that are well cemented with neat Portland cement (without additives) (Kutchko et al., 2007). Further investigation is required to evaluate the effect of cement additives, fractures or channels in the cement, and geomechanical stress. Thus chemical degradation of the cement well plug and/or cement sheath itself is considered the slowest, and leakage due entirely to chemical degradation of neat (without additives, particularly bentonite) cement may not be a significant concern (Strazisar et al., 2008). There are currently two technology strategies in use that are special formulation of cements for CO2 injection; the first one is silica flour-rich cement which is meant to reduce the reaction of CO2 with components of the cement such as calcite; and the second is a very low-porosity and low permeability cement with interlocked granular matrix designed to prevent entry of CO2-rich fluid. It is important to mention, however, that work is still in progress to improve the standards of both cement types.

14.8.3 Debonding Failure at the interface between the cement and formation rock or debonding of cement-formation is the primary initiator for failure in wellbore due to changes in pressure, temperature and stress (Fig. 14.6). Pilkington (1992) claimed that up to 90 % of all wellbores reviewed under a comprehensive evaluation of cement quality logging procedures had a microannulus indicative of debonding. Various mechanisms may contribute to this type of failure mode, but mainly cement shrinkage leads to drop in radial stress and hence an increase in tangential stress (Fig. 14.6). When the radial stresses are lower than the pore pressure, circumferential fracture will develop at the rock–cement interface due to increase in the tangential stress (Dusseault et al., 2000). This is a condition of the pore pressure being greater than the © Woodhead Publishing Limited, 2010

456

Developments and innovation in CCS technology Vertical stress

R – rock C – cement WC – well casing CP – cement plug Radial stress

Tangential stress

CP

R

C

WC

C

R

14.6 Illustration of rock–cement–casing interfaces including stresses around a wellbore.

minimum horizontal stress which means that the hydraulic fracture criterion is reached where circumferential fractures that are perpendicular to the radial pressure (minimum horizontal stress) and typically no wider than 10–20 mm develop at the rock–cement interface (Dusseault et al., 2000). Microannulus of 25 mm is enough to create a detectable gas leak (Gray et al., 2009). Differences between lateral stress gradients and pressure gradients provide forces for vertical growth of the fracture. The excess pressure that develops at the upper leading tip of the fracture increases with the (vertical) height and the fracture tends to become gas filled as gas slowly diffuses into it, increasing the driving force. This could lead the fracture to rise and the gas to enter to shallow strata or leak at the surface, thus compromising the injection well integrity. Typical bond strength (i.e. the tensile resistance of the cement–rock interface) is quite small, generally less than 1–2 MPa (Dusseault et al., 2000).

14.8.4 Casing shear failure In some extreme cases, the changes in stress state in the target formation are so great that the casing collapses, either broken by shear forces or plastically deformed to the degree that it blocks the function of the wellbore. One such function could be to seal flow of fluids at deep reservoirs. Casing shear or collapse caused by overpressure in the formation is of interest in the case of a CO2 injection because it can cause a local lateral movement between the top of the target reservoir and the bottom of the primary cap rock or



457

reactivation of earlier faults and hence shear collapse of the casing (Fig. 14.7a). Additional causes are creep flow of some plastic reservoir rocks such as halite and, finally, natural seismic events or induced seismic events may cause relative movement and motion in the underground that damages wellbore casing and cement. One group of industry experts (Dusseault et al., 2001) identifies the dominant casing-deformation mechanisms as being localized horizontal shear at weak lithology interfaces within the overburden; localized horizontal shear at the top of production and injection intervals; and casing buckling within the producing/storage interval, primarily located near perforation. Casing shear is caused by rock shear. Rock shear is caused by changes in stress, pressure and temperature induced by typical petroleum recovery activities such as depletion, injection and heating. During CO2 injection, reactivation of pre-existing faults may trigger shear failure (Fig. 14.7b). In addition to this, the saline water displaced by the CO2 injection may invade the caprock, hence decreasing the shear strength and friction coefficient of the caprock (shale) causing creep deformation under ambient differential stresses. The second cause could be significant reservoir pressure differences due to the CO2 injection processes which may (a)

Adjacent wells

Caprock/shale Shear failure on interface

Reservoir with increased pore pressure zone

Injection well

14.7 Schematic illustration of casing shear: (a) injection-induced interface shear stress along the contact between the caprock and the reservoir; and (b) low-angle dipping fracture plane reactivation due to injection-induced shearing.


458

Developments and innovation in CCS technology (b)

Aggressive CO2 injection

Low-permeability shale

High shear stress zone

Highpressure and low effective stress zone

Reservoir

Low angle dipping fracture plane

14.7 Continued

be accompanied by strains. Such events have been documented to be the causes for casing shear failures, due to high-pressure water injections, in big oilfields; about 18 % of the total wells showing casing shear of which nearly 70 % occurred at the bottom of thick upper shale layer in contact with the reservoir (Han et al., 2006). Injection also leads to shearing by reservoir expansion that leads to shearing near lithology bounding interfaces where stresses are concentrated. The bounding strata, being relatively impermeable seal rocks, do not experience a similar stress change; therefore, they have no tendency to expand. As a pressured zone propagates from the injection point and the permeable rock tries to expand outward, a large shear stress is imposed on the interface between the reservoir and the bounding strata (Fig. 14.7a). If this shear stress exceeds the interface strength, slip ensues and casings in offset wells can be impaired (Dusseault et al., 2001), thus blocking the function of the injection well.

14.8.5 Corrosion Carbon dioxide, like hydrogen sulphide (H2S), will form a weak acid and becomes corrosive when dissolved into water. It is well known that the corrosion rate of steel in wet CO2 can be quite high and depends on a number



459

of factors, including CO2 partial pressure, temperature, flow regime and rate, the pH of the liquid phase and the presence of other impurities such as amine, H2S and O2. Gaseous or supercritical dry CO2 is not corrosive. The few exceptions to this are wells completed for acid gas service, geothermal service, steam injection service or, in fact, for CO2–EOR service. A corrosive breach of the innermost casing above the shallowest cement plug in wellbores will allow high-pressure, buoyant wet CO2 fluids to rise to the top of the wellbore if it is not filled with cement, and a complete wellbore integrity failure can occur. At a potential storage site, rough corrosion rate estimation for the casing in the radial wellbore direction using data compiled from different sources, showed ranges from ~0.3–< 30 mm/year (Mulders, 2007). Experimental results from corrosion studies by Det Norske Veritas show that corrosion rates vary from 0.01 to 10 mm/year, depending on the presence of water and amines (Sridhar et al., 2008). These preliminary data were generated to determine the effect of small amounts of water in supercritical or dense phase CO2 and showed that the corrosion rate can be significant at water concentrations above 100 ppm by weight (10 mm/year) and that small amounts of amine carried over from capture processes may reduce the corrosion rate significantly (0.01 mm/year). In general, there is a paucity of experimental data on the corrosion rate of carbon steel in supercritical and dense phase CO2 containing water, amines and other impurities. The variations in experimental results for corrosion rate values suggest that the corrosion mechanisms involved are different, probably due to different experimental settings in the laboratory. The few data available in the literature on supercritical CO2 are relevant, to a large extent, to a water phase only.

14.9

Technologies for monitoring injection well integrity

14.9.1 Injection well integrity A number of standard technologies are available for monitoring the integrity of active injection wells. Cement bond logs are used to assess the bond and the continuity of the cement around the well casing. Periodic cement bond logs can help detect deterioration in the cemented portion of the well and may also indicate any chemical interaction of the acidic formation fluids with the cement. The initial use of cement bond logs as part of the well integrity testing can indicate problems with bonding and even the absence of cement. In the absence of cement bond log data, variable density log data may be used for the same purposes. Alternatively, the cement quality and top of cement might be predicted by cement job parameters, i.e. losses and hole size versus cement volume pumped.


460


Injection well integrity can also be monitored using geophysical well integrity logs. These can include casing corrosion logs and temperature and acoustics measurements. The minimum packer setting depth is determined by the minimum formation strength to withstand the maximum pressure during the well injection lifetime in order to avoid an underground blowout that cannot be detected with the surface monitoring systems. In addition to these, geophysical well integrity logs can provide measurements of downhole petrophysics, caprock pressure, saturation and fluid chemistry. Very detailed information provided around the well can be used to calibrate performance modelling. Prior to converting a well to other uses, such as CO2 injection, the well usually undergoes testing to ensure its integrity under pressure. These tests are relatively straightforward, with the well being sealed top and bottom (or in the zone to be tested), pressured up and its ability to hold pressure measured. In the case of storage in depleted oil and gas fields, the cement can be affected by reservoir subsidence and other geological instabilities in addition to normal degradation due to ageing. The fluid composition and pressure in the reservoir can change over time, exposing the cement to pressure and composition different from initial test conditions. This is the motivation for collecting baseline monitoring data prior to the injection of CO2. Injection takes place through a pipe that is lowered into the well and packed off above the perforations or open-hole portion of the well to ensure that the injectant reaches the appropriate level. The pressure in the annulus, the space between the casing and the injection pipe, can be monitored to ensure the integrity of the packer, casing and the injection pipe. Changes in pressure or gas composition in the annulus will alert the operator to problems. As noted above, the injection pressure is carefully monitored to ensure that there are no problems. A rapid increase in pressure could indicate problems with the well, although industry interpretations suggest that it is more likely to be loss of injectivity in the reservoir. Temperature logs and ‘noise’ logs are also often run on a routine basis to detect well failures in natural gas storage projects. The noise log or survey, also sometimes called the sound survey (e.g. Sonar Log, Borehole Audio Tracer Survey (BATS), Acoustic Sonde Log and others), is essentially a very sensitive detector of the sound produced by fluid flow. The sounds of moving fluids or the hiss of escaping gas are caused by disturbances in a liquid/gas interface or by turbulence in the fluid stream. Noise logs use very sensitive microphones to be able to measure sound produced by fluid flow in a well. Measurements of sound within the audible range of frequencies (20–20 000 Hz) are usually most indicative of turbulent flow behind pipes. Rapid changes in temperature along the length of the wellbore are diagnostic of



461

casing leaks. Similarly, ‘noise’ associated with leaks in the injection tubing can be used to locate small leaks (Lippmann and Benson, 2003).

14.9.2 Injection rates and pressures To ensure that injection of CO2 is taking place at the rates desired, monitoring of the condition of the injection well is necessary. Measurements of CO2 injection rates are a common oil field practice, and instruments for this purpose are available commercially. Measurements are made by gauges either at the injection wellhead or near distribution manifolds. Typical systems use orifice meters or other devices that relate the pressure drop across the device to the flow rate. Top and downhole performance logs are used to monitor injection rates, pressure and temperature changes. Downhole seismic surveys (vertical or offset) and seismic profiles along single or between boreholes may help to map fluid pressure distribution around the well. Measurements of injection pressure at the surface and in the formation are also routine processes. Pressure gauges are installed on most injection wells through orifices in the surface piping near the wellhead. Downhole pressure measurements are routine, but are used for injection well testing or under special circumstances in which surface measurements do not provide reliable information about the downhole pressure. A wide variety of pressure sensors are available and suitable for monitoring pressures at the wellhead or in the formation. These instruments are used to monitor injection pressures through shutoff valves that will stop or curtail injection if the pressure exceeds a pre-determined safe threshold or if there is a drop in pressure as a result of a leak. Surface pressures can be used to ensure that downhole pressures do not exceed the threshold of reservoir fracture pressure. Modern systems such as fibre-optic pressure and temperature sensors are expected to provide more reliable measurements and well control.

14.9.3 Microseismic monitoring Microseismic monitoring can detect induced seismicity due to CO2 injection and provide information on pore pressure change and geomechanical stability (fault/fracture creation or reactivation) within a few tens of meters or even a few meters caused by local stress accumulation. In addition to this, the use of microseismic monitoring techniques has potential value for wellbore integrity monitoring. Extremely small micro-earthquakes associated with rock fracturing or cement sheath crack may be detected using the most sensitive borehole monitoring arrays that are fully digitalized geophones connected or often cemented into the injection well. Discrete micro-earthquakes with magnitudes of the order of 3–4 on the Richter scale can be picked up by static arrays of sensors. Microseismic images from geophones in single or © Woodhead Publishing Limited, 2010

462


multiple monitoring wells may detect cracks in the cement sheath. In addition to borehole arrays, arrays of geophones can be deployed at the surface.

14.10 Future trends Many of the technologies required for large-scale geological storage of CO2 already exist. Current drilling and well construction practices for CO2 injection wells are based on existing knowledge and practices from the oil and gas industry. The technology for injection wells has evolved to a highly sophisticated state, such that it is now possible to drill and complete vertical and extended reach wells (including horizontal wells) in deep formations, wells with multiple completions and wells able to handle corrosive fluids. Cement blends that resist degradation are being developed for field application in the near future. Monitoring tools that can provide an early warning system are key elements in assessing wellbore integrity. Thus it is anticipated that improvements and development in monitoring technology will evolve as experience is gained from large-scale CO2 injection projects in the future. Corrosion is a potential threat for wellbore integrity, especially over long periods of time. Future activities should aim at understanding corrosion problems under reservoir thermodynamic conditions. This is among the major risks involved in CO2 storage engineering.

14.11 Sources of further information and advice The current state-of-the-art knowledge about the science and technology of underground fluid injection including a chapter devoted to CO2 injection is presented in a textbook edited by Tsang and Apps (2005) which we advise for use as a reference for developing knowledge about underground injection of fluids. The basic knowledge of CO2 geological storage is well documented in the IPCC (2005) special report, and we recommend this for those who wish to acquire a deep knowledge of the science of geological storage of CO2. Publications from the International Journal of Geenhouse Gas Control (www.elsevier.com/locate/ijggc), the Society of Petroleum Engineers (www. spe.org) and Oil & Gas Science and Technology – Revue de l’IFP (http:// ogst.ifp.fr) are important resources for developing further knowledge on all aspects related to CO2 storage in underground geological formations. Many of the issues from these journals are specifically devoted to wellbore integrity and CO2 injection. We further recommend readers to refer to these journals to acquire a through knowledge of underground injection and storage of CO2. Readers are highly recommended to look into relevant references in the text to get further knowledge about specific topics of interest.



463

14.12 Acknowledgements The contributors want to thank the Det Norske Veritas (DNV) Corporate for the financial and moral support and to all the authors whom their work is referred to prepare this material.

14.13 References Ahmad M U and Smith J A (1988) ‘Earthquakes, injection wells and the Perry Nuclear Power Plant, Cleveland, Ohio’, Geology, 16, 739–742. Al-Homadhi E S (2001) ‘The effect of the initial rock permeability on the extent of injectivity reduction due to brine injection through fractured formation’, Oil & Gas Science and Technology, 56(2), 135–143. Bachu S, Gunter W D and Perkins E H (1994) ‘Aquifer disposal of CO2: hydrodynamic and mineral trapping’, Energy Conversion and Management, 35(4), 269–279. Bradshaw J G, Allinson G, Bradshaw B E, Nguyen V, Rigg A J, Spencer L and Wilson P (2003) Australia’s CO2 geological storage potential and matching of emissions sources to potential sinks, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Oxford, UK, Elsevier (Pergamon), Vol. 1, 633–638. Cailly B, Le-Thiez P, Egermann P, Audibert A, Vidal-Gilbert S and Longaygue X (2005) ‘Geological storage of CO2: a state-of-the-art of injection processes and technologies’, Oil & Gas Science and Technology – Rev. IFP, 60(3), 517–525. Christman P G and Gorell S B (1990) ‘Comparison of laboratory and field-observed CO 2 tertiary injectivity’, Journal of Petroleum Technology, February, 226–233. Clark J E, Bonura D K and Voorhees R F V (2005) ‘An overview of injection well history in the United States of America’, in Tsang C F and Apps A J (eds), Underground Injection Science and Technology, Amsterdam, the Netherlands, Elsevier, 3–11. Dusseault M B, Gray M N and Nawrocki P A (2000) ‘Why oilwells leak: cement behavior and long-term consequences’, SPE 64733, International Oil and Gas Conference and Exhibition, 7–10 November, Beijing, China. Dusseault M B, Bruno M S and Barrera J (2001) ‘Casing shear: causes, cases’, cures, SPE Drilling & Completion, 16(2), 98–107. Ennis-King J and Paterson L (2001) Reservoir engineering issues in the geological disposal of carbon dioxide, in Williams D J, Durie R A, McMullan P, Paulson C A J and Smith A Y (eds), Proceedings of the Fifth International Conference on Greenhouse Gas Control Technologies: GHGT5, Collingwood, VIC, CSIRO Publishing, 290–295. EPA (2008) ‘Federal requirements under the underground injection control (UIC) program for carbon dioxide (CO2) geologic sequestration (GS) wells; proposed rule’, Federal Register, 73(144 ), 43492–43541. Fabriol H (2002) Feasibility study of microseismic monitoring (Task 5.8), BRGM Commissioned Report BRGM/RP-51293-FR (Confidential). Flett M A, Gurton R M and Taggart I J (2005) Heterogeneous saline formations: Longterm benefits for geo-sequestration of greenhouse gases, in Rubin E S, Keith D W and Gilboy C F (eds), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Cheltenham, UK, IEA GHG, Vol. 1, 501–510. Gray K E, Podnos E and Becker E (2009) ‘Finite element studies of near-wellbore region during cementing operations: Part I’, SPE Drilling & Completing, 24, 127–136. Grigg R B (2005) ‘Long-term CO2 storage: using petroleum industry experience’, in


464


Thomas DC and Benson SM (eds), Carbon Dioxide Capture for Storage in Deep Geologic Formations – Results from the CO2 capture project’, Oxford, UK, Elsevier, 853–866. Han H, Dusseault M and Xu B (2006) Simulation of tectonic deformation and large-area casing shear mechanisms – Part B: geomechanics1, 06–1004, The 41st U.S. Symposium on Rock Mechanics (USRMS): ‘50 Years of Rock Mechanics – Landmarks and Future Challenges’, June 17–21, Golden, CO. Healy J H, Ruby W W, Griggs D T and Raleigh C B (1968) ‘The Denver earthquakes’, Science, 161, 1301–1310. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage, Working Group III of the Intergovernmental Panel on Climate Change, Metz, B, Davidson, O, de Coninck, H C, Loos, M and Meyer, L A (eds), Cambridge, UK, Cambridge University Press. Jarrell P M, Fox C E, Stein M H and Webb S L (2002) Practical Aspects of CO2 Flooding, SPE Monograph Series No. 22, Richardson, TX, Society of Petroleum Engineers. Jimenez J A and Chalaturnyk R J (2003) Are disused hydrocarbon reservoirs safe for geological storage of CO2?, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Oxford, UK, Elsevier (Pergamon), Vol. 1, 471–476. Kaldi J G and Gibson-Poole C M (2008) Storage Capacity Estimation, Site Selection and Characterisation for CO2 Storage Projects, CO2CRC Report No. RPT08-1001, Canberra, ACT, Cooperative Research Centre for Greenhouse Gas Technologies (CO2CRC). Kleinitz W, Koehler M and Dietzsch G (2001) ‘The precipitation of salt in gas producing wells’, SPE 68953, SPE European Formation Damage Conference, 21–22 May, The Hague, the Netherlands. Korbol R and Kaddour A (1994) ‘Sleipner West CO2 disposal: injection of removed CO2 into the Utsira formation’, Energy Conversion and Management, 36(6–9), 509–512. Kutchko B, Strazisar B, Lowry G, Dzombak D and Thaulow N (2007) Impact of wellbore cement degradation on CO2 storage integrity, American Geophysical Union, 2007 Fall Meeting, 10–14 December, San Francisco, CA, abstract #U43C-1398. Law D H S and Bachu S (1996) ‘Hydrogeological and numerical analysis of CO2 disposal in deep aquifers in the Alberta Sedimentary Basin’, Energy Conversion and Management, 37(6–8), 1167–1174. Lippmann M J and Benson S M (2003) Relevance of underground natural gas storage to geologic sequestration of carbon dioxide, Department of Energy’s Information Bridge, US Government Printing Office (GPO), available at: http://www.osti.gov/energycitations/ servlets/purl/813565-MVm7Ve/native/813565.pdf (accessed December 2009). Magoon L B and Dow W G (1994) ‘The petroleum system’, American Association of Petroleum Geologists Bulletin, Memoir 60, 3–24. Mulders F (2007) Analysis of abandoned well integrity at a potential CO2 storage site, 3rd Well Bore Integrity Network Meeting, 12–13 March, Santa Fe, NM, available at: http://www.co2captureandstorage.info/docs/WBI3Presentations/FMulders.pdf (accessed December 2009). Müller N (2007) Subsurface mineralisation: halite precipitation during CO2 injection, Shell Exploratory Research (EPT-RXX), CATO Workpackage WP 4.1, available at: http:// www.co2-cato.nl/cato-2/publications/subsurface-mineralisation-halite-precipitationduring-co2-injection (accessed January 2010). Perry K F (2005) ‘Natural gas storage industry experience and technology: potential application to CO2 geological storage’, in Thomas DC and Benson SM (eds), Carbon



465

Dioxide Capture for Storage in Deep Geologic Formations—Results from the CO2 Capture Project, Oxford, UK, Elsevier London, Elsevier, 815–826. Pilkington P E (1992) ‘Cement evaluation–past, present, and future’, Journal of Petroleum Technology, February, 132–140. Riddiford F A, Tourqui A, Bishop C D, Taylor B and M Smith (2003) A cleaner development: The In Salah Gas Project, Algeria, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Oxford, UK, Elsevier (Pergamon), Vol. 1, 606–606. Rutqvist J and Tsang C F (2005) ‘Coupled hydromechanical effects of CO2 injection’, in Tsang C F and Apps A J (eds), Underground Injection Science and Technology, the Netherlands, Amsterdam, Elsevier, 649–679. Rybalchenko A I, Pimenov M K, Kurochkin V M, Kamnev E N, Korotkevich V M, Zubkov A A and Khafizov R R (2005) ‘Deep injection disposal of liquid radioactive waste in Russia, 1963–2002: results and consequences’, in Tsang C F and Apps A J (eds), Underground Injection Science and Technology, Amsterdam, the Netherlands, Elsevier, 13–19. Sedlacek R (1999) ‘Untertage Erdgasspeicherung in Europa’, Erdol, Erdgas Kohle, 115, 573–540. Shi J-Q and Durucan S (2005) A numerical simulation study of the Allison Unit CO2ECBM pilot: the effect of matrix shrinkage and swelling on ECBM production and CO2 injectivity, in Rubin E S, Keith D W and Gilboy C F (eds) (2005), Proceedings of the Seventh International Conference on Greenhouse Gas Control Technologies: GHGT7, Cheltenham, UK, IEA GHG, Vol. 1, 431–442. Simpson A J and Paige R W (1991) Advances in Forties Field water injection, The Offshore Europe, 3–6 September, Aberdeen, UK, SPE 23140. Skinner L (2003) ‘CO2 blowouts: an emerging problem’, World Oil, 224(1), 38–42. Solomon S, Carpenter M and Flach T A (2008) ‘Intermediate storage of carbon dioxide in geological formations: a technical perspective’, International Journal of Greenhouse Gas Control, 2, 502–510. Spycher N and Pruess K (2005) ‘CO2-H2O mixtures in the geological sequestration of CO2. II. partitioning in chloride brines at 12–100 °C and up to 600 bar’, Geochim. Cosmochim. Acta, 69(13), 3309–3320. Sridhar N, Thodla R and Ayello F (2008) Approaches to Assessment of Well Bore Leakage in CCS Sites, Report No. 8602008-04, Høvik, Norway, DNV Research & Innovation. Strazisar B, Kutchko B, Dzombak D, Lowry G and Thaulow N (2008) Degradation rate of well cement and effect of additives, Fourth Wellbore Integrity Network Meeting, 18–19 March, Paris, France. Streit J E and Hillis R R (2003) Building geomechanical models for the safe underground storage of carbon dioxide in porous rock, in Gale J and Kaya Y (eds), Proceedings of the Sixth International Conference on Greenhouse Gas Control Technologies: GHGT6, Oxford, UK, Elsevier (Pergamon), Vol. 1, 495–500. Tsang C F and Apps A J (2005) Underground Injection Science and Technology, Amsterdam, the Netherlands, Elsevier. USEPA (1994) Determination of Maximum Injection Pressure for Class I Wells, Region 5 – Underground Injection Control Section Regional Guidance #7. van der Meer L G H (1995) ‘The CO2 storage efficiency of aquifers’, Energy Conversion and Management, 36(6–9), 513–518. van der Meer L G H (1996) ‘Computer modeling of underground CO2 storage’, Energy Conversion and Management, 37(6–8), 1155–1160.


466


van der Meer B and Egberts P (2008) ‘Calculating subsurface CO2 storage capacities’, The Leading Age, April, 502–505. van eck T, Goutbeeka F, Haak H and Dosta B (2006) ‘Seismic hazard due to smallmagnitude, shallow-source, induced earthquakes in the Netherlands’, Engineering Geology, 87(1–2), 105–121.


15

Carbon dioxide (CO2) capture and storage technology in the cement and concrete industry S. G h o s h a l, McGill University, Canada and F. Z e m a n, New York Institute of Technology, USA Abstract: Carbon dioxide (CO2) capture technologies such as postcombustion capture, oxygen combustion and isolated calcination and mineral sequestration of CO2 in precast concrete products through accelerated CO2 curing appear to be promising avenues for managing CO2 emissions by the cement and concrete industry. The use of CO2 for carbonation of cementbased waste stabilization and solidification, and for carbonation of waste cement kiln dust and waste concrete also offers opportunities for mineral sequestration of CO2. These emerging technologies are discussed in this chapter. Key words: calcination CO2, CO2 capture, mineral sequestration of CO2, oxygen combustion, isolated calcination, post-combustion capture, CO2 storage, accelerated CO2 curing, CO2 uptake in cement.

15.1

Introduction

Concrete, an aggregate mixture bound with cement, is the basis for urban development. As a commodity, it is second only to water and, through its use in infrastructure, it is a vital component of the built environment. Current production is over 2 Giga tonnes (Gt) per year with growth projected at over 50 % by 2020 (CSI, 2002). There is no ready substitute for cement and, therefore, for concrete as a construction material that can be produced on a similar scale. The sustainability challenge for cement is the production of its main component, clinker. Clinker is produced using high-temperature pyroprocessing, which requires energy and releases significant amounts of carbon dioxide (CO2). Like many heavy industries, the cement plant derives its process energy from fossil fuels. However, cement also has significant process emissions, i.e. CO2 emissions not originating from the oxidation of fuel. In fact, the majority of greenhouse gases produced during the manufacturing of cement occur via the calcination or thermal decomposition of limestone, the dominant source material. The International Energy Agency (IEA) estimated that the cement industry released 1.8 Gt of CO2 in 2005, 6 % of the global total (EIA, 2007). The Cement Sustainability Initiative (CSI) estimates 469 © Woodhead Publishing Limited, 2010

470


that this value will grow to 2.9 Gt by 2020 and almost 5 Gt by 2050 (CSI, 2002). As a result, the greenhouse gas emissions from cement production are divided between calcination (50 %) and fuel consumption (40 %) with indirect emissions from electricity production and vehicular traffic forming the remainder (10 %) (Worrell et al., 2000). The fuel consumption is further split between the calcination reaction, which accounts for 60 %, and the kiln burner, which accounts for the remaining 40 % (Barker et al., 2009). Current efforts by the cement industry have focused on reducing the clinker content of cement (blending), using non-carbonate lime sources such as steel slag, and using alternate fuels, including potentially carbon neutral fuel sources such as waste sawdust, to provide heat for the kilns (CSI, 2002; van Oss and Padovani, 2003). These measures are cost-effective and should allow the industry to meet near-term targets but cannot provide the deep reductions necessary for climate stabilization. Addressing CO2 emission reductions through CO2 capture and storage technologies, specifically post-combustion capture and oxygen combustion, appears to be a promising route for the cement industry, particularly since the presence of calcination CO2 necessitates mitigation techniques beyond alternative fuels and renewable energy. Furthermore, the use of CO2 for accelerated curing of non-reinforced concrete modular building products or concrete blocks appears to be a viable method for mineral sequestration of captured CO2 and CO2 from flue gases. The use of CO2 for carbonation of cement-based waste stabilization and solidification and for carbonation of waste cement kiln dust also offers opportunities for mineral sequestration of CO2. These emerging technologies are discussed in this chapter.

15.2

Basic principles

The majority of the CO2 associated with cement production is associated with two distinct sources, the decomposition of limestone and the oxidation of fuels. In order to provide a common basis for the subsequent discussion, it is worth relating the composition of the various compounds associated with cement and concrete. A schematic representation of the various compounds, and their constituents, is presented in Fig. 15.1. The term ‘raw meal’ refers to the mixture of natural minerals fed into the cement kiln. It normally consists of a mixture of limestone-bearing rocks with the silica component derived from sands or clays. Clays used may also contain sufficient levels of iron and alumina. These latter two raw materials can be supplemented with high-purity compounds such as bauxite. The product of a cement kiln is referred to as clinker. The clinker is composed of a series of calcium-based minerals. We have used the standard industrial notation for calcium oxide (C), silica (S), alumina (A) and iron oxide (F). The clinker minerals are commonly referred to as alite (tricalcium silicate


CO2 capture and storage in cement and concrete industry 100 %

Air

Gypsum

Other

C4AF C3A

80 % Crushed stone Composition (wt %)

471

Limestone

C2S 60 %

Sand

40 %

Iron oxide Clinker

20 %

C3S

Water Alumina

Cement

Silica

0 %

Concrete

Cement

Clinker

Raw meal

15.1 Proportion of various components associated with cement and concrete use.

or C3S) and belite (dicalcium silicate or C2S) with the other components referred to as aluminate (tricalcium aluminate or C3A) and ferrite (calcium alumino-ferrites or C4AF). Once it is ground up with gypsum, it is referred to as cement. A mixture of 90 % clinker and 10 % gypsum is referred to as Ordinary Portland Cement (OPC). Concrete is an aggregate mixture that uses cement as the binding agent. A typical cement kiln is designed in a counter current fashion to provide suitable heat exchange and maximize efficiency. Beginning at ambient temperatures, the raw meal is fed into the exhaust end of the kiln resulting in the transfer of heat from the combustion gases and the incoming meal. The heating of the raw meal results in the evaporation of water and volatilization of sulphur compounds. The calcination of limestone is the first high-temperature reaction and is a precursor to the clinker formation process as the lime required for reaction with the silica is a product of the thermal decomposition of limestone, according to Equation 15.1. The mass loss associated with the generation of CO2 during calcination necessitates the processing of at least 1.4 t of raw meal, containing over 1.1 t of limestone, for the production of 1 t of clinker.

CaCO3(s) Æ CaO(s) + CO2(g) DH° =178.3 kJ/mol

[15.1]

The enthalpy or heat required to drive the calcination in Equation 15.1 is the sum of the sensible heat required to bring the raw materials to reaction


472


temperature (13.2 kJ/mol) and the energy to drive the reaction (165.1 kJ/ mol) (Lide, 2000). In total, the reaction requires the addition of 2.0 Mega joules (MJ) of heat per kg of clinker to the process at temperatures around 800–850 °C. It is worthwhile noting that the theoretical energy penalty for clinker production is 1.76 MJ/kg clinker (Worrell et al., 2000). The difference arises from the fact that the clinker formation reactions, the reaction of the lime and silica, are exothermic. In addition, the clinker formation occurs at 1400 °C, well above calcination temperatures and downstream of the calcination reaction, which precludes direct heat transfer. The most efficient kilns today consume slightly less 3.0 MJ/kg of clinker with energy lost through the kiln walls, the exhaust gases and the hot clinker leaving the kiln. Modern kiln designs have resulted in a significant reduction in energy consumption per unit clinker. The gains have been achieved by altering the design from a long kiln, either wet or dry, to the multistage pre-heater, precalciner kiln shown in Fig. 15.2. A pre-calciner kiln contains a tertiary air duct that moves air to the bottom of the riser tower. The majority (60 %) of the fuel used in the plant is combusted in the pre-calciner (ECRA, 2007). The significant design changes from the long kilns are the use of a suspension pre-heater for transferring heat from the exhaust gases to the incoming raw meal, the addition of a separate calciner and more efficient clinker coolers to transfer heat to the incoming air. Overall, the unit energy consumption has been reduced from 4.9 MJ/kg clinker in 1968 (Cembureau, 1998) to less than 3.0 MJ today (ECRA, 2007). The near-term CO2 emissions reduction strategy for the cement industry involves improving plant efficiency, using alternative fuels and reducing the clinker content of cement (blending). Improving plant efficiency has already been implemented in much of the developed world. Alternative fuels could, in theory, avoid all fuel emissions but will depend on the availability and content of such fuels. The potential for blending has been estimated between 5–20 %, again depending on availability (Worrell et al., 2001). Each of these options is discussed in the recent ‘Technology Papers’ issued by CSI/ECRA (CSI/ECRA, 2009).

15.3

Capture of carbon dioxide (CO2) from cement plants

Cement plants are similar to power plants in that they produce exhaust streams dominated by nitrogen containing significant amounts of CO2 at atmospheric pressure. There are, however, several differences worth noting. The first is that the CO2 concentration in the flue gas is noticeably higher in a cement plant (14–33 %) as compared to a natural gas turbine (3–4 %) or a coal-fired boiler (12–14 %) (ECRA, 2007). Second, a cement plant does not contain a low-pressure steam cycle, as in a coal power plant or natural gas combined


Environmental aspects Grinding ∑ Dust ∑ Noise ∑ Electricity

Clinker production ∑ Dust ∑ Gases: SO2, NOx, CO2, micropolutants ∑ Noise ∑ Heat ∑ Fuels

Grinding cement ∑ Dust ∑ Noise ∑ Electricity ∑ Raw materials

Storage/shipping ∑ Dust ∑ Noise ∑ Fuel

Cement silo Cement grinding Additions Clinker storage Cooler Rotary kiln

Pre-heating Quarries

Grinding

Drill Pre-homogenizaton Crushing Dumper

Quarry and dry-process cement plant

Social aspects

Economic aspects ∑ Shareholder returns ∑ Local taxes & wages ∑ Suppliers businesses ∑ Widely used product ∑ Community investment

15.2 General layout of a modern pre-heater/precalciner dry rotary kiln; R Rivet (CSI, 2002).

473

∑ Employment ∑ Health and safety ∑ Training ∑ Local community impacts

CO2 capture and storage in cement and concrete industry


Quarrying ∑ Dust ∑ Noise ∑ Vibration ∑ Landscape impact ∑ Raw materials: limestone clay sand

474


cycle system. The low-pressure steam loop is often considered an energy source for the post-combustion capture of CO2. Third, the cement plant uses its exhaust gases for direct heat exchange to dry the incoming raw meal. Fourth, CO2 is produced from two distinct sources in a cement plant, fuel oxidation and limestone decomposition. The dual origin of cement CO2 has some important ramifications for potential CO2 capture technologies. As discussed in Chapter 1, pre-combustion capture is one of the main technologies for CO2 capture and storage (CCS). The basic premise of this category is to de-carbonize the energy carrier, often by conversion to hydrogen, so that the exhaust gases do not contain CO2. Assuming hydrogen could be used in a cement kiln, the exhaust gases would still contain the CO2 released during calcination. In addition, hydrogen flames have very different thermal properties when compared to coal burners. As a result, analyses of CO2 capture for cement plants generally conclude that pre-combustion capture is not a suitable technology. The other options, post-combustion and oxygen combustion, are both considered feasible for the cement industry although much research is still required.

15.3.1 Post-combustion capture The use of post-combustion capture systems is an attractive option for the cement industry, as it does not involve any modifications to the existing plant. It can be viewed as an additional gas clean up added after the removal of category pollutants, such as SOx and NOx. The feasibility of such a system was investigated for a coastal cement plant in Norway (Hegerland et al., 2007). The study provides a good summary of the modifications required for a cement plant. The exhaust gases can be diverted to a capture plant prior to reaching the exhaust stack. The stack gases will require additional cleaning to remove any remnants of SOx or NOx, which accelerate sorbent degradation. Cement plants also contain cement kiln dust (CKD), an alkaline mixture of the raw meal. These must also be virtually eliminated. The improved gas clean up can be accomplished by upgrading existing equipment or using additional equipment. The latter may be more economical if a slipstream system is used, i.e. a portion of the exhaust is diverted for CO2 capture. The system considered by Hegerland et al. (2007) consisted of an amine sorbent regenerated using low-pressure steam. The lowest cost sorbent was chosen to minimize sorbent replacement costs owing to the high dust load in cement exhausts. A small portion of the steam load could be met by heat recovery/steam generation from the exhaust gases, typically at temperatures above 300 °C. A boiler, however, generated the bulk of the steam. In the study, the boiler was fired with fossil fuels resulting in a 25 % increase in the amount of CO2 sent to storage. If the fuel is coal, then additional gas cleaning capacity would be required as the boiler exhaust gases are also



475

fed into the CO2 capture system. The capture system is expected to remove the vast majority of the CO2 entering the scrubber, normally estimated at 85–90 %. The addition of the boiler CO2 reduces the amount of CO2 avoided to approximately 77 %. The cost of this system is highly dependent on the cost of the fuel and amount of additional gas clean up required. A separate study of the post-combustion option concluded that, given the additional fuel needed for the boiler, it is worth considering constructing a power plant to generate electricity for the plant. This would provide a low-pressure cycle for sorbent regeneration and offset emissions associated with power generation as well (IEA, 2008). The addition of the power plant results in a 360 % increase in coal consumption yet leads to a similar (77 %) level of avoided CO2 emissions. The addition of the power plant doubles the cost of cement and would require significant carbon prices to justify its implementation. Post-combustion capture technologies are being seriously considered in the power sector, which produces order of magnitude more CO2 emissions than the cement sector. As a result, the cement industry will benefit from research and development sponsored by the power sector. Although CO2 capture using amine sorbents is considered the most cost-effective postcapture system available, significant advances are being made for Ca/Mg carbonates, engineered membrane and sorbent-based gas separation systems (Ebner and Ritter, 2009). Post-combustion capture systems are likely to be feasible but result in an increase in the amount of CO2 sent to storage and do not produce any benefits for the cement company. As such, their feasibility will be directly tied to the price of CO2 emissions.

15.3.2 Oxygen combustion Oxygen combustion is another method for capturing the CO2 from cement plants where the gas separation occurs upstream of the kiln as opposed to post-combustion capture, which is essentially an additional gas clean up process. The important difference is that the gases separated are oxygen and nitrogen instead of CO2 and nitrogen. This involves changes in the clinker production process, which poses challenges but also creates opportunities for efficiency improvements. The end result of oxygen combustion is a gas stream dominated by CO2 and therefore suitable for storage. In oxygen combustion, high-purity oxygen is used for fuel oxidization as opposed to air. The removal of nitrogen from the system results in the potential for very high flame temperatures, as no thermal ballast is present. This requires the recycling of stack gases back to the burner to control temperatures in the kiln. The result is that the CO2 concentration in the kiln will rise from 20–30 % to over 80 %. The effect of raising the partial pressure of CO2 on the calcination (thermal decomposition) of limestone


476


has been understood for almost a century (Johnston, 1910). The temperature at which the reaction (Equation 15.1) proceeds (i.e. CO2 is released) rises from 670 °C at pCO2 = 0 atm to 900 °C at one atmosphere of CO2. This necessitates operating the pre-calciner at higher temperatures to ensure complete calcination of the raw meal. The effect of raising the CO2 partial pressure on overall clinker formation is not known at this time. More work is needed to fully categorize the quality of clinker produced under high CO2 concentrations. Oxygen is produced in quantities sufficient to ensure the complete oxidation of all the fuel used in the process. As such, it is only indirectly tied to the CO2 released during the calcination reaction, which can be viewed as a by-product of combustion. In contrast, post-combustion systems require sorbent to capture both combustion and calcination CO2. As a result, the cement industry is considered the best location for the use of oxygen for CCS (Gronkvist et al., 2006). Furthermore, oxygen enrichment is already practised in the cement industry as a method to increase production from existing kilns (Wrampe and Rolseth, 1976). Oxygen combustion, therefore, combines a theoretically lower energy penalty, albeit in the form of electricity as opposed to fuel, with an increase in production providing some additional revenues to offset costs. The dual origin of CO2 produced in cement manufacturing allows for a dual approach to reducing CO2 emissions. One option is to separate the calcination and clinker formation reactions and produce oxygen solely for calcination purposes. In a modern kiln, this has the advantage of reducing the amount of oxygen required, the percentage of the process requiring isolation from the atmosphere and the required modifications to the kiln. The alternative is to produce enough oxygen to oxidize all of the fuel used in cement manufacturing. The advantage here is that production increases are possible and the majority of the CO2 emissions can be mitigated. The disadvantage is that the entire kiln system must be isolated from the atmosphere to prevent air infiltration including the rotating kiln. Encasing the kiln in a shell has been considered in the past, for heat recovery (Lusche et al., 1982). While oxygen production using cryogenic air separation unit (ASU) is an established technology, its application in CCS involves some production changes (Pfaff and Kather, 2009). There is no co-production of nitrogen associated with CCS and the oxygen can be delivered at atmospheric pressure, which reduces the energy consumption of the process. In addition, the purity can potentially be lowered resulting in further energy savings. The most cost-effective ASU is a distillation system separating oxygen, nitrogen, and argon, gases with similar boiling points. The similar boiling points require a complex process to produce high-purity streams. In CCS, the produced CO2 is generally liquefied, via increased pressure, for transportation via pipelines. At this point, incondensable components such as nitrogen and argon can be



477

separated from the CO2. The final purity level required is be a function of the cost of producing higher purity oxygen, the effects of the impurities (Ar, N2) on the process, and the cost of CO2 purification associated with those impurities. Calcination in atmospheres dominated by CO2 is also known to produce highly sintered lime (Garcia-Labiano et al., 2002). Sintering is a surface phenomenon that reduces the surface area of the lime particles and thus limits reactivity with the CO2 in the gas phase. The effect of the higher CO2 partial pressures on the production of clinker has been investigated (Zeman, 2008). Additional experiments demonstrated that no carbonation of the free lime in the clinker occurs during cooling in a CO2 environment. This supports the use of CO2, recycled from the exhaust, as the cooling fluid in the clinker cooler section of the plant.

15.3.3 Isolated calcination Although conceptually similar to oxygen combustion, isolated calcination is fundamentally different. It is more appropriately viewed as a pre-treatment of the raw meal rather than the conversion of the clinker production process. The premise of this option is to calcine the limestone separately from the cement kiln and, in so doing, mitigate the dominant source of CO2 (66 % including associated fuel) without redesigning the kiln itself. The limestone calcination occurs in a separate facility with CCS. The lime is then conveyed to the pre-heater and mixed with the sources of silica, alumina and iron prior to being fed into the kiln. A schematic of the isolated calcination system is presented in Fig. 15.3. The substitution of lime for limestone in the raw meal will have some effects on the operation of the cement kiln. These include smaller particles entering the pre-heater and a smaller portion of the cooler air entering the kiln through the elimination of tertiary air. It can also be expected that the pre-heater tower will be less efficient owing to the reduced mass flow. Approximately one third of the raw meal mass is associated with the bound CO2. Isolated calcination will require a more detailed study of the transfer from the calciner to the kiln. A central question will be the leakage of CO2 to the environment, which occurs any time conditions inside a reactor are significantly different from ambient conditions. The calciner will contain lime in a high-concentration CO2 atmosphere at temperatures above 900 °C. The lime will have to be separated and introduced into the raw meal. There are two approaches to separating the lime when considering heat transfer. One is to separate at high temperatures and introduce the hot lime into the kiln. This would have the effect of minimizing re-carbonation, transferring heat to the kiln and reducing the heat transfer in the pre-heater. The alternative is


478

Developments and innovation in CCS technology CO2 recycle Fuel, O2 Oxygen lime kiln CO2 storage

CaCO3

CaO

Fuel, air

S, A and F Conventional cement kiln

Clinker

15.3 Schematic representation of isolated calcination.

to cool the lime at the calciner, in order to maximize the thermal efficiency of that unit. This would introduce low-temperature lime to the kiln but may result in re-carbonation (i.e. lower kiln efficiency), improved heat transfer in the pre-heater, limestone in the raw meal (i.e. higher kiln fuel usage) and cooler air. The choice of operating parameters discussed above offers some flexibility to isolated calcination. In addition, the size of the separate calciner is not absolutely linked to kiln capacity, meaning that isolated calcination can be viewed in a manner similar to slip streaming for post-combustion capture. The calciner can be sized to produce any percentage of the lime required for the kiln, from 0–100 %, in order to meet current regulatory requirements. If a self-contained unit is constructed, then modular designs may be feasible to deal with eventual tightening of CO2 emissions allowances, i.e. higher carbon costs. Barker et al. (2009) considered the case of isolated calcination in their report to the IEA (IEA, 2008). The electricity consumption increased by 120 %, mainly for oxygen production, and CO2 compression and purification. The conversion avoided 51 % of CO2 emissions at an estimated cost of $34 per tonne and an avoided cost of $40 per tonne. The avoided cost refers to the total cost of the project over the difference in emissions between the base case and the case with CCS. The amount of CO2 avoided is usually less than the amount captured owing to increased energy consumption and life-cycle emissions. As such, isolated calcination may be an effective nearterm option; however, it is unable to reduce CO2 emissions from the cement industry to near zero levels.



15.4

479

Accelerated carbon dioxide (CO2) curing of concrete

The proof of the concept of accelerated CO2 curing of concrete has been known since the 1970s (Young et al., 1974), but it is only recently being considered for commercial implementation, as a result of growing interest in CO2 sequestration. In this process, CO2 can be sequestered in mineral form during manufacturing of prefabricated concrete products and concrete masonry units. To date, CO2 sequestration has been carried out primarily in laboratory studies, which are aimed at assessing the performance of the CO2 cured concrete products and the rate and extent of carbonation. Cement is a potential CO2 sink because of the presence of calcium silicates (Fig. 15.1), which can be reacted with CO2 to form thermodynamically stable calcium carbonates and associated silica gel. Accelerated CO2 curing of concrete involves curing or hardening the cement, water and aggregate mixture in fresh concrete by exposure to a gas rich in CO2 (e.g., waste gas streams captured from cement plants). The calcium silicates in concrete react with dissolved CO2 in the presence of moisture while the silica aggregates are inert to the carbonation reaction and only affect the physical parameters of the concrete such as porosity, strength and permeability. The application of accelerated CO2 curing is limited to non-reinforced concrete, such as concrete blocks or cement boards reinforced with fibers. Such a technology may reduce total CO2 emissions from major point sources while developing value-added products. Theoretically, production lines for cement-based products can be installed next to the cement kilns or large stationary sources of CO2, e.g., power plants, in order to achieve the economical and environmental advantages of this technology. Approximately 14 million tonnes of cement is estimated to be used currently for manufacturing of precast concrete products in the USA and Canada, and these products are candidates for accelerated CO2 curing (Monkman, 2008). Accelerated CO2 curing of concrete is an active curing process for concrete production, and does not account for the CO2 uptake that may occur passively from atmospheric CO2 uptake in cured, hydrated concrete products. In the latter process, calcium hydroxide formed from hydration of calcium silicates reacts with atmospheric CO2, and this carbonation reaction lowers the pH of pore waters and leads to corrosion of steel reinforcement members (Shao et al., 2006a). Accelerated curing of concrete is currently typically achieved in industrial production by steam curing of fresh concrete mixtures in autoclaves, and is a common technique for production of modular concrete construction products and concrete blocks. In this method, the heat from the steam causes rapid hydration reactions and cement setting. However, the production of steam is energy-intensive, whereas accelerated CO2 curing that utilizes CO2 emissions


480


can reduce overall greenhouse gas emissions and achieve more rapid curing than with steam. In a recent study, Shi and Wu (2008) demonstrated that the four hour carbonation of moulded fresh concrete had compressive strengths comparable to that of the same concrete mix that was steam cured for 20 hours. The durability of carbonation cured concrete has been studied by Logan (2006). The carbonation curing was found to reduce the weathering carbonation shrinkage of concrete by 33 % compared to hydration curing when CO2 mass uptake in cement was only 10 %. This level of carbonation also reduced the mass loss during freeze/thaw durability testing by 90 %.

15.4.1 Chemical reactions in carbon dioxide (CO2) curing The components of cement subject to carbonation are calcium silicates C3S and C2S, and the following reactions are reported to occur (Papadakis et al., 1989):

3CaO·SiO2 + 3CO2(aq) + mH2O Æ SiO2·mH2O + 3CaCO3(s) [15.2]

2CaO·SiO2 + 2CO2(aq) + mH2O Æ SiO2·mH2O + 2CaCO3(s) [15.3]

Some of the C3S and C2S are hydrated to calcium silicate CSH, and this is a co-product along with calcium carbonate, but extended carbonation eventually converts the CSH to calcium carbonate and silica gel as follows (Fernandez Bertos et al., 2004):

3CaO·2SiO2·3H2O + 3CO2(aq) Æ SiO2(gel) + 3CaCO3 + 3H2O

[15.4]

Detailed reaction stoichiometry and kinetics for the above reactions have been discussed by Young et al. (1974), Fernandez Bertos et al. (2004) and Goodbrake et al. (1979). The reactions noted above are exothermic and occur spontaneously.

15.4.2 The carbonation process A schematic for the carbonation process is shown in Fig. 15.4. Prior to carbonation, the fresh concrete mix of cement, water and aggregates (e.g. sand and gravel) is compacted in a mould to form (harden) the product to be carbonated and maintained for up to several hours. Calcium silicates and water would react to primarily form CSH during this initial hydration period (Lea, 1970). Following this, the compacted samples are extracted from the mould and placed in the carbonation chamber. Various types of carbonation chambers have been employed, as shown in Fig. 15.4. The carbonation chamber Type A is a batch reactor connected to a CO2 source, which feeds CO2 in the chamber. The CO2 pressure in the chamber drops as


CO2 capture and storage in cement and concrete industry Step 1

481

Step 2 Mixer

Pressure

Cement Sand

Compaction

Water Step 3 A

Pressure gauge

B

Pressure gauge

C CO2 flow

CO2 gas

Concrete

CO2 gas

Concrete

CO2 analyzer Concrete sample

CO2 gas

15.4 The sequential steps involved in typical accelerated CO2 curing of concrete. Step 1: preparation of fresh concrete mixture. Step 2: compaction of the fresh mixture in a mould. Step 3: carbonation of compacted concrete in various reactor configurations.

CO2 is absorbed by the concrete sample. Such reactors have been employed by the majority of the studies (Shao et al., 2006b; Shi and Wu, 2008). In Type A reactors, the CO2 diffuses from the headspace in the samples, and the CO2 uptake in the pores provides a sustained driving force for CO2 diffusion into the concrete sample. The heat of the exothermic carbonation inside the sample may cause migration of moisture from the sample to the headspace in this type of carbonation chamber, negatively affecting the extent of carbonation reaction. Type B carbonation chambers operate in a similar manner, but the CO2 is made to flow through the chamber. Humidifying the influent CO2 may allow for regulating moisture loss and for reducing the temperatures in the sample. Young et al. (1974) have employed such carbonation reactors. Type C reaction chambers are different from the two other types because the advection of CO2 gas occurs through the concrete sample, rather than over it. This provides uniform distribution of CO2 through the concrete sample, and high CO2 pressures are not required to ensure diffusion of CO2 in the concrete matrix. Such reactors have been employed by Huntzinger et al. (2009) and Kashev-Haghigi and Ghoshal (2010). A closed reaction chamber (autoclave) or Type A reactor pressurized with a CO2-rich gas is commonly used for carbonation of the moulded concrete products (Monkman and Shao, 2006; Shi and Wu, 2008). Carbonation is performed over several hours during which the carbonation reactions as


482


described by Equations 15.2–15.4 occur. To ensure a significant extent of carbonation, maximum CO2 concentrations in the headspace of the reaction chamber should be maintained by ensuring a continuous supply of the CO2rich gas supply.

15.4.3 Critical operational parameters in accelerated (carbon dioxide (CO2) curing The extent and kinetics of carbonation may be significantly influenced by pore water content of the fresh concrete sample, the CO2 concentration in the gas phase, the porosity of the fresh concrete moulded product and the amount of alkali metals in the specific cement blend and in the aggregates. Water is essential for the carbonation reaction as CO2 gas and calcium ions from the cement minerals need to be dissolved in water. With increasing water volume in the sample, greater mass of both reactants are available with a higher amount of water in the concrete. However, as the pore water saturation increases, there is less pore volume available for CO2 gas penetration in the sample and thus the rate of CO2 dissolution decreases, resulting in a lower carbonation reaction rate. Young et al. (1974) reported one of the first studies which used CO2 for accelerated curing of 1 inch diameter calcium silicate mortars. They studied the compressive strength of compacted mortar samples of C3S and C2S after a few minutes exposed to CO2. Strength developed rapidly in compacted samples after 81 minutes of exposure, but the rate of carbonation was different in samples depending on the water content in the sample and gas phase. Samples maintained at a water:calcium silicate ratio of 0.125 and carbonated with a pure CO2 stream of relative humidity of 50 % provided the fastest rate of compressive strength development. Klemm and Berger (1972) found the optimal water:cement ratio to be 0.09 in similar sized samples of cement mortars (Klemm and Berger, 1972). The ideal relative humidity for carbonation reactions was also reported between 50 % and 70 % in Type A and B reactors (Burkan Isgor and Razaqpur, 2004). At lower RH values, CO2 cannot be dissolved adequately and at higher humidity, CO2 diffusion in pores is inhibited by water. The CO2 concentration plays a significant role in the extent of carbonation. Shao et al. (2006b) compared the carbonation of 14 mm thick concrete samples under pure CO2 and under a 25 % CO2–75 % N2 mixture at 0.5 MPa pressure in a Type A reactor. Carbonation under a pure CO2 atmosphere provided for uniform carbonation across the thickness of the concrete samples, the total CO2 uptake was 16 % by weight of the cement after two hours of CO2 curing and the carbonation extent was uniform in the core and the outer surface of the sample. In contrast, with 25 % CO2 atmosphere, the CO2 uptake was only 2 % by weight of the cement in the interior of the concrete sample,



483

whereas the outer layers of the concrete sample had a CO2 uptake comparable to the sample carbonated in a pure CO2 atmosphere. The strength gain of the concrete sample exposed to the 25 % CO2 atmosphere was significantly lower. However, a longer carbonation period of these samples should provide higher and more uniform CO2 uptake. Increasing CO2 concentrations can increase CO2 uptake to only a limited amount, as discussed in the section below. Shi and Wu (2008) found that there were small but measurable increases in the CO2 uptake when concrete samples were exposed to pure CO2 at pressures ranging from 0.07–0.4 MPa, and when the duration of carbonation was increased. The maximum CO2 uptake was approximately 17 % by weight of cement.

15.4.4 Potential limitations to carbon dioxide (CO2) uptake The CO2 mass uptake during accelerated CO2 curing has been reported to range between 10 and 15 % of the dry mass of cement (Monkman and Shao, 2006; Shi and Wu, 2008). This suggests that only 20–30 % of the carbonation efficiency is achievable. In 2004, in the USA alone over 16.4 million metric tonnes (mt) of cement was delivered to concrete product manufacturers as reported by the USGS (van Oss, 2004). Of this, 6.4 Mt was for brick and block production, 3.5 Mt for precast and prestressed, 2.1 Mt for pipe and 4.2 Mt for other concrete products. Therefore, at the currently achieved CO2 uptake of 15 % by weight of cement in the bricks and blocks produced in the USA in 2004 can reach up to – 1 Mt compared to the 45.7 Mt CO2 emission from cement manufacture in USA, based on 2004 data (EIA, 2007). This represents a 2.2 % reduction in the cement-related CO2 emissions in the USA. Although this may seem to be a small fraction of the total emissions, it should be noted that the accelerated CO2 curing of concrete produces a value-added product. Furthermore, there are significant opportunities for improving the CO2 uptake efficiency during accelerated CO2 curing, and for the usage of carbonated aggregates. The CO2 uptake potential in cement is a function of the relative mass of oxides present in raw material (XCaO,MgO,SO3,Na2O,K2O) and is estimated by Steinour formula (Steinour, 1959):

XCO2,Tot = 0.785 (XCaO – 0.700XSO3) + 1.091XMgO

+ 1.420XNa2O + 0.935XK2O

where:

XCO2,Tot = CO2 uptake capacity (wt%)

=

wt of CO2 theoretically taken up wt of cement


[15.5]

484


Based on the relative mass of oxides within a typical cement, e.g., Type 10 Portland cement, the theoretical CO2 uptake capacity (XCO2,Tot) of cement was found to be 49.6 wt %, i.e. CO2 can be reacted up to half of cement’s weight. The significantly lower uptake observed during carbonation experiments is attributable to several reasons: (i) deposition of calcium carbonate precipitates on calcium silicate surfaces leading to limited access of carbonic acid to the reactive surface area; (ii) pore blockages resulting from precipitated solids during carbonation leading to losses in reactive surface areas; and (iii) loss of water from the concrete sample due to temperature increases due to the exothermic carbonation reactions. These processes are illustrated in Fig. 15.5. The loss of water is more likely to be higher in reaction chambers that are operated without CO2 gas flow such as in Fig. 15.5(a), because of limited opportunity for temperature reductions resulting from heat outflow from the reactor with the exiting gases. Longer periods of accelerated CO2 curing and higher CO2 concentrations supplied to the reactor may provide marginal improvements to carbonation efficiency (Shi and Wu, 2008) by improving the diffusive mass transport of CO2 through calcium carbonate deposits or into small and restricted pores. Monkman (2008) suggested that one potential approach for enhancing the CO2 uptake in concrete products during manufacturing is the substitution of sand and other inert aggregates with appropriate aggregate materials (e.g. ladle slag fines, a waste product of steel manufacturing) that are rich in alkali metals (Monkman, 2008). Precarbonation of the aggregates, prior to mixing of the fresh concrete, and CO2 curing of the resulting aggregate mixture, provided an overall doubling of the amount of CO2 uptake in the concrete.

Solid phase (cement/aggregate)

CO

Liquid phase (pore water) Gas phase (air/CO2) Carbonation deposit CO2 flow

15.5 Schematic of possible pore structure changes with calcium carbonate deposition in pore surfaces following accelerated CO2 curing. Circles denote examples of calcium carbonate deposition that lead to pore blockage and reduced access to reactive calcium silicates surfaces in the cement matrix.



485

15.4.5 Potential for direct carbonation curing of concrete using flue gases Carbonation curing by pure CO2 requires the capture of CO2 from the industry flue gas, which is fairly costly with the current technologies. From the existing CO2 capturing technologies, chemical gas adsorption technology seems to be more economical compared to other separation processes such as cryogenic separation, dry bed absorption and membrane separation (Zhou, 2005). In the cement industry, the CO2 capture costs are estimated to reach up to US$50–250 per tonne CO2 (Holloway, 2005). Therefore, the potential of using flue gas directly from cement plants and/or other large stationary sources is of interest in accelerated CO2 curing of concrete. The carbonation process with flue gas involves some addition chemical reactions because of SO2 and NO2. SO2 is dissolved in the pore water and then dissociates. The HSO3– ion is then oxidized to SO42–, which reacts with calcium ion to form gypsum (CaSO4 · 2H2O) (Gómez et al., 2007). Therefore, gypsum and calcium carbonate are the two main solid products that will thus be formed during carbonation, and will also contribute to concrete pore blockage and reduction in reactive surface area of concrete. Typical cement plant flue gases contain SOx and NOx concentrations that are three orders of magnitude lower than the CO2 concentrations. As a result, the presence of these gases in the flue gas is unlikely to affect the mass transport of CO2 and, they are unlikely to be important reactive components in the carbonation process. The carbonation curing of cement products by as-captured flue gas from a typical cement kiln was compared with pure CO2 gas by Wang (2007). It was found that the carbonation curing by pure CO2 had some detrimental effects on the long-term strength of concrete, mainly because of the high water loss, the difficulty of water to diffuse into the densified cement matrix formed due to the high carbonation rate and also the formation of microcracks in cement paste due to the high heat generated during carbonation. Thus accelerated CO2 curing carbonation of concrete by flue gases may also provide some performance benefits.

15.4.6 Carbon dioxide (CO2) curing of waste cement materials There are significant opportunities for CO2 uptake through accelerated CO2 curing in waste materials. A primary example is CKD which is produced at the rate of approximately 0.15–0.2 per tonne of clinker produced (van Oss and Padovani, 2003). In a recent study Huntzinger et al. (2009) demonstrated that the CO2 uptake efficiency is approximately 80 % in CKD during accelerated CO2 curing, thus providing a sequestration capacity of up to 10 % of the


486


CO2 emitted in the calcination process, or 5 % of total emissions, from the US cement industry. Reinforced concrete waste is a large component of construction debris with an estimated material flow of 68 Mt per year in the US alone (Stolaroff et al., 2005). Stolaroff et al. (2005) suggest an active carbonation system where concrete waste is submerged in water, which is circulated through a spray system. The alkaline water spray removes CO2 from the air, which then precipitates as calcite or reacts with the calcium in the waste materials. The active carbonation of concrete waste was estimated to mitigate 6 % of industry emissions at a cost of $8 per tonne CO2. This estimate was based on a preliminary analysis that did not include any costs associated with grinding and/or haulage. Fernandez Bertos et al. (2004) suggest that there are significant opportunities for accelerated CO2 curing during the solidification and stabilizations of soils and sludge contaminated by toxic metal pollutants in cement matrices. Thus, the accelerated CO2 curing technology can be applied to processes other than the manufacture of concrete masonry units.

15.5

Future trends

The cement industry, like most heavy industries, is undergoing a global shift in production from the developed to developing markets. The existing trend is for the cement industry to find alternate fuels for the kilns and produce blended cements as means for reducing the CO2 associated with concrete. These trends will likely continue as long as sources of fuel and cementitious materials are available. There may also be a consolidation in the number of facilities owing to consolidation in the industry. If the dominant cement manufacturers continue to purchase smaller companies, it is reasonable to assume that production will also be consolidated in larger plants, such as the Ste. Genevieve facility in Missouri. Much of the development of CCS in the cement industry will be tied to developments in the carbon markets or the implementation of a carbon tax. Much of this development will depend on post-Kyoto agreements. In any case, it is unlikely that carbon prices will rise to the levels needed for the implementation of CCS, greater than $50 per tonne. Lower prices will spur the deployment of high-efficiency fans and motors as well as improvements in facility maintenance such as leakage prevention. These types of emissions reductions can be considered more beneficial as they avoid storing CO2 underground, or in another form, that will require continuous monitoring. On the other hand, accelerated CO2 curing of precast concrete products, particularly by flue gases, produces high-performance concrete products, and thus is not significantly limited by the lack of CO 2 incentives for implementation.



487

CO2 capture initiatives discussed above, in conjunction with mineral sequestration of CO2 in precast concrete products and cementitious wastes, and with the use of alternative fuels that have low carbon or are carbon neutral, there is a theoretical possibility of cement plants evolving into a zero emissions facility in the future.

15.6

Conclusions

The greenhouse gas emissions from the global cement industry are second only to the power sector and will eventually require mitigation. More importantly, the cement industry does not have readily available alternatives such as wind or nuclear power. While some near-term options, such as efficiency, alternative fuels and blending, offer some reductions, the drastic reductions necessary for climate stabilization will likely come from CCS. Post-combustion systems can be retrofitted to existing facilities, as they do not alter the clinker production process. They can only be an economic burden on the plant and result in a 25 % increase in the amount of CO2 sent to storage. Oxygen combustion, on the other hand, fundamentally alters the process with significant potential benefits. Aside from increasing production, which has the potential to offset some of the costs, the recovery of heat from the exhaust stack and improved combustion conditions could increase plant efficiency, potentially decreasing the CCS burden. Isolated calcination can be viewed as a near-term, retrofit option that can provide significant reductions but is ultimately insufficient for stabilization. Finding an effective solution, with the lowest cost, is imperative as cement, and therefore clinker, are used ubiquitously and are key components of global development goals. The conversion to oxygen combustion will necessitate the addition of an oxygen production facility in addition to the CO2 compressing stations required for CCS. These additions will double the electrical demand of the plant to over 200 kWh per tonne of clinker. Even assuming coal-based electricity without CCS, the fugitive emissions from the electricity consumption are potentially balanced by improved plant operations, increase safety and reduce the specific fuel consumption. Even today’s best available technology operates at a thermal efficiency of approximately 70 %. By returning the exhaust heat to the kiln, reducing the amount of vent air and utilizing the heat radiated from the outside of the kiln shell, the potential exists to significantly increase the thermal efficiency of the process. The avoidance of nitrogen in the kiln also reduces the emphasis on NOx reduction. The solution to this problem has generally been to introduce a low-NOx (temperature) burner, which adversely affects clinker quality. The use of oxygen will allow for a short, hot burning zone that will be able to produce consistently high-quality clinker. In addition, the conversion process opens exciting new doors for


488


fuel gasification and heat recovery that in turn opens the door to additional fuels and process steps that may further improve energy efficiency. The implementation of CCS in the cement industry is an expensive proposition that will require much study. Such large-scale changes to the industry will likely be associated with a ‘next generation’ cement plant. The cement industry is often considered to be the most expensive place to implement CCS and will likely benefit from the current focus on the power sector. However, unlike the power sector, which has several options for carbon free electricity (solar, wind, nuclear), the cement industry is without such options. Blending and alternate fuels will continue to provide near-term reduction with efficiency improvements also a cost-effective option. In the long term, alternate fuels in combination with CCS could provide a feasible negative emissions technology, i.e. a method to actively reduce atmospheric CO2 levels.

15.7


Risks associated with climate change are spurring much work across the broad spectrum of human industry. The cement industry is no different. There are many published papers in peer reviewed journals and the cement industry was mentioned in the IPCC Special Report on Carbon Capture and Storage. In addition, three comprehensive studies that extensively cover the application of CCS technologies to the cement industry were released in 2008 by the European Cement Research Academy (ECRA), the International Energy Agency (IEA) and Columbia University, sponsored by the Cement Sustainability Initiative (CSI). The report by the European Cement Research Academy provides a complete overview of the various CCS technologies and their feasibility for the cement industry (ECRA 2007). The International Energy Agency report provides similar information with an additional energy, material and cost analysis for post-combustion capture and isolated calcination (IEA, 2008). The major conclusions of the report are summarized in a conference paper (Barker et al., 2009). A complete review of the changes and opportunities associated with oxygen combustion is discussed in detail in a report for the Cement Sustainability Initiative of the World Business Council on Sustainable Development (Zeman and Lackner, 2008). The Cement Sustainability Initiative, in conjunction with ECRA, published a series of technology papers discussing all of the options available to the cement industry (www.wbcsdcement.org). A good overview of accelerated curing of accelerated CO2 curing of concrete at the laboratory scale and its potential for CO2 sequestration is provided by Shao et al. (2006).



15.8

489

References

Barker D, Turner S, Napier-Moore P, Clark M and Davison J (2009) ‘CO2 capture in the cement industry’, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 87–94. Burkan Isgor O and Razaqpur A G (2004) ‘Finite element modeling of coupled heat transfer, moisture transport and carbonation processes in concrete structures’, Cement and Concrete Composites, 26(1), 57–73. Cembureau (1998) Climate Change, Cement and the EU, Brussels, Belgium, Cembureau. CSI (2002), Our Agenda for Action, Geneva, Switzerland, World Business Council for Sustainable Development. CSI/ECRA (2009) Development of state of the art-technology in cement manufacturing: trying to look ahead, CSI/ECRA – Technology Papers, Dusseldorf, Geneva, CSI/ ECRA. Ebner A D and Ritter J A (2009) ‘State-of-the-art adsorption and membrane separation processes for carbon dioxide production from carbon dioxide emitting industries’, Separation Science and Technology, 44, 1273–1421. ECRA (2007), Carbon Capture Technology – Options and Potentials for the Cement Industry, Dusseldorf, Germany, European Cement Research Academy GmbH. EIA (2007) Emissions of greenhouse gases in the United States 2006, Washington, DC, Energy Information Administration, US Department of Energy. Fernandez Bertos M, Simons S J R, Hills C D and Carey P J (2004) ‘A review of accelerated carbonation technology in the treatment of cement-based materials and sequestration of CO2’, Journal of Hazardous Materials, 112(3), 193–205. Garcia-Labiano F, Abad A, Diego L F D, Gayan P and Adanez J (2002) ‘Calcination of calcium based sorbents at pressure in a broad range of CO2 concentrations’, Chemical Engineering Science, 57, 2381–2393. Gómez A, Fueyo N and Tomás A (2007) ‘Detailed modelling of a flue-gas desulfurisation plant’, Computers & Chemical Engineering, 31(11), 1419–1431. Goodbrake C J, Young J F and Berger R L (1979) ‘Reaction of hydraulic calcium silicates with carbon dioxide and water’, Journal of the American Ceramic Society, 62(9–10), 488–491. Gronkvist S, Bryngelsson M and Westermark M (2006) ‘Oxygen efficiency with regard to carbon capture’, Energy Procedia, 31, 3220–3226. Hegerland G, Pande J, Haugen H A, Eldrup N, Tokheim L-A and Hatlevik L-M (2007) ‘Capture of CO2 from a cement plant – technical possibilities and economical estimates’, in Gale J, Rokke N, Zweigel P and Svenson H (eds), Proceedings of the Eighth International Conference on Greenhouse Gas Control Technologies: GHGT8, Oxford, UK, Elsevier, CD-ROM. Holloway S (2005) ‘Underground sequestration of carbon dioxide – a viable greenhouse gas mitigation option’, Energy, 30(11–12), 2318–2333. Huntzinger D N, Gierke J S, Kawatra S K, Eisele T C and Sutter L L (2009) ‘Carbon dioxide sequestration in cement kiln dust through mineral carbonation’, Environmental Science & Technology, 43(6), 1986–1992. IEA (2008), CO2 Capture in the Cement Industry, Cheltenham, UK, IEA Greenhouse Gas R&D Programme. Johnston J (1910), ‘The thermal dissociation of calcium carbonate’, Journal of the American Chemical Society, 32(8), 938–946. © Woodhead Publishing Limited, 2010

490


Kashef-Haghighi S and Ghoshal S (2010), ‘CO2 sequestration in concrete through accelerated carbonation curing in a flow-through reactor’, Industrial & Engineering Chemistry Research, 49, 1143–1149. Klemm W A and Berger R L (1972) ‘Accelerated curing of cementitious systems by carbon dioxide : Part I. Portland cement’, Cement and Concrete Research, 2(5), 567–576. Lea F M (1970) The Chemistry of Cement and Concrete. London, UK, Edward Arnold. Lide D R (2000) CRC Handbook of Chemistry and Physics, Boca Raton, FL, CRC Press. Logan C O (2006) Carbon dioxide absorption and durability of carbonation cured cement and concrete compacts, Master of Engineering, Montreal, QC, Canada. McGill University. Lusche M, Endres G and Euskirchen J (1982) ‘Recent experience in the utilization of radiant waste heat from industrial kilns’, ZKG International, 35(9), 447–450. Monkman S (2008) Maximizing carbon uptake and performance gain in slag-containing concretes through early carbonation, Doctor of Philosophy, Montreal, QC, Canada, McGill University. Monkman S and Shao Y (2006) ‘Assessing the carbonation behavior of cementitious materials’, Journal of Materials in Civil Engineering, 18(6), 768–776. Papadakis V G, Vayenas C G and Fardis M N (1989) ‘A reaction engineering approach to the problem of concrete carbonation’, AIChE Journal, 35(10), 1639–1650. Pfaff I and Kather A (2009) ‘Comparative thermodynamic analysis and integration issues of CCS steam power plants based on oxy-combustion with cryogenic or membrane based air separation’, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 495–502. Shao Y, Mirza M S and X W (2006a) ‘CO2 sequestration using calcium-silicate concrete’, Canadian Journal of Civil Engineering, 33(6), 776–784. Shao Y, Zhou X and Monkman S (2006b) ‘A new CO2 sequestration process via concrete products production’, IEEE EIC Climate Change Conference, Ottawa, ON, Canada, 1, 216–221. Shi C and Wu Y (2008) ‘Studies on some factors affecting CO2 curing of lightweight concrete products’, Resources, Conservation and Recycling, 52(8–9), 1087–1092. Steinour H H (1959) ‘Some effects of carbon dioxide on mortars and concrete-discussion’, Journal of the American Concrete Institute, 30, 905–907. Stolaroff J, Lowry G and Keith D (2005) ‘Using CaO- and MgO-rich industrial waste streams for carbon sequestration’, Energy Conversion and Management, 46, 687–699. van Oss H G (2004), Cement, Restom, VA, U.S. Geological Survey Minerals Yearbook. van Oss H G and Padovani A C (2003) ‘Cement manufacture and the environment, Part II: environmental challenges and opportunities,’ Journal of Industrial Ecology, 7(1), 93–123. Wang S (2007), Carbonation of cement-based products with pure carbon dioxide and flue gas Master of engineering, Montreal, QC, Canada, McGill University. Worrell E, Martin N and Price L (2000), ‘Potentials for energy efficiency improvements in the US cement industry’, Energy, 25, 1189–1214. Worrell E, Price L, Martin N, Hendriks C and Meida L O (2001) ‘Carbon dioxide emissions from the global cement industry’, Annual Review of Energy and the Environment, 26(1), 303–329.



491

Wrampe P and Rolseth H C (1976) ‘The effect of oxygen upon the rotary kiln’s production and fuel efficiency: theory and practice’, IEEE Trans. Ind. Apps, IA–12(6), 568–573. Young J F, Berger R L and Breese J (1974) ‘Accelerated curing of compacted calcium silicate mortars on exposure to CO2’, Journal of the American Ceramic Society, 57(9), 394–397. Zeman F S (2008) ‘Clinker formation in atmospheres dominated by nitrogen and carbon dioxide’, ZKG International, 61(5), 58–67. Zeman F S and Lackner K S (2008), The Reduced Emission Oxygen Kiln, New York, NY, Lenfest Center for Sustainable Energy at Columbia University. Zhou X D (2005), CO2 uptake through carbonation of concrete using simulated and as-captured flue gas, Master of Engineering, Montreal, QC, Canada, McGill University.


16

Carbon dioxide (CO2) capture and storage technology in the iron and steel industry

J - P. B i r a t, ArcelorMittal Global R&D, France Abstract: This chapter presents the activity of the iron and steel industry in the field of carbon dioxide (CO2) capture and storage (CCS) and, more generally, of CO2 mitigation. Emissions from the sector amount to around 5 % of anthopogenic emissions, but over the past 40 years, recycling and strong energy conservation policies have achieved a 55 % cut. However, unless or until the world is able to move to a closed loop economy, at the end of the 21st century at the earliest, steel will have to keep being produced from ore. Since CO2 emissions cannot be cut any further by energy conservation measures, which have almost reached their thermodynamic limits, it is necessary to introduce breakthrough technologies. In the short term, this will mean introducing CCS in the core of the blast furnace process operation (called the TGR–BF), with the potential of retrofitting existing steel mills with this technology from the 2020s on. In regions where natural gas is available and cheap enough, a similar concept can be applied (ULCORED process). Completely new technologies, such as the Smelting Reduction Hisarna Process (with CCS), are also being explored. In addition, and looking much further ahead (deployment beyond 2030?), other technologies not relying on CCS but using either hydrogen (hydrogen prereduction) or electricity (electrolysis) or biomass, can be conceived. Key words: steel, steelmaking, ironmaking, CCS, PSA, cryogenics, blast furnace, ULCOS, breakthrough technology.

16.1

Introduction

Steel is among the major structural materials in the world, its production coming second only to that of cement. Iron and steel have been used for several thousand years, to make artifacts, from buildings to automobiles and from guns to cans, but also the tools and machines from which all other artifacts are made. Steel is ubiquitous, and the history of mankind is completely interwoven with that of the material. Behind the name of steel hide several thousand different alloys, the largest ever family of materials. The steel industry is sophisticated, modern and capital-intensive with production amounting in 2008 to 1.3297 Gt. It features some of the most impressive engineering reactors, such as the blast furnace (BF), which is unique and probably as powerful and complex as a nuclear reactor or the large rockets used into raise heavy payloads into orbit. Anthropogenic emissions of greenhouse gases (GHG), which amounted to 49 Gt of CO2 equivalent worldwide in 2004,1 are traditionally split among 492 © Woodhead Publishing Limited, 2010

CO2 capture and storage technology in iron and steel industry

493

economic sectors, with industry accounting for 19.4 %. The steel industry represents 6–7 % of global anthropogenic CO2 emissions according to the IPCC,1 but only 4–5 % according to the IEA,2 i.e. 1/4 to 1/3 of the whole industry sector. These estimates include both direct emissions by the steel mills themselves and indirect emissions, generated by the energy sector to produce the electricity that the mills consume. This accounting method does not, however, encompass a life-cycle assessment whereby the emissions from steel mills are balanced against the benefits of using steel, by comparing a society that uses steel with a hypothetical one that does not. The CO2 intensity of the steel sector today is 1.9 tCO2/tcrude steel, but in precentage terms the steel industry is a small emitter compared to the energy (25.9 %), transport (13.1 %), forestry (17.4 %) or agriculture (13.5 %) sectors. However, the CO2 stream is generated by a relatively small number of large emitters, each one emitting between 1 and 10 Mt per year.

16.2

Carbon dioxide (CO2) emissions of the steel sector

Why does the steel industry generate CO2? There are two main reasons: on the one hand, energy is needed to produce steel, more often than not generated from fossil fuels and, on the other hand, reducing agents are necessary to produce steel from iron ores, the cheapest, most easily available reductant being the carbon of coal. Steel is typically made in an integrated steel mill (ISM) (see Fig. 16.1), located close to a large river or a deep sea harbour, which typically receives coal and iron ore (oxide, either hematite or magnetite) by ship, usually from

16.1 ArcelorMittal Florange’s blast furnace skyline, in France.


494


far afield, Brazil, Australia or South Africa, for example. The raw materials are blended and prepared to meet the specification for steel production in large plants, where the ore is sintered (sinter plant) or pelletized and the coal pyrolyzed into metallurgical coke (coke oven batteries) or ground into pulverized coal (PC grinders). Coke, PC and sinter, plus some bulk ore and sometimes pellets, are fed to a very large, circular and vertical reactor called a blast furnace (BF). A counter flow of hot air is blown through the furnace via a series of tuyeres located at the bottom third – between the bosh and the hearth. A blast furnace has a diameter of 6–15 m, a height of 35 m, a working volume of 800–4800 m3, a production of pig iron (hot metal) of 1000–15 000 t/day and a power of 1 GW – thus it stands amongst the largest engineering reactors in all industrial sectors. The blast furnace taps hot metal, a saturated iron-carbon melt, at 1500 °C, liquid slag (later commonly used as a raw material for the cement industry) and gases, since the high temperature chemistry in the reactor (Boudouard equilibrium) produces a gas that contains similar amounts of CO and CO2. This gas is carefully collected, mixed with coke oven and steelmaking gases, and sent through an internal grid to reheating furnaces and a power plant to combust it fully into CO2. The blast furnace may operate continuously (with rare shutdowns) for more than 25 years. To make steel, it is necessary to oxidize the carbon (as well as other elements like silicon and manganese) out of the melt. This is done in a further steelmaking plant, where pure oxygen is blown into the hot metal through a top lance or bottom tuyeres. The reactor is called a converter or BOF (basic oxygen furnace). It is a batch process, tapping regularly, twice every hour, up to 400 t of liquid steel at around 1600 °C at each cycle. There are many more plants in the steel mill to customize the composition to produce hundreds of different steel chemistries (secondary metallurgy); steel is soilidified in a continuous caster and then hot and sometimes coldrolled down to the desired shape and dimensions. Metallurgical properties are achieved by various heat treatments carried out on-line or off-line, continuously or in batch. The most energy-intensive part of the mill is the blast furnace area, and CO2 is generated there and also at every downstream plant where combustion takes place: hence the complexity of the CO2 sources shown in Fig. 16.2. Figure 16.3 shows the various production routes used to make steel today and their share worldwide and in France. The BF or ISM route produces steel from primary raw materials, i.e. iron ore, and requires both energy and reducing agents in the form of coke and pulverized coal. The electric arc furnace (EAF) route is different in that it produces steel from secondary raw materials, i.e. iron scrap. The physical requirement of the mill, therefore, is mainly energy, in the form of electricity, coal and oxygen, to melt the scrap, without the requirement for oxide reduction. As soon as liquid steel



288 kg 5–10 % CO2

CO2

Coal 12 kg 57 kg limestone 133 kg 30 % CO2

CO2

Sinter strand Pellet plant

BF gas Blast furnace

329 kg CO2 25 % CO2

285 kg

CO2 25 % CO2

Coke oven gas

Coke plant Limestone 109 kg

709 kg

1255 kg eq CO2 in BF gas

Hot blast Stoves Coal 187 kg

Coke Lime kiln

Total CO2 emission: 1815 kg/t rolled coil

Coal = 1710 kg CO2 Limestone = 105 kg CO2

72 kWh 138 kg scrap

Coal 382 kg

495

CO2 20 % CO2

Power plant Hot strip mill

CO2 84 kg 10 % CO2

Converter Flares etc. gas 63 kg Steel plant

16.2 simplified flow sheet of an integrated steel mill, showing carbon-bearing material input, CO2 emissions, expressed in volume (kg/t of hot-rolled coil) and concentration in the flue gas (volume %).

is available the process can move on into charted territory with the same casting and rolling techniques as before. The direct reduction (DR) route is based on ore and uses natural gas as a reducing agent and fuel, along with electricity for the EAF. Thus it requires a different kind of shaft reactor called a pre-reduction furnace. Production is less, between 1 and 2 Mt annually. Steelmaking and downstream plants are same as in the ISM route. The carbon dioxide intensity of the three routes is, respectively, 1.97, 0.45 and 1.10 tCO2/tcrude steel.3 An ISM is a complex series of interconnected plants, where CO2 is emitted from many stacks (10 or more). Figure 16.2 gives a simplified carbon balance, showing the major entry sources (coal and limestone) and the stack emissions, in volume (kg/t of hot rolled coil) and concentration of CO2 (volume %). The major CO2 stream comes out of the blast furnace and accounts for 69 % of all steel mill emissions to the atmosphere. This is indeed where most of the reduction takes place and where most of the energy is needed. This is also the preferred place for dealing with CO 2, in a Pareto-type approach.* The top gas of the blast furnace is roughly 25 % *It is possible in theory to apply CO2 capture and sequestration (CCS) to every smokestack of the ISM, which would cut CO2 emissions by 100 %. The BF however is the largest source of CO2 and the CO2 stream it generates is quite concentrated (see below). Most of the effort to develop CCS for the ISM therefore concentrates on applying CCS to the BF which, moreover, requires an industry-specific approach. Further reductions would mean applying standard technologies to other smokestacks, similar to those used in power plants for example, starting at the coke ovens and sinter plant. © Woodhead Publishing Limited, 2010

Lump ore Sinter

Coal


Coke

Scrap

Pellets DR

Natural gas, oil or coal

Steel making

Hot metal Oxygen

Coal Natural gas

Natural gas, oil

Blast O2

Air

Fine ore

Pellets

BF Iron making

Lump ore

Secondary steel production

Shaft Rotary kiln Fluidized furnace furnace bed DRI Scrap

Scrap

Scrap

OHF

EAF

BOF

3 %

66 %

0 %

61 %

Crude steel

EAF

6 %

25 %

World, 2007

0 %

39 %

France, 2007

16.3 Production routes to make steel today, with production shares in the world and in France, an example chosen to show that the world data average scattered country ones (BF = blast furnace; OHF = open hearth furnace; BOF = basic oxygen furnace; EAF = electric arc furnace; DR = direct reduction).


Raw material preparation

Fine ore

496

Primary steel production


497

of CO2, the rest being CO at a similar concentration with a complement of nitrogen. All the other stacks put together account for 31 % of the emissions: they exhibit rather low CO2 concentration, typical of the flue gas in a combustion chamber. Of course, the BF top gas never ends up directly in a stack, as the embedded energy is recovered in a power plant, which is part of the mill complex. Compared to an ISM, a DR steel mill generates lesser quantities of CO 2 at the stack, but, in common with the ISM mill, it generates emissions downstream at the steel shop and rolling mills. An EAF mill generates even smaller amount of CO2, from the steelshop on: most of its emissions are actually due to the electricity production needed to power the EAF.

16.3

Strategies to control carbon dioxide (CO2) emissions from the steel sector

A state-of-the-art steel mill is very much an optimal system in terms of consumption of fuels and reducing agents. The blast furnace itself operates close to its thermodynamic limits, and the whole mill has a potential of energy savings of roughly 10 % only. This is due to several decades of cost management, as high energy prices have driven the industry to optimize its processes as close as possible to their physical limits. The industry rightfully claims energy savings and, correspondingly, CO2 cuts ranging between 50 and 60 % over the last 40 years, depending on the local conditions: this is the highest level of energy conservation achieved by any industrial sector. Therefore, cutting CO2 emissions further, to the level that post-Kyoto policies require, raises specific challenges: it is necessary to uncouple energy savings and CO2 mitigation targets, a feature unique to the steel sector. It goes without saying that the use of steel scrap should be retained at the high level it has reached today. It is estimated that the collection rate of obsolete scrap is currently around 85 %; this forms the basis of a strong recycling economy, complete with scrap dealerships and a specific steel production route based on the EAF. Put simply, value is created by the recycling of virtually all available scrap. In the long term, this situation will continue. It should also be pointed out that the indirect emissions related to electricity production will be reduced with time. For example, ULCOS (for Ultra Low CO2 Steelmaking) has shown that, under a strong carbon constraint, the carbon intensity of the European electricity grid will drop from 370 gCO2/kWh in 2006 to 144 g in 2050, a specific drop of 55 %, which will be translated into the same level of reductions in indirect emissions.4 The major source of CO2 emissions from steel mills remains the orebased route, and this will retain an important role in the long term, at least until a recycling society can replace the 20th and 21st century economy of


498


production growth which is mainly driven by population growth – probably some time in the 22nd century or at the very end of the present one. 5 Methods of curtailing emissions from the ore-based route must be found, and it is clear from the previous sections that there is no simple process, available off the shelf, that can accomplish this. Major paradigm shifts in the way steel is produced have to be imagined and the corresponding breakthrough technologies designed and developed by strong R&D programs. The largest such program designed for movement in this direction, ULCOS, has been running in the EU since 2004.5, 6 ,7, 8 The analysis that ULCOS has proposed in terms of breakthrough technologies is shown in Fig. 16.4, which explains how reducing agents and fuels have to be selected from three possibilities, carbon, hydrogen and electrons, mostly in the form of electricity. Bacteria would also sit on the electron apex, if microbiological metallurgy was considered. The mock ternary element in the figure is intended for didactic clarity: all existing energy sources can be represented on the triangle sides (e.g. coal is close to carbon on the carbon–hydrogen line, natural gas is closer to hydrogen, hydrogen from water electrolysis is on the hydrogen–electricity line, etc.). Present day steel production technology is based on coal, i.e. mostly on carbon, on natural gas, a mix of carbon and hydrogen, and on electric arc furnaces, indicated by the light grey boxes in the figure. To identify CO 2-

Existing technology

CO2 capture & storage

Carbon Coke

Use of C from sustainable biomass

Coal

Syngas

Blast furnace

Plasma in Blast Furnace

Natural gas pre-reduction Natural gas H pre-reduction 2 H2 Hydrogen

Decarbonizing

Electric arc furnace

Electrolysis H2 by electrolysis of H2O Electricity Electrons

New technology

16.4 Pathways to breakthrough technologies for cutting CO2 emissions from the ore-based steel production routes.



499

lean process routes, only three major solution paths stand out: (i) a shift away from coal, called decarbonizing, whereby carbon would be replaced by hydrogen or electricity, in processes such as hydrogen reduction or electrolysis of iron ore (indicated by the dark grey boxes) (ii) the introduction of CCS technology; (iii) the use of sustainable biomass. These are shown in the light grey boxes in the diagram. ULCOS has investigated about 80 different variants of these concept routes in the initial phase of its research program, using modeling and laboratory approaches to evaluate their potential, in terms of CO2 emissions, energy consumption, operating cost of making steel and sustainability.9 Among these, six families of process routes have been selected within the ULCOS program for further investigation and eventual scale-up to a size where commercial implementation can take over: ∑

a blast furnace variant whereby the top gas of the blast furnace goes through CO2 capture, but the remaining reducing gas is reinjected at the base of the reactor, which is moreover operated with pure oxygen rather than hot blast (air). This has been called the top gas recycling blast furnace (TGR–BF). The CO2-rich stream is sent to storage (see Fig. 16.5). ∑ a smelting reduction process based on the combination of a hot cyclone and a bath smelter (called HIsarna) and incorporating some of the technology of the HIsmelt process.10 The process also uses pure oxygen and generates off-gas which is almost ready for storage (see Fig. 16.6). ∑ a direct reduction process, called ULCORED, which produces direct reduced iron (DRI) in a shaft furnace, either from natural gas or from coal gasification. Off-gas from the shaft is recycled into the process after CO2 has been captured, and the CO2 then leaves the DR plant in a concentrated stream and goes to storage (see Fig. 16.7). ∑ two electrolysis variants, ULCOWIN and ULCOLYSIS, which operate, respectively slightly above 100 °C in a water alkaline solution populated by small grains of ore (electrowinning process) or at steel-making temperature with a molten salt electrolyte made of a slag (pyroelectrolysis). ∑ two more options are available: one uses hydrogen for direct reduction, when and if it is available without any carbon footprint; the other is based on the use of sustainable biomass, the first embodiment of which is charcoal produced from sustainable eucalyptus plantations situated in tropical countries. In the short term, the TGR–BF seems the most promising solution, as existing blast furnaces can be retrofitted with the new technology, thus limiting the extensive capital expenditures that would be necessary to switch over to the breakthrough technologies. Moreover, the very principle of the process delivers energy savings because the capture of CO2 and the


500


16.5 Schematics of the TGR–BF process; the furnace is in the center and is shown in a separate window, where the CO2 separation unit and gas reheater are shown feeding two rows of tuyeres; the cones show iron ore and coal; CO2 is sent to underground storage through a pipeline (from www.ulcos.org).

16.6 Schematics of the HIsarna process; the reactor, in the center, is also shown in a more detailed window, where the cyclone sits on top of the bath smelter and char from a screw reactor feeds carbon into the furnace; the cones show iron ore and coal; CO2 is sent to underground storage through a pipeline (from www.ulcos.org).



501

16.7 Schematics of the ULCORED process: the reactor is in the center; to the left is the pelletizing plant for iron ore; the grey box is the CO2 separation unit and the gas flows through a pipeline to the underground storage site (from www.ulcos.org).

recycling of the purified gas displace high-temperature chemical equilibria (Boudouard reaction) and use coke and coal with a higher efficiency inside the BF than is possible with conventional operation. This balances the extra costs incurred by the capture and storage, to some extent. The concept has, in addition, been tested on a large-scale laboratory blast furnace in Luleå, with positive outcome.11 Where natural gas is available, ULCORED is an attractive option. A 1 t/h pilot is planned to be erected in Luleå in the next few years by LKAB, an ULCOS partner, with a view to fully validating the concept. Somewhat later and probably only for greenfield steel mills, the HIsarna process will also be an option. An 8 t/h pilot is to be erected and tested in the course of the ULCOS program. The electrolysis processes have been developed from scratch within the ULCOS program and therefore are still operating at laboratory scale. Although they hold the promise of zero emissions, if they have access to green electricity, time is required to scale them up to a commercial size (10–20 years). Hydrogen steelmaking will depend heavily on the availability of ‘green hydrogen’, while the use of charcoal, from growing countries, would require the set up of complex logistics, including heavy infrastructure across several continents. Discussions have been centered until now on the major sources of CO2, allowing cutting of emissions for the whole steel mill by more than 50 %. It


502


is possible to cut emissions further by treating the other stacks of the steel mill: the cost of abatement would of course, be higher. With this rationale, though, zero emissions could potentially be achieved. There are also other programs addressing this challenge: along with ULCOS, they are part of the CO2 Breakthrough Program of worldsteel – the international Iron & Steel Association, a Forum for the various initiatives to exchange information about their progress.12 ∑

∑ ∑

∑

∑

∑

Japan has a large national program led by the Japanese Iron and Steel Federation (JISF) called COURSE 50, which focuses on the development of a new amine scrubbing technologies for blast furnace gas and the use of hydrogen separated from coke oven gas, a by-product of the steel industry.13 The ready-to-use technology concepts should be available by 2030. POSCO, in Korea, runs its own program, with various dimensions including the adaptation of CCS to the COREX® process and the development of an ammonia-based scrubbing process.14 The American Iron and Steel Institute (AISI), in North America, runs a program where high-temperature electrolysis is examined at MIT, hydrogen reduction of iron ore in the laboratory, preparatory to transposing to a flash furnace reactor, at Utah University, mineral sequestration at Columbia University and CO2 collection from EAF fumes using lime at Missouri Rolla University.15 A Canadian program, run by the Canadian Steel Producers Association (CSPA), has a strong focus on the use of biomass in iron and steelmaking as a substitute for fossil fuels, as biomass per capita is quite important in this large country.16 ArcelorMittal Brazil has been reporting its development of a biomass steel production route based on sustainable plantations of eucalyptus trees, production of charcoal and small charcoal blast furnaces,17 which is already used in Brazil but on a small scale (300 000 tpy BF) and is most probably a local solution. There are also participating programs from Bao Steel in China, China Steel in Taiwan and SAIL in India.3

The rationale for all of these programs is similar to that of ULCOS. However, they are less advanced in terms of making breakthrough technologies available and their progress is not yet widely reported. These are the main reasons why this chapter is heavily based on the ULCOS approach, which is very typical of what is done elsewhere. The long lead time needed for the development of breakthrough technologies shows that there is no simple recipe for cutting the present CO2 emissions of the steel industry by 50 % or more (the objective of the ULCOS program): new technologies have to be developed, which means a high level of risk,



503

necessarily lengthy development time, large budgets for R&D and then large capital expenditures to convert steel mills to the breakthrough processes. Moreover, the economic viability of these solutions will depend on the worldwide price of CO2 and on the implementation of a level playing field for climate policies in order to avoid ‘carbon-havens’ and consequent carbon leakage, especially out of Europe. With all these caveats, the steel industry can cut its emissions significantly and continue to provide a material that the world needs in order to maintain the standard of living of its citizens and cut CO2 emissions in other sectors.

16.4

Carbon capture and storage (CCS) for the steel sector

This section will refocus on CCS for the steel sector, because it is the main topic of this book but also because implementing CCS seems to be the quickest route – expected to be operational by the 2020s – to delivering significant cuts in the CO2 emissions of the sector. It is important to note that CCS will be implemented in the steel industry without matching any of the existing CCS categories, which have been defined with energy generation in mind: indeed, in the steel sector, the major part of the generation of CO2 is related to the reduction of the iron oxides that constitute iron ore. Oxyfuel combustion, pre- or post-combustion capture and chemical-looping do not mean much in an industrial context where there is neither combustion nor oxidation – except very locally inside the reactors. Figure 16.8 presents the various CCS concepts applied to the steel industry and to a combustion process. The appropriate concept to apply to the TGR–BF is that of in-process CO2 capture, with oxygen operation. The oxygen part is similar, but not identical to oxyfuel operation. The recycling part is original and is the key reason why some energy savings, and the corresponding cut in operating costs, are gained. The same concept applies to the ULCORED process, which also includes use of pure oxygen and in-process recycling of the shaft top gas, in addition to other features such as a series of shift reactors in the recycling loop. The HIsarna process is slightly different from the two other processes as it does not involve a recycling loop for the gas: the smelter gas is oxidized at the cyclone level, where some reduction is carried out along with combustion to pre-heat and melt the ore. There is a counter-current flow of the gas against the iron stream, in which its chemical energy is fully exhausted. Figure 16.9 shows the carbon and CO2 mass balances of a steel mill operating with TGR–BF. Emissions are cut by 65 % compared to the nonULCOS benchmark steel mill of Fig. 16.2 (and by 56 % in the steel mill


504

Developments and innovation in CCS technology Post-combustion capture

Pre-combustion capture

Steelmaking process

Gases

CCS CCS

CCS

Material units

CCS in industrial processes, including combustion

Process reactor

Air or O2

Combustion chamber

16.8 Implementation of CCS in process industries including combustion.

itself, due to the carbon saving introduced by the process). Capturing the flue gas of an extra stack, like that of the sinter plant, would bring the reduction level upto 75 %. The most striking feature of the top gas stream from which CO2 is recovered is the high concentration of CO 2, around 35 %, which is significantly more than in the top gas of the conventional blast furnace.

16.5

Carbon dioxide (CO2) capture technologies for the steel sector

Capture technologies call mainly on the following phenomena (as further detailed in related chapters): ∑

sorption separation either by adsorption or absorption under different physical and chemical processes; or ∑ cryogenic separation by fractional liquefaction; or ∑ physical separation achieved by pushing the gas through membranes or molecular sieves. Physical and chemical adsorption (respectively physisorption and chemisorption) processes depend on the nature of the CO2 trapping material


CO2 capture and storage technology in iron and steel industry Coal = 1255 kg CO2 Limestone = 105 kg CO2 Natural gas = 95 kg CO2

586 kWh 138 kg scrap 255 kg 1–10 % CO2 CO2

57 kg Coal 24 kg 30 % CO2 Limestone 133 kg CO2 Sinter strand Pellet plant

TGR-BF gas Blast furnace

CO2

71 kg 10% CO2 CO2

Recycled gas

Coal 144 kg

Coke Lime kiln

CO2 capture

36 kg 5 % CO2

Coke oven gas

Coal 258 kg

Total CO2 production: 1455 kg/t rolled coil 812 kg CO2 to storage Net CO2 emission: 643 kg/t rolled coil

Stoves

Coke plant Limestone 109 kg

505

Natural gas 1.7 GJ

Hot strip mill

CO2 153 kg 20 % CO2

Converter Flares etc. 71 kg gas Steel plant

16.9 Simplified flow sheet of an integrated steel mill operating with a TGR–BF, showing carbon-bearing material input (darker shaded boxes), CO2 emissions, expressed in volume (kg/t of hot-rolled coil) and concentration in the flue gas (%).

(respectively, a reactive liquid vs a porous solid) and on the physics that will be used to restore and release the CO2 following capture (by lowering the pressure or increasing the temperature in the first case and heating up in the second case). The natures of the bonds between the sorbate and the surface differ under physisorption and chemisorption processes; physical processes utilize, for example, hydrogen and dipole–dipole bonds, while chemical processes utilize, for example, covalent and ionic bonds. All of these separation processes operate either in batch, by capturing CO 2 in a first stage and then releasing it in a second stage, or continuously in the filtration process used with membranes. Other solutions propose to capture CO2 and store it immediately, for example by mineral sequestration.18 Some of these technologies are mature and have been used on a large scale in specific industrial applications related to CO2 or other gases such as hydrogen: this is the case with amine scrubbing (based on AMDEA), and with physical adsorption systems such as PSA (pressure swing adsorption) or vpsa (vacuum pressure swing adsorption), complemented by a cryogenics unit designed to purify the CO2 stream. Others are under development or are still only operating at too small a scale to constitute a benchmark on which to design a system for the steel industry: examples are liquid ammonia scrubbing,19 capture in clathrates20 or by lime,21 and membrane technologies.


506


The following discussion will be limited to mature technologies, which will clearly introduce a bias in the conclusions, as CCS is a technology that will be deployed from 2020, a period when some of the emerging technologies may have matured. There are strong expectations, for example, that new chemicals, amines or other families of molecules will be developed to replace AMDEA with much lower needs for steam and energy. The same is said of membranes and, in fact, most research teams claim large development potentials, some of which are bound to come to fruition. The technology proposed by the ULCOS program to incorporate capture in the TGR–BF is physisorption. Figures 16.10 and 16.11 show the equipment which was used in the experimental blast furnace (EBF) during the largescale TGR–BF experiments carried out by ULCOS in Luleå and the physical principle of its operation.11 (The EBF is a 1.1 m BF that produces 1.5 t/h of hot metal. A production BF produces between 50 and 500 t/h, depending on its size.) It is a VPSA built by Air Liquide, a partner of the program. This technology was chosen for two reasons: it was the simplest and cheapest solution, here and now, to collect CO2 from the off-gas of the blast furnace and to produce a recycling stream of gas where the concentration in the reducing gas, mainly CO, would be maximized. In these experiments, the CO2 stream was not stored. In a larger scale experiment, like the one that is currently being organized to follow up on the EBF trials in the ULCOS program at the size of a

L = 20 m W = 16 m

16.10 View of the VPSA of Air Liquide used during the ULCOS TGR–BF experiments in Luleå.


Schematic diagram of pressure swing cycle

Classification of adsorption processes according to pressure levels HP and LP HP

CO2 lean gas


PSA (pressure swing adsorption) HP

HP

VPAS (vacuum presure swing adsorption) HP

1.5 bar 1 bar

CO2 containing gas LP

CO2-rich gas

Desorption (regeneration)

VSA (vacuum swing adsorption)

LP Time

Adsorption

LP

Repressurization

PSA/VPSA: separation performances

LP PSA

VPSA 80 %

90 %

16.11 Principle of the PSA CO2-scrubbing techniques (left) and various domains of application and performances of the variant techniques, PSA, VPSA and VSA. (Air Liquide in ULCOS).


Pressure

507

508


commercial blast furnace*22 – a necessary step to scale up the technology – the gas will be stored in a deep saline aquifer and therefore a higher level of CO2 purity is required. Further purification of the stream using cryogenics, i.e. fractional liquefaction of the gas, is hence necessary, which generates an extra stream of reducing gas which is also recycled in the BF. In this case, the optimized system consists of a combination of a PSA and a cryogenics unit. Detailed studies carried out by ULCOS show that chemisorption technologies, like amine scrubbing, physisorption ones, like VPSA or PSA, and cryogenics have different domains of optimality, the concentration of CO2 in the stream of gas to be treated being one of the most important. At the level of concentration found in the BF case, and even more clearly in the TGR–BF case, the physisorption schemes are the best, in terms of technical performance and cost, both operating and capital. This is also true if CCS is applied to a DR route, like ULCORED.23 In the HIsarna case, which directly delivers a very high concentration of CO2, a cryogenics unit is enough. On the other hand, if CCS needs to be applied to the other stacks of the steel mill, then an amine scrubbing unit would be the best solution. These conclusions might be seen as Eurocentric, i.e. dependent on the local conditions in terms of price of energy and steam. Indeed, it should be noted that the Japanese COURSE 50 program, which is also developing a capture technology for the BF top gas, has selected a chemisorption solution.24 The different conclusions could also be due to different time horizons considered for estimating the potential of the capture technologies. Indeed, the amine washing considered in the ULCOS program is based on the present state of the art of this fairly common technology, i.e. on the use of commercial MDEA amines that exhibit an energy need for restoring the sorbant of 3.2 GJ/tCO2. R&D is underway to improve this performance (down to 1.8 GJ/t), to work at lower temperatures and to use wasted heat: the Japanese COURSE 50 national program, for example, has promised to deliver this new technology by the 2020s. An important point is the concentration in CO2 reached by the various technologies. The EU directive on CO 2 storage does not specify this concentration because it is a complex question, related to what technologies can deliver under actual industrial circumstances and the intention of the legislator that other waste gases should not be collected and stored along with CO2. The degree of CO2 purity clearly impacts on cost; it should be determined in terms of storage, i.e. the ability to turn CO2 into a condensed *The pilot planned for Eisenhüttenstadt is 0.6 Mt/year and the Florange Demonstrator 1.4 Mt/year.



509

fluid at an acceptable cost, giving a threshold above 90 % but not necessarily 99 % or more. The features of the various capture technologies available today for the steel industry are shown in Table 16.1. The ranges still need to be validated by experiments over a sufficient period of time but are relatively broad, from 96–100 %. This indicates that future legislation should lay down broad targets in order to enable flexibility in choosing the best technology and should not specify an exact target or particular technology.

16.6

Carbon dioxide (CO2) storage for the steel sector

Storage of CO2 can take place in geological reservoirs (geostorage), in the ocean or by the mineralization of some other compounds, chemical reactants or rocks (ex situ storage).

16.6.1 Ocean storage Ocean storage is currently going through a moratorium and so is not being seriously considered as an option by industry at the present time.

16.6.2 Mineral sequestration Mineral sequestration is an option which has been examined seriously in both the ULCOS21 and the worldsteel programs.25, 26 The concept is simple: some minerals, such as magnesium-rich ultramafic rocks (peridotites, serpentines, gabbros, etc.), can react spontaneously (negative enthalpy of reaction) with CO2 and form carbonates, which stand below CO2 on the oxido-reduction scale:26 the compounds are usually stable, and the only difficulty of these schemes is to master the kinetics of reactions, which naturally take place in the realm of geology, with corresponding timescales. Some of the reactions may involve lime or magnesia and bicarbonates may also be formed. A scheme specific to the steel industry27 proposes to use slag, especially steel-making BOF slag, as the reactant that will be used to absorb CO2 by a chemical reaction: there is a phase in that slag, called larnite (Ca2SiO4) and present at the level of 30–40 %, which can react with CO2:

½ Ca2SiO4 + CO2 Æ CaCO3 + ½ SiO2

with an enthalpy of – 22 kcal. In addition to larnite, slag may contain as much as 6 % free lime (CaO), which also reacts with CO2 to form the same calcium carbonate. The use of slag has been studied in the ULCOS program,28 where it was shown that the reaction can proceed at moderate temperatures (90 °C), high pressures (100 bar) and moderate reaction times (90 min) if


CO CO2 N 2 H 2

%vol %vol %vol %vol

(dry) 45 (dry) 37 (dry) 10 (dry) 8


PSA VPSA Recycled gas (process gas) CO yield % CO %vol CO2 %vol N 2 %vol H 2 %vol %vol H2O CO2-rich gas captured CO %vol (dry) CO2 %vol (dry) N 2 %vol (dry) H 2 %vol (dry) Suitable for transport and storage? CCS process Electricity consumption kWh/t CO2 Capture process kWh/t CO2 Compression for storage (110 bar) kWh/t CO2 LP steam consumption GJ/t CO2 Total energy consumption GJ/t CO2

VPSA + Amines + compression and compression cryogenic flash

88.0 71.4 2.7 13.5 12.4 0

90.4 68.2 3.0 15.7 13.0 0

97.3 68.9 3.0 15.6 12.6 0

12.1 79.7 5.6 2.5 No

10.7 87.2 1.6 0.6 No

3.3 96.3 0.3 0.1 Yes?

100 100 – 0 0.36

105 105 – 0 0.38

292 160 132 0 1.05

99.9 67.8 2.9 15.1 12.1 2.1

PSA + cryogenic distillation + compression 100 69.5 2.7 15.4 12.4 0

0 100 0 0 Yes

0 100 0 0 Yes

170 55 115 3.2 3.81

310 195 115 0 1.12


Input gas

510

Table 16.1 comparison of the mature CO2 capture technologies for the steel industry. The top small table shows the composition of the input gas in the case of a TGR-BF


511

the slag is ground (50 mm) to liberate the calcium silicate, mixed with water to produce a slurry and kept agitated during the reaction process. Seventy per cent carbonatation is achievable under these conditions, which means that 1 t of slag can capture 250 kg of CO2. Comparing this amount of stored CO2 with the steel mill emissions and the amount of slag which is generated in parallel shows that only 1.3 % of the total CO2 generated by the steel mill (0.1 CO2 Mt compared to total emissions of 7.2 Mt/y) can be sequestered in this way. The ULCOS program conclusion was that this was not meeting the challenge and did not match in any way the 50 % mitigation target that was its goal. Now, if mineralization was to provide more sequestration, then more reactant would have to be used, roughly 100 times more. This shows the level of the logistics involved, as it would amount to 25 times the mass of steel produced. Proponents of mineralization do not suggest moving the rock to the steel mill, but rather the gas to the mine. This, however, is a proposal that needs more detailed elaboration before it can be considered as an option compared to geostorage.

16.6.3 Utilization of carbon dioxide (CO2) Schemes based on the utilization of CO2 will not be discussed here for several reasons. Direct use of CO2 as a gas or as a supercritical fluid has a very small potential for CO2 mitigation compared to the emissions of the steel sector and, moreover, these will eventually be released to the atmosphere which makes it difficult for such schemes to qualify as a storage solution. Transforming of CO2 into less oxidized species, for example fuels (like ethanol), requires energy in larger quantities than those that were released when CO2 was produced from coal or other fossil fuels: that some of the schemes suggest using renewable sources does not make them globally more attractive.

16.6.4 Geostorage Beyond the initial studies, where surveys of potential storage sites have been investigated from existing knowledge of the geology of vast regions, the search for sites that can adequately deliver the storage service that a large steel mill would expect over the long term remains in its infancy. The analogy with the concept of resources and reserves, used in the case of oil, gas or ore fields and deposits, is strong: the initial estimates, for example, of the Joule II project that gave a figure of 806 Gt in Europe are similar to resources.29 The storage capacity that a steel mill needs to identify prior to launching a site validation relates to reserves and, more precisely, proven reserves! These constitute the bottom and the tip respectively, of the pyramid


512


shown in Fig. 16.12. The ULCOS program compared a database of existing steel mill sites in Europe with one of geological structures supposed to be favorable for storage, i.e. onshore and offshore saline aquifers with or without lateral seals, low-enthalpy geothermal reservoirs, deep methane-bearing coal beds and abandoned coal and salt mines, exhausted to near-exhausted oil and gas reservoirs.30 Figure 16.13 gives as an example a part of Western Europe, where the major steel mills are shown with their underlying geological structure, where the North Sea basin, the Ardennes massif, the Ruhr graben and the Paris basin are the geologists’ preferred sites offering storage opportunities. Potential storage sites are present almost everywhere in Europe, close to the existing steel mills (100–200 km).30 Several potential storage sites have been selected: Scunthorpe (CORUS), Sidmar-Gent (ArcelorMittal), Taranto (RIVA), Ijmuiden (CORUS) and Fossur-Mer (ArcelorMittal). Figure 16.14 shows the area close to ArcelorMittal Works in Gent, where the Dinantian aquifer, the Bunter aquifer, the Cretaceous aquifer and coal fields are all candidates for storage, with potential injection points in the dashed area.30 More detailed investigations are necessary to identify a site. These usually include the analysis of existing geological data from previous drilling campaigns that were conducted when the area was searched for oil or gas and modeling of the site to assess its size, capacity (geographical extension and porosity) and injectability.30 Then dedicated campaigns of geophysical measurements are necessary to complement the existing data, which are usually insufficient. Some experimental drillings complete the process, leading to a ‘Site specific’ estimates with economic layer

Better quality injection site and source – sink match

Increasing cost of storage

Capacity (tm) = A.D. f.hst.rCO2

16.12 Resource and reserve concepts as applied to CO2 storage (BRGM in ULCOS). (After McCabe, 1998)


Corus Scunthorpe


Corus IJmuiden

Corus Port Talbot ARCELOR Gent

ARCELOR Bremen


N

Corus Redcar

MITTAL Hamburg

HKM TKS Duisburg Huckingen MITTAL Ruhrort salzgitter Flachst Salzgitter

ARCELOR Dunkerque CARSID Marcinelle ARCELOR Liege ARCELOR Florange

ROGESA Dillingen Saint-gobain Pont-a-Mousson

513

16.13 Steel mills (stars and dots) showing potential for CCS.

514


The Cretaceous aquifer

Sidmar NV

The Bunter aquifer

The Dinantian aquiser Cockerill Sambre SA Duferco Clabecq SA

16.14 Geological structures favorable for storage in Belgium, close to the Gent Mill of ArcelorMittal (Sidmar).

full assessment of the potential of the site. This stage requires an exploration permit. Actual injection can then theoretically start, if an operation permit has been secured. These steps have been started by various players in the steel industry but most are not yet in the public domain. Transportation to the sites would be carried out either by dedicated pipes or by batch transport, mainly ships (barges on rivers and sea vessels on the ocean, although some projects are planning to start operation by hauling CO2 in trucks). In the longer term, the collection of CO2 and its transportation towards a field of large storage sites will most probably be organized at a regional level. Networks will be built in various regions and form a series of convergent but separated grids, which have been compared to a sunflower field, as shown in Fig. 16.15.31 The emergence of such networks can be inferred from various initiatives, led by industrial operators having expertise in geological ventures, like the oil or the gas industry, or which have streams of CO2 to store (like the steel or the power industry), by regional associations (like the Yorkshire forward initiative,32 or the Fos or le Havre ones).33 At the end of the process, various operators will propose storage as a service to local regional communities, in consortia led by industry, regional organizations or financial operators.



515

16.15 The ‘daisy grid’ model of CCS in Europe.

16.7

Perspectives on carbon capture and storage (CCS) and carbon dioxide (CO2) abatement in the steel sector

The challenge of implementing CCS in the steel industry is related to the need to develop new breakthrough technologies and to integrate them with the new technologies emerging in other sectors. The power sector, which most of the discussion on CCS tends to focus on, cannot provide a role model for the steel industry as the features of its core technologies are quite different. Steel, in particular, is unlikely to apply end-of-pipe CCS, which would be similar to post-combustion capture in the power sector, but will instead seek to integrate capture into its own processes – so-called ‘in-process capture’. This is the avenue followed at least by the ULCOS consortium. There is considerable risk associated with this approach, both the positive risk of achieving large cuts in emissions that would then be implemented in the industry from 2020 on, and a negative risk of partial success, or failure or of unfavorable economic conditions that would prevent the implementation of a technology successfully developed at a purely technical level. Many commentators have stressed either the positive or the negative side, but they should be taken in context, not at face value. On the negative risk side, the following points could be made:31, 34, 35 ∑ The technological risk is related to the scale-up of process concepts such as those proposed in the ULCOS program. Larger scale experiments are underway but will take time (5–10 years) and large budgets (of the order of 500 M7). They have a lengthy agenda of critical issues to


516

∑ ∑

∑

∑

∑ ∑


solve, including the development of detailed technologies needed for the scale-up. The deployment of the technology, if and when successful, will require large capital investment in existing blast furnaces. An estimate of 71300 billion has been stated in20. This investment can only generate a payback beyond some level of CO2 value. This depends on the value of CO2, through an ETS system or a Pigovian tax, and on the cost of the CCS technology. CO2 values change on a daily basis and can be found in the press. The cost of the technology is covered below. International competition in a worldwide, very fluid steel market will create competition distortion between producers in some countries who have to contend with a high value of CO2 and producers in other countries who do not. This might eventually result in the steel industry going offshore to countries with free carbon, what is called carbon leakage. Some measures could be conceived to avoid this risk, the simplest one being a level playing field internationally, where carbon havens would be outlawed, but there are other solutions, such as a carbon border tax. 36 The regulatory situation remains complex, although regions and countries, including the EU, are taking steps to clarify the conditions for CO2 storage.37 Steel producers insist that access needs to be easy and available at or near technical cost: CO2 storage sites ought to be considered as a public good and access to them provided as a public service. They also stress the point, in Europe, that the ETS rights should be owned by the steel producer in order for it to collect the corresponding carbon value on the carbon market by selling them or using them elsewhere. The timeline for implementation is very tight compared to that of the reduction targets set by the EU. The general public is not very aware of CCS, and NGOs do not all support it. This means that stakeholders should be engaged early in the development of a CCS project in order to make it possible.

The positive side has been at the core of this chapter. If the previous conditions are met, and steel operators insist very strongly on that, then the deployment of CCS technologies in the steel industry could start taking place in the early 2020s. Full implementation in various parts of the world will depend on their regional policies regarding GHG emissions, but it will take time due to the large capital expenditures that will be incurred. A natural way of retrofitting BF to the TGR–BF technology will be to use the occasion of a BF relining to switch over to the new process route. Since a typical BF campaign lasts 15–20 years, it will take at least that long to carry out the transition. How much reduction can be implemented at the level of the sector is a complicated question, which can only be answered by discussing various



517

scenarios, since technical, economic and political boundary conditions dictate the answer. These scenarios predict almost anything, from the impossibility of decreasing emissions if all of the risks fall on the negative side, to the possibility of decreasing specific emissions by 50 % or more and even of decreasing sectoral emissions by 50 %, in spite of a large increase in production expected by 2050,3, 38 if conditions are especially favorable for this. Going beyond describing these scenarios briefly entails moving into the realm of politics. A key issue is the cost of CCS and the value of CO2 in the long term. The latter is usually described in terms of 20–50 7/t of CO2 for the near and middle term, but economists38 predict much higher values for the long term (400–600 7/t in 2050) under strong carbon constraints (50 % reduction in emissions at world level). The cost of CCS is also a complex issue, mainly because there is little experience in the world of actual operation. In spite of this, many publications come up with estimates of CCS costs,2, 34, 39 which are usually on the lower side. This is rather puzzling as costs are very difficult to come by, mainly because they are only estimates on putative technologies and are therefore subject to very large uncertainties. Moreover, costs and prices are different things and there is no guarantee that prices will be close to actual costs, plus a reasonable margin to reward the various services included in CCS. The results published in the literature should, therefore, only be considered as presenting one scenario among many other possible ones. The proposition that industry can adapt easily to any kind of target if sufficient pressure is applied is an optimistic one and one that underestimates the scale of the CCS challenge and therefore sends the wrong signals to the politicians and legislators. This does not represent the true situation and therefore does not necessarily help in making the transition to CCS easier or even possible!

16.8

Conclusions

The steel industry has been aware of the climate change threat since the late 1980s and began proposing solutions early on,40 CCS was identified from the start as a powerful potential solution to the problem. Cooperative programs have been launched in Europe and in the rest of the world to tackle the issue at various scales, and commercial-scale demonstrator experiments are now underway, which may lead to implementation and deployment from the 2020s onwards. This is a long-term agenda, full of promise but also of risks and traps, a situation which is probably similar to what other sectors are experiencing. Risks are related to the complexity of the issue, which calls for the development and the implementation of breakthrough technologies under time constraints which are very short. The message of this chapter is not that CCS is unlikely


518


to happen in the steel industry, quite the opposite. However, the optimism which prevails in many policy-driven publications is overrated, and some researchers are actually becoming aware of this situation.41

16.9 Acknowledgements The author wishes to thank the colleagues from the ULCOS consortium, who helped him frame this vision of CCS in the case of the Steel Industry, especially Tore Torp, of Statoil and Chris Treadgold of CORUS, who chaired ULCOS groups dedicated to CCS, Anne Berthelemod, Elisabeth Marlière and Michel Devaux of Air Liquide, Yves Le Nindre and Didier Bonijoly of BRGM, Nicolas Baglin and Jean Borlée of ArcelorMittal.

16.10 Sources of further information and advice Worldsteel, Bruxelles ESTEP, Brussels ULCOS web site IEA, Paris Eurofer, Brussels AISI, Washington

http://www.worldsteel.org/ http://cordis.europa.eu/estep/ http://www.ulcos.org http://www.iea.org/index.asp http://www.eurofer.org/ http://www.steel.org//AM/Template. cfm?Section=Home JISF, Tokyo http://www.jisf.or.jp/en/index.html CSPA, Ottawa http://www.canadiansteel.ca/ ArcelorMittal, Luxembourg http://www.arcelormittal.com/ Corus, London http://www.corusgroup.com/en/ ThyssenKrupp Steel, Düsseldorf http://www.thyssenkrupp-steel.com/en/ Voestalpine, Linz http://www.voestalpine.com/ag/de.html Riva, Milan http://www.rivagroup.com Dillinger Hütte, Dillingen http://www.dillinger.de/ Saarstahl, Völklingen http://www.saarstahl.com/ Air Liquide http://www.france.airliquide.com/ BRGM http://www.brgm.fr/ LKAB, Luleå http://www.lkab.com/ ?openform&lang=EN MEFOS, Luleå http://www.mefos.se/ SSAB, Stockholm http://www.ssab.com Ruukki, Helsinki http://www.ruukki.com

16.11 References 1 IPCC, Climate Change 2007: Mitigation, Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Metz



2

3 4 5

6 7

8

9

10 11 12

13

14

15

16 17

18 19

519

B, Davidson O R, Bosch P R, Dave R, Meyer L A (eds), Cambridge University Press, Cambridge, UK, 2007 IEA, Energy Technology Perspectives 2008: Scenarios and Strategies to 2050, OECD, Paris, 2008; total greenhouse gas emissions from human activities in 2004 were 49 billion tonnes (IPCC Working Group III, Climate Change 2007). Birat J P, worldsteel, personal communication, February 2009. Bellevrat E and Menanteau P, Scenarios and economic modeling, Revue de Métallurgie, 9, 2009, 318–324. Birat J-P, Addressing the climate change challenge: ULCOS breakthrough program, 157th ISIJ meeting, International Organized Sessions, Environmental and Energy Technology/High Temperature Processes, 30 March 2009, Tokyo, Japan, CAMPISIJ, 01/03/2009. www.ulcos.org. Birat J-P, Borlée J, Korthas B, van der Stel J, Meijer K, Günther C, Halin M, Bürgler T, Lavelaine H and Treadgold Ch, ULCOS program: a progress report in the Spring of 2008, SCANMET III, 3rd International Conference on Process Development in Iron and Steelmaking, 8–11 June 2008, Luleå, Sweden. Birat J-P, The sustainability of steel. Iron as a backbone of urban and societal metabolism, International Steel Technologies Symposium, 2-5 November 2008 Taiwan. Birat J-P, Lorrain J-P, de Lassat Y, The “CO2 Tool”: CO2 emissions & energy consumption of existing & breakthrough steelmaking routes, Revue de Métallurgie, 9, 2009, 325–336. www.hismelt.com. Zuo G and Hirsch A, The trial of the top gas recycling blast furnace at LKAB’s EBF and scale-up, Revue de Métallurgie, 9, 2009, 387–392. Christmas I, IISI-37/03/Report of the Secretary General, 6 Oct 2003, available at: http://www.worldsteel.org/?action=newsdetail&jaar=2003&id=83 (accessed January 2010). Matsuzaki S, Yonezawa K, Saito K and Naito M, Possibility of hydrogen reduction in iron-making process (COURSE 50 program in Japan), 157th ISIJ Meeting, International Organized Sessions, Environmental and Energy Technology/High Temperature Processes, 30 March 2009, Tokyo, Japan, CAMP-ISIJ, 01/03/2009. Lee S, Jung J, Kim K and Kim S, Improvement of reduction rate of iron ore using hydrogen-enriched synthetic gas, 157th ISIJ Meeting, International Organized Sessions, Environmental and Energy Technology/High Temperature Processes, 30 March 2009, Tokyo, Japan, CAMP-ISIJ, 01/03/2009. Schorsch L, North American Steel Industry Searches for Steelmaking Breakthrough, American Metal Market addressing issues of concern to the metals community, April 1, 2008. Olynyck S, The future of energy efficiency in canada and abroad, ENERGY 2007, 27–28 November 2007, Ottawa, Canada. Birat J P, Mitigation of greenhouse gas emissions in the steel sector, with a focus on biomass issues, III Conférencia regional sobre mudanças globais: América do Sul, 4–8 November 2007, Saõ Paulo, Brazil. Grumbach M, formerly ArcelorMittal, Traitement chimique des gaz d’échappement, personal communication, 2008. Young Kim J, Hana K and Dong Chu H, CO2 absorption with low concentration ammonia liquor, in Gale J, Herzog H and Braitsch J (eds), Greenhouse Gas Control


520

20

21

22

23

24

25 26

27 28 29 30

31 32 33 34

35

36

37

Developments and innovation in CCS technology Technologies 9, Proceedings of the Ninth International Conference on Greenhouse Gas Control Technologies (GHGT9), Energy Procedia, 1, 2009, 757–762. Hong Duca N, Chauvya F and Herri J-M, CO2 capture by hydrate crystallization – a potential solution for gas emission of steelmaking industry, Energy Conversion and Management, 48(4), 2007, 1313–1322. Birat J-P, Steel and CO2 – the ULCOS Program, CCS and mineral carbonation using steelmaking slag, invited paper, 1st International Slag Valorisation Symposium, 6–7 April 2009, Belgium, Leuven. ULCOS, European Steel Technology Platform – new steel initiatives for a sustainable society, March 2009, available at: http://www.ulcos.org/en/docs/Estep%20press%20 release%20March%205%20EN.doc (accessed January 2010). EC, ULCOS II has been shaping up, Brussels, Belgium, 2009, available at: ftp:// ftp.cordis.europa.eu/pub/estep/docs/ulcos_ii_sc_estep_en.pdf (accessed January 2010). Yonezawa K, Longitudinal vision of steelmaking industries from an environmental standpoint in Japan through COURSE 50 project, SCANMET III, 3rd International Conference on Process Development in Iron and Steelmaking, 8–11 June 2008, Luleå, Sweden. www.scanmet.info Goff F, Lackner K S, Carbon dioxide sequestering using ultramafic rocks, Environmental Geosciences, 5, 1998, 89–101. Rawlins C H, Richards V L, Peaslee K D and Lekakh S N, Sequestration of CO2 from steelmaking offgas by carbonate formation with slag, AISTech Proceedings, 1–4 May, 2006, 1133–1144. BRGM, personal communication. Bourgeois F and Bodéna F, Experimental & modelling advances in mineral carbonation for CO2 capture and storage, personal communication, 2005. Le Nindre Y-M, BRGM, personal communication, 2008. BRGM, Que signifie vraiment le stockage geologique de CO2, Orleans, France, available at: http://www.brgm.fr/brgm/CO2/fichiers/CO2GeonetFr_protected.pdf (accessed February, 2010) and Y-M. Le Nindre, BRGM, personal communication, 2006. Birat J-P, CCS and the steel industry, International Conference on CCS regulation for the EU and China, 18–19 February 2009, Brussels, Belgium. http://www.yorkshire-forward.com/ (accessed January, 2010). Baglin N, ArcelorMittal, personal communication, 2009. IEA, Energy Technology Perspectives 2008: Scenarios and Strategies to 2050, OECD, Paris, 2008, in support of the G8 Plan of Action, IEA, OCDE, 2008, see page 279 et seq. Christmas I, Fit to take on the new challenges, Metal Bulletin weekly, 12 January 2009, available at: http://www.worldsteel.org/index.php?action=storypages&id=317 (accessed January, 2010). Sarkozy renews pressure for CO2 border tax, Euractiv.com, 14th September 2009, available at http://www.euractiv.com/en/climate-change/sarkozy-renews-pressureco2-border-tax/article-185387 (acccessed February, 2010). Directive EC 2009/31/EC of the European Parliament and of the Council of 23 April 2009 on the geological storage of carbon dioxide and amending Council Directive 85/337/EEC, European Parliament and Council Directives 2000/60/EC, 2001/80/EC, 2004/35/EC, 2006/12/EC, 2008/1/EC and Regulation (EC) No 1013/2006, Official Journal of the European Union, L140, 114–135, available at: http://eur-lex.europa.



38

39 40 41

521

eu/LexUriServ/LexUriServ.do?uri=OJ:L:2009:140:0114:0135:EN:PDF (accessed February, 2010). EPE, Scénarios de transitions vers un monde économe en carbone en 2050 : quels enjeux pour l’industrie, Entreprises pour L’Environnement, Paris, France, 2008, available at: http://www.epe-asso.org/pdf_rap/EpE_rapports_et_documents95.pdf (accessed February, 2010). Pathways to a Low-Carbon economy, McKinsey & company, Version 2 of the Global Greenhouse Gas Abatement cost curve, 2009. Birat J-P, Antoine M, Dubs A, Gaye H, de Lassat Y, Nicolle R and Roth J-L, Vers une sidérurgie sans carbone?, Revue de Métallurgie, 90, 1993, 411–421. Hansson A and Bryngelsson M, Expert opinions on carbon dioxide capture and storage–A framing of uncertainties and possibilities, Energy Policy, 37, 2009, 2273–2282.


Index

absorber, 161 absorption compounds, 170 absorption processes, 156–61 absorption stripper system, 162 accounting approach, 115 acid gas removal, 262–5, 272 ACR-700 nuclear power plant, 139 ACR 700 nuclear reactor, 134 ACT Map, 59 adsorbents carbon-based adsorbents, 184–5 hydrotalcites, 185–6 mesoporous and microporous, 184–6 porous crystals, 186 zeolites, 185 adsorption, 391–2 advanced separation processes, 392 advanced zero emission power plant, 300–2 cycle, 287, 288, 305, 335–6, 339, 344 carbon side turbine and additional firing on nitrogen side, 342 full CO2 capture, 340 AGA equation see American Gas Association equation Air Liquide, 323, 506 Air Products, 326, 335, 351 and Chemicals, 324 air separation technology, 287 air separation unit, 261, 271, 276, 299, 321, 351, 400 alite, 470–1 alumina supported sorbents, 189–90 aluminate, 471 American Gas Association equation, 416, 417 amine scrubbing, 359, 505 amine sorbents, 475

amines, 315 ammonia, 315 ammonium bicarbonate, 169 anthracites, 184, 188 antisublimation, 206 API 5L Grade X65, 422 API 5L Grade X70, 422 aquifers, 46 Arcelor Brazil, 502 ASME B31.8, 428 ASPEN, 32, 371, 373 Aspen Plus simulations, 341 autothermal reforming, 321, 336, 367–8, 400 natural gas and light hydrocarbons, 367–8 AZEP see advanced zero emission power plant aziridine, 189 Bao Steel, 502 basic oxygen furnace, 494 BATS see Borehole Audio Tracer Survey Battelle Two–Curve Model, 420 methane showing toughness effect, 421 bauxite, 371 Beck hydro station, 130 Beggs and Brill equation, 416, 417 belite, 471 Benedict–Webb–Rubin–Starling equation, 413 bentonite, 366 Berlin Natural Gas Storage Project, 437 BGL slagging gasifier, 248 black water treatment, 272 blast furnace, 492, 499 blending, 470

523 © Woodhead Publishing Limited, 2010

524

Index

BLUE MAP Scenarios, 59 boilers oxyfuel combustion, 289–93 potential applications, 293 Rankine cycles with indirect heating, 289 tests on oxyfuel combustion in a coalfired boiler, 289–93 application to fluidised beds, 291 circulating fluidised bed with CO2 recirculation, 292 CO2 purification, 290–1 retrofit and efficiencies, 291–3 booster stations, 425 Borehole Audio Tracer Survey, 460 Boudouard reaction, 501 Brayton cycles, 287, 288, 307 bridging technology, 88 brownmillerite, 327 bubbling fluidised bed, 372 building proximity distance, 428 bulk diffusion, 195 BWRS equation see Benedict–Webb– Rubin–Starling equation CaCO3–CaO cycle, 371 calcination, 477 cycle, 193 calcium carbonate, 485 calcium hydroxide, 454 calcium silicate hydrate, 454 CANada Deuterium Uranium reactor technology, 98 CANDU 6, 134 capacity constraint, 120, 121, 122 CAPEX, 392, 399 capillary module, 222 capital cost, 39, 40 capture cost, 171 capture ready, 312 carbon-based adsorbents, 184–5 carbon capture technologies, 102 carbon dioxide absorption chemical solvents characteristics, 157 rate in amine-based solvents, 168 rate in amine blends, 169 basic building blocks in transport processes, 387–95 capture processes and technologies in power plants, 15

compression, transport and injection processes and technologies, 16 concrete accelerated curing, 479–86 density as function of pressure, 385 equations of state, 413 equilibrium partial pressure, 169 EU Directive for geological storage, 81–3 expansion, 393 from cement plants, 472–8 gas composition oxyfuel capture, 412 post-combustion capture, 411 pre-combustion capture, 412 geological sequestration, 17 injection processes and technology, 435–62 phase diagram oxyfuel type capture CO2 quality streams, 414 pure CO2, 410 phase properties, 409–13 product stream, 393–5 pure CO2 properties, 409–10 quality recommendations for transport, 386–7 pipeline, 388 ship, 388 recompression distance, pressure and temperature gradient along the pipe, 418 stream composition, 82 terrestrial and ocean sequestration, 18 transport stream composition, 410–12 utilisation, 511 carbon dioxide capture, 28, 31–6 advanced absorption processes and technology, 155–79 absorption processes, 156–61 absorption stripper system scheme, 162 advantages and disadvantages, 170–1 amine-based technology, 160 applications and future trends, 172 chemical solvents characteristics, 157 CO2 equilibrium partial pressure, 169 CO2-H2O-amine phase equilibrium, 166 estimated capture costs comparison, 171


Index

expected cost for large-scale plants, 172 post-combustion maturity, 179 R&D areas for different capture technologies, 173–4 solution pH and equilibrium CO2 loading, 166 solvents heat of reaction, latent heat of vaporisation and reaction rates constants, 165 technology advancements, 166–70 technology description, 161–5 trays for methyldiethanolamine activators, 167 advanced adsorption processes and technology, 183–98 advanced membrane separation processes and technology, 203–36 capillary module, 222 CO2/N2 selectivity vs CO2 permeability, 215 CO2 permeate concentration, 228, 230 CO2 recovery ratio vs stage cut, 228 comparing membrane modules, 223, 225 cost considerations, 230–4 cryogenic CO2 capture, 206–13 current and emerging technologies, 235 feed mixture concentration, 233 frosting temperatures in flue gases, 212 future trends and conclusions, 234–6 hollow fibre module, 223 inorganic membranes main characteristics, 209–11 membrane costs as function of pressure, 233 membrane equipment configuration, 225 membrane modules, 221–3 membrane properties, 231 membrane systems performance, 213–16 permeate concentration vs CO2 recovery fraction, 232 power plant integration design, 225–9 principles, 204

525

process overview, 207–8 Robeson’s curve for CO2/CH4 separation, 215 spiral wound module, 224 advanced oxygen production systems for power plants, 320–54 advantages and limitations, 347, 351–2 air separation technologies, 322–6 future trends, 352 OSM technology for oxyfuel power plants, 326–31 power generation systems integrated with OSM units, 331–47 chemical looping combustion systems and technology in power plants, 358–74 basic principles, 359–62 technologies and potential applications, 362–4 CLC systems and technology in power plants advantages and limitations for natural gas and syngas, 364–6 CaS–CaSO4 system, 371–2 future trends, 373–4 hydrogen manufacture, 366–8 use with solid fuels, 368–71 economic and operational parameters coal-power plants, 127 IGCC power units, 132 long term out-of-province hydroelectric imports, 135 natural gas power plants, 128 NGCC power plants, 133 nuclear power units, 130, 134 pulverised coal power plants, 131 energy supply planning, 93–151 Canada’s emission trend and Kyoto’s emission target, 94 emerging energy challenge and a case from Ontario, Canada, 93–7 energy conservation strategy, 113–15 functionalised sorbents, 186–92 alumina supported, 189–90 carbon supported, 187–8 glass fibre supported, 190 polymer and resin supported, 187 porous crystals, 191 silica supported, 188–9 templated sorbents, 191–2


526

Index

xerogel supported, 190–1 zeolite supported, 190 future trends, 105–13 coal price forecast, 114 electricity demand forecast, 105–8 forecasted peak and energy demand, 107 fuel prices forecast, 108 load-duration curve linear approximation, 106 natural gas price forecast, 112 typical load-duration curve, 105 gasification processes and synthesis gas treatment, 243–77 advantages and limitations, 273–6 applications, 259–61 basic principles, 244–58 building blocks for complete systems, 261–70 future trends, 276–7 power plant as an example for a complete system, 270–3 illustrative case study, 124–50 annual CO2 emission, 138 annual cost of electricity, 148 case study results, 135–50 coal power plants fuel-switched to natural gas, 138 data, 125–35 electricity production generated to meet base-load demand, 144 electricity production generated to meet peak-load demand, 145 electricity sector annual expenditure, 146 entire fleet annual CO2 emissions, 149 estimated refurbishment cost for nuclear units, 128 new power stations construction, 136 nuclear units capacity profile, 129 percent of electricity production, 144 power allocated to meet base-load demand, 140 power allocated to meet peak-load demand, 141 total electricity production, 142 total expenditure, 147 total power allocated from each supply technology, 137

total power allocation, 140 membrane materials and design, 216–21 CO2 membrane gas absorption principle, 220 facilitated transport membranes, 219 hybrid membranes, 218–19 membrane contactors, 220–1 mixed-matrix membranes, 218 polymeric membranes performance, 217 specific surface area of contactors, 220 mesoporous and microporous adsorbents, 184–6 carbon-based, 184–5 hydrotalcites, 185–6 porous crystals, 186 zeolites, 185 Ontario current installed capacity, 103 current installed generation capacity, 94 demand growth and generating capacity portfolio, 96 existing coal-fired power plants, 105 forecasted annual base-load demand, 111 forecasted annual energy demand, 109 forecasted annual peak demand, 110 operational and out-of-service nuclear units, 104 oxyfuel combustion systems and technology in power plants, 283–315 advantages and limitations, 307–12 basic principles, 285 future trends, 313–15 technologies and potential applications, 287–307 planning model, 115–24 construction lead time constraint sample matrix, 121 indices, sets, variables, and parameters, 117 model formulation, 116 regenerable sorbents, 192–6 carbonation/calcination cycle schematic, 193 natural minerals, 192–5 synthetic, 195–6


Index

supply technologies and carbon capture and storage, 98–105 carbon dioxide capture and storage, 101–5 comparison of different technologies, 100–1 supply technologies, 98–101 carbon dioxide capture and sequestration, 243 carbon dioxide capture and storage, 1–20, 97, 383, 408 carbon dioxide capture, 31–6 economics of fossil energy plants, 32 electric power plants characteristics, 33 electricity cost, 36 emission rates and costs associated with power plants, 36 plant level modelling, 32 pre-combustion capture options, 32 processes, 31 production of hydrogen and other fuels, 35–6 carbon dioxide injection, 44–56 aquifer and injection parameters, 46 for storage only, 45–50 injection site design, 45 input parameter ranges, 51 levelised costs sensitivity to various parameters, 52 oil fields reservoir characteristics and levelised cost, 54 reservoir characteristics, 52 carbon dioxide transport, 36–44 actual and calculated pipeline diameter, 39 cost analysis reference parameters, 43 electricity cost from fossil power plants, 37 pipeline and ship costs, 44 pipeline and ship transport costs for offshore injection, 43–4 pipeline transport, 37–40 ship transport, 41–2 carbon management and stabilisation routes, 8–11 CCS systems schematic diagram, 10 stabilisation wedges concept for reducing carbon emissions, 11

527

total cost of early commercial projects, 15 cement and concrete industry, 469–88 accelerated CO2 curing of concrete, 479–86 basic principles, 470–2 CO2 capture from cement plants, 472–8 future trends, 487–8 coal-based hydrogen system build-out scenario, 58 optimisation, 58 energy supply planning, 101–5 carbon capture technologies, 102 Ontario’s current energy mix, 103–5 future trends, 19–20 general system set up oxyfuel systems with CCS, 333 oxygen selective membrane systems, 334 greenhouse gas emissions and global climate change, 2–8 average annual atmospheric CO2 concentrations, 4–5 carbon dioxide intensity by region and country 1980-2030, 7 EU-15 CO2 emissions 2006, 3 global average air and ocean temperatures, rising global average sea levels and melting of sea-ice, 6 implementation in process industries including combustion, 504 iron and steel industry, 492–518 abatement perspectives, 515–17 CO2 capture technologies, 504–9 CO2 storage, 509–15 steel sector, 503–4 steel sector CO2 emissions, 493–7 strategies to control CO2 emissions, 497–503 legitimation: results and gaps, 74–80 current status, 75–7 knowledge gaps, 78–80 main arguments and misconceptions, 77–8 legitimation and market formation regulatory and social analysis, 64–88 implementation outlook, 86–8 technological innovation system functions, 66


528

Index

market formation and direction of search, 80–6 EU and CCS market formation, 86 financing demonstrations, 83 towards mass markets, 84–5 projects in EU, 175–8 reservoir characteristics baseline case and at three depths, 52 oil fields, 54 primary coalbed methane resins, 55 safety regulation, 80–3 EU directive for the geological storage of CO2, 81–3 international frameworks, 81 techno-economic analysis and modelling, 27–61 coal-based hydrogen system optimisation, 58 component level engineering/ economic models, 30 costs range for CCS system components, 57 fossil energy system with CCS, 29 future trends, 59–61 reduction in CO2 emissions, 60 system modelling, 56–9 technological maturity and innovation system, 67–73 economic incentives, 69 stages of maturity, 68 technological innovation system, 70–3 technological maturity, 67 transport process, 389 transport technology development and innovation, 11–17 CCS component technologies, 12 CO2 compression, transport and injection processes and technology, 16 economics, regulation and planning, 14–15 global CCS projects, 13 industrial applications, 16–17 processes and technologies in power plants, 15–16 utilisation technology development and innovation, 17–19 advanced concepts, 18–19 carbon dioxide geological sequestration, 17 maximising and verifying storage in

underground reservoirs, 17–18 terrestrial and ocean sequestration and environmental impacts, 18 carbon dioxide injection, 45–50 analogues for CO2 storage and best practices from other sectors, 437–8 aquifer and injection parameters, 46 controlling parameters for injectivity, 441–9 casing shear schematic illustration, 457–8 injection well pressure and reservoir constraints, 447–9 injectivity loss, 443–5 permeability, 441–3 pressure development at the end of injection cycle, 442 rock–cement–casing interfaces including stresses around wellbore, 456 wellbore design, 445–7 deep saline aquifers, 46 site characterisation costs, 46–7 surface equipment, 47–8 well drilling costs, 47 depleted gas reservoirs, 49–50 different storage formations, 449–51 depleted oil and gas fields, 450 hydro-mechanical impact, 451 injection in coal seams, 450–1 saline formations, 449 enhanced fossil fuel recovery, 50–6 enhanced coalbed methane, 54–6 enhanced gas recovery, 51–2 enhanced oil recovery, 53–4 site design, 50 field operations, 451–3 enhanced oil recovery and CO2 reinjection, 453 Sleipner CO2 storage site simplified illustration, 452 for storage only, 45–50 future trends, 462 injection site design, 45 injection well technologies, 438–41 input parameter ranges, 51 levelised cost, 48 levelised costs sensitivity to various parameters, 52 oil fields reservoir characteristics and levelised cost, 54


Index

processes and technology, 435–62 reservoir characteristics, 52 Salah Gas Project schematic, 447 technologies for monitoring injection well integrity, 459–62 injection rates and pressures, 461 injection well integrity, 459–61 microseismic monitoring, 461–2 underground fluid injection, 436 well integrity, 453–9 casing shear failure, 456–8 cement degradation, 454–5 corrosion, 458–9 debonding, 455–6 induced seismic events, 453–4 carbon dioxide looping, 192 carbon dioxide recovery commercial plants, 159 historical development, 160 ratio vs stage cut for three different flow mode patterns, 228 carbon dioxide sequestration, 102 carbon dioxide transport energy requirements for six transport processes as function of feed gas pressure, 398 as function of feed gas volatile content, 399 as function of temperature for heat rejection, 399 gas purification, compression and liquefaction processes and technology, 383–405 future trends and future work, 404 interface between capture and transport, 400–2 quality recommendations for pipelines and ship transport, 386–7 sensitivity analysis, 395–400 ship to pipeline and pipeline to ship processes, 402–3 transport pressures selection, 385–6 infrastructure and pipeline technology, 408–30 CO2 phase properties, 409–13 future trends and future work, 429 pipeline transport, 414–23 ship transport, 423–5 transport economics, 425 large-scale transport infrastructure, 425–8

529

cumulative plot of emissions with sources, 426 pipelines offshore, 427 regulations, 427–8 overview and building blocks in processes, 387–95 carbon dioxide expansion, 393 carbon dioxide product stream, 393–5 compression and cooling, 390–1 condensation, 392 pumping, 392 unwanted components removal, 392 volatile gases removal, 393 water and other liquids removal in vapour–liquid separator drums, 391 water removal by adsorption, 391–2 process in CCS chain, 389 process superstructure, 394 recycling combustibles back into the capture process, 401 transport processes energy requirements, 395 carbon havens, 516 carbon leakage, 516 carbon monoxide shift, 265–7 three-stage system, 266 carbon supported sorbents, 187–8 carbon tax, 85 carbonation cycle, 193 process, 480–2 casing shear, 456–8 cement, 471 cement and concrete industry accelerated CO2 curing of concrete, 479–86 carbonation process, 480–2 chemical reactions, 480 critical operational parameters, 482–3 pore structure changes with calcium carbonate deposition, 484 potential for concrete direct carbonation curing using flue gases, 485 potential limitations to carbon dioxide uptake, 483–4 sequential steps involved, 481 waste cement materials CO2 curing, 485–6


530

Index

CCS technology, 469–88 basic principles, 470–2 future trends, 487–8 CO2 capture from cement plants, 472–8 isolated calcination, 477–8 oxygen combustion, 475–7 post-combustion capture, 474–5 modern pre-heater/pre-calciner dry rotary kiln layout, 473 various components associated with cement and concrete use, 471 cement bond logs, 459 cement job parameters, 459 cement kiln dust, 474, 485 Central Basin Pipeline, 416 ceramic autothermal recovery, 322, 325 chemical-looping combustion, 192, 303, 322, 324–5, 360 advantages and limitations for natural gas and syngas, 364–6 basic principles, 359–62 carbon dioxide capture in power plants, 358–74 CaS–CaSO4 system, 371–2 cycle, 287, 289 future trends, 373–4 hydrogen manufacture, 366–8 natural gas and light hydrocarbons ATR, 367–8 metal oxide looping cycle, 360 solid fuels combustion process schematic, 372 system layout, 304 technologies and potential applications, 362–4 Grace 10 kW chemical-looping reactor, 363 use with solid fuels, 368–71 chemical-looping combustor, 362 chemical looping reforming, 236 chemical-looping with oxygen uncoupling, 370 chemisorption, 505 Chinese bituminous coal, 369 chromium steel, 441 cladding, 419 Clean Energy Systems cycle, 287, 293 clean gas shift, 267 Climate Change Act, 7 clinker, 469, 470 CLOU see chemical-looping with oxygen uncoupling

CO2 Breakthrough Program, 502 CO2 prevented emission recuperative advanced turbine energy cycle, 307 coal, 8, 99 plasticisation, 450 power, 139 power plants, 99, 100, 103, 105, 141 economic and operational parameters, 127 fuel-switched to natural gas, 138 power stations, 99, 100 CO2–ECBM technology, 452 cogeneration, 99, 100 CO2–H2O steam cycle, 341 cold flash gas, 393 Commission Directive 96/61/EC, 20 Commission Directive 85/337/EEC, 20 Compact Gasifier, 276 compression, 390–1 concrete, 469 condensation, 392 ConocoPhillips, 255 construction lead time, 120–2 sample matrix, 121 converter, 494 cooling, 390–1 COOPERATE see CO2 prevented emission recuperative advanced turbine energy cycle COREX process, 502 corrosion, 458–9, 462 Cortez pipeline, 416 COS hydrolysis, 267 CPLEX 10, 116 cryogenic air separation unit, 307, 323–4, 476 cryogenic CO2 capture, 206–13 cryogenic distillation, 287 daisy grid model, 515 debonding, 455–6 decarbonising, 499 Decew hydrostation, 130 deep saline aquifers, 46 density log data, 459 depositional process, 449 desulphurisation, 262 diagenetic process, 449 diethanolamine, 187 digitalised geophones, 461 direct reduced iron, 499


Index direct reduction process, 494, 495, 499 directional drilling system, 446 distillation, 393 dry feed system, 248 Dynamis project, 422 E-Gas, 252, 255, 256 ECBM see enhanced coal bed methane recovery ECN membranes, 329 Econamine process, 160 EE50, 115 EE100, 115 EE100 Plus Standards, 115 EIA’s AE2005, 108, 113 electric arc furnace route, 495 electrowinning process, 499 energy conservation strategy, 113–15 enhanced coal bed methane recovery, 17, 446, 450 enhanced oil recovery, 384, 408, 439, 446 entrained flow gasifiers, 247 entrepreneurial experimentation, 73 EOR see enhanced oil recovery ethylenediamine, 187 ETS see EU Emissions Trading Scheme EU 2003/87/EC, 84 EU Directive for Geological Storage of CO2, 81–3, 85 EU Emissions Trading Scheme, 69, 83 monitoring and reporting guidelines, 84–5 EU Energy and Climate Package, 84 EU Framework 6 ENCAP project, 410 European ENCAP project, 297, 315 experienced-based approach, 115 experimental blast furnace, 506 facilitated transport membranes, 219 feed systems, 248 ferrite, 471 FeTiO3, 370 fixed bed gasifiers see moving bed gasifiers Flour Daniel Econamine process, 158 flue gas desulphurisation, 291 2-D Fluent model, 372 fluid bed gasifiers, 247 fly ash, 99, 188 fossil energy plants economics with carbon dioxide capture, 32–5

531

avoided cost of CO2 emissions, 35 CO2 capture cost, 35 cost of electricity, 34–5 plant capital investment cost, 32–4 fossil-fueled power plants, 15, 126 cost of electricity, 37 fossil fuels, 8 fracture pressure, 447, 448 fuel prices forecast, 108 coal, 114 natural gas, 112, 113 fuel-selection, 122–3 functionalised sorbents, 186–92 Future Energy GSP see Siemens process gas reservoirs, 49–50 gas separation, 213 gas turbine technology, 277 gas turbines, 98 oxyfuel combustion, 296–300 cycles, 308–12 oxyfuel concepts comparison, 310, 312 post and pre-combustion capture and oxyfuel GT cycles, 311 with direct heating and externally generated oxygen, 296–300 Graz cycle, 299–300 Matiant cycle, 297–9 technology development, 297 with internally generated oxygen, 300–4 advanced zero emission plant, 300–2 AZEP cycle using MCM and HX, 302 chemical-looping combustion, 303 mixed ceramic membrane ionising oxygen, 301 gasification advantages and limitations, 273–6 availability, 275 capital requirements, 275–6 efficiency, 273–4 environmental impact, 274–5 and synthesis gas treatment for carbon dioxide capture, 243–77 future trends, 276–7 applications, 259–61 chemicals including synthetic fuels, 260 IGCC block flow diagram, 259


532

Index

methanol plant block flow diagram, 260 polygeneration, 261 power, 259–60 basic principles, 244–58 chemistry and thermodynamics, 244–5 commercial processes, 252–8 definition, 244 process characteristics, 246 process realisation, 245–52 building blocks for complete systems, 261–70 air separation, 261 flows in natural gas and syngas-fired gas turbine, 268 gas turbine output, 269 MDEA flowsheet, 263 syngas- and hydrogen-fired combustion turbines, 268–70 syngas treatment, 261–7 three-stage CO shift system, 266 two-stage Selexol flowsheet, 264 cold gas efficiency, 249 commercial processes, 252–8 E-Gas, 255 ECUST, 257–8 GE energy, 253–4 Lurgi, 258 Shell and Prenflo technologies, 254–5 Siemens, 255–6 power plant, 270–3 combined cycle power plant, 273 gas generation, 271–3 IGCC block flow diagram with carbon dioxide capture, 271 raw gas analyses, 272 process realisation, 245–2 bed type, 245–8 Conoco Philips E-Gas gasifier, 256 feed preparation, 248 gasifier containment systems, 251 GE quench gasifier, 253 Lurgi dry bottom gasifier, 258 operating temperature, 248–9 oxidant, 249–50 primary gas cleaning, 252 primary syngas cooling, 250–1 reactor containment, 250 Shell coal gasification process, 254 Siemens SFG gasifier, 257

two-stage gasification, 252 syngas treatment, 261–7 acid gas removal, 262–5 carbon monoxide shift, 265–7 COS hydrolysis, 267 gasifier containment systems, 251 GE energy, 253–4 GE quench gasifier, 253 General Algebraic Modelling System, 116 General Electric J79 gas turbine, 296 geological storage, 81–3 geomechanical processes, 450 geophysical well integrity logs, 460 geostorage, 511–12, 514–15 glass fibre supported sorbents, 190 global climate change, 2–8 global warming potential, 1 Grace 10 kW chemical-looping reactor, 363 Graz cycle, 287, 288, 299–300 power plant layout, 300 green hydrogen, 502 GreenGen, 14 greenhouse gas effect, 2, 307, 320, 486 greenhouse gas emissions, 2–8 greenhouse gases, 492–3 gypsum, 371, 471, 485 H2 Fleet Leader, 269 halite precipitation, 443–4 heat recovery steam generator, 288, 391 HEATEX, 162 hexagonal mesoporous silica, 188 high-pressure steam turbine, 295 high-temperature syngas desulphurisation process, 276 HIsarna process, 499, 501, 503, 508 schematics, 500 HIsmelt process, 499 hollow fibre module, 223 hybrid membranes, 218–19 hydrate, 423 hydroelectric power stations, 99–100, 101, 128, 130 hydrogen, 366–8 hydrogen delivery system, 59 Hydrogen Energy 390 MWe CCS/EOR, 254 hydrogen-fired combustion turbines, 269–70 hydrostatic pressure, 447 hydrotalcites, 185–6


Index

533

Hygensys, 236 HYSIS, 32 IGCC see integrated gasification combined cycle illmenite, 365 immobilised amine sorbents, 187 in-process capture, 515 IN-STRIP, 162 Independent Electricity System Operator, 106 injection rate, 45 injection well, 436 CO2 injection and well integrity, 453–9 casing shear failure, 456–8 cement degradation, 454–5 corrosion, 458–9 debonding, 455–6 induced seismic events, 453–4 integrity monitoring technologies, 459–62 injection rates and pressures, 461 injection well integrity, 459–61 microseismic monitoring, 461–2 pressure and reservoir constraints, 447–9 technologies, 438–41 simplified vertical CO2 injection well, 439 injectivity loss, 443–5 geochemical reaction products and effects, 444 halite precipitation, 443–4 impurities effects on injection stream, 444–5 mobilised fine particles, 444 Integrated Environmental Control Model, 130 integrated gasification combined cycle, 9, 100, 244, 270, 305, 323, 365 block flow diagram, 259 integrated steel mill, 493–4 simplified flow sheet, 495, 505 intermediate-pressure steam turbine, 295 ion diffusion, 195 ion transport membrane, 307, 321 Iron & Steel Association, 502 iron and steel industry carbon dioxide capture technologies, 504–9 mature CO2 capture technologies comparison, 510

principle of pressure swing adsorption CO2-scrubbing techniques, 507 vacuum pressure swing adsorption of Air Liquide, 506 carbon dioxide emissions, 493–7 ArcelorMittal Florange’s blast furnace skyline, 493 ISM simplified flow sheet, 495, 505 steel-making production routes, 496 carbon dioxide storage, 509–15 CO2 utilisation, 511 daisy grid model of CCS in Europe, 515 geological structures favourable for storage in Belgium, 514 geostorage, 511–12, 514–15 mineral sequestration, 509, 511 ocean storage, 509 resources and reserve concepts, 512 steel mills, 513 CCS technology, 492–518 and CO2 abatement perspectives, 515–17 for steel sector, 503–4 strategies to control CO2 emissions, 497–503 breakthrough technologies for cutting CO2 emissions, 498 HIsarna process schematics, 500 TGR–BF process schematics, 500 ULCORED process schematics, 501 isolated calcination, 477–8 schematic representation, 478 Japanese COURSE 50, 502, 508 Joule II project, 511 Joule Thomson effect, 418 JT-valve, 393 Kalinin Atomic Power Plant, 436 kerosene, 296 Kerr–McGee/ABB Lummus amine technology, 156 Kirovo-Chepetsk Chemical, 436 Kyoto protocol, 93 Lambton coal power plant, 139 large-scale fuel channel replacement, 103 LEANGAS, 162 LEANSOL, 162 legitimation, 73


534

Index

defined, 74 market formation regulatory and social analysis, 64–88 implementation outlook, 86–8 technological innovation system functions, 66 results and gaps in social scientific research, 74–80 current status on public opinion, 75–7 knowledge gaps regarding public opinion and involvement, 78–80 main arguments and misconceptions, 77–8 light hydrocarbons, 296 auto thermal reforming, 367–8 limestone, 194–5 linearisation method, 119 liquefaction, 393 Liquefied Energy Chain, 404 liquefied natural gas, 404, 423, 424 liquefied petroleum gas, 423, 424 lithium zirconate, 196 LNC, 331 lock hopper system, 276 looping cycle, 359 low-pressure steam turbine, 295 low temperature frost evaporators, 212 Lurgi dry bottom gasifier, 248, 258 MARKAL, 31, 59 market formation, 71, 80–6 mass markets EU emissions trading scheme, 84 EU ETS monitoring and reporting guidelines, 84–5 Norway, 84–5 regulation for incentivising CCS, 84–5 Matiant cycle, 287, 288, 297–9, 305, 307 T–S diagram with regeneration and reheat, 299 zero emission gas turbine cycle scheme, 298 MCM-41, 189 membrane contactors, 220–1 membrane modules, 221–3 commercially available CO2-separation membrane modules, 226 conceptual scheme, 222 other parameters for design, 226 membrane separation, 234

advanced processes and technology for CO2 capture, 203–36 CO2 membrane materials and design, 216–21 cost considerations, 230–4 cryogenic carbon dioxide capture, 206, 212–13 design for power plant integration, 225–9 future trends, 234–6 membrane modules, 221–3, 223–5 membrane systems performance, 213–16 membrane wall, 250 mercury, 275 removal, 272 MESSAGE, 31 metal organic frameworks, 186 methane, 54–6 methyldiethanolamine, 263 micro turbine, 347 microseismic monitoring, 461–2 microseismicity, 453 Milano cycle, 307, 346 mineral sequestration, 509, 511 Mitsubishi, 251 Mitsubishi-Kansai technology, 158 mixed Brayton/Rankine cycle, 288, 299 mixed conducting membranes, 321, 327, 339 mixed-integer linear programming, 115, 116 mixed-integer non-linear programming, 97, 116 mixed ionic electronic conducting membranes, 321 mixed-matrix membranes, 218 MOF-177, 186 molecular basket, 189 molecular sieves, 185 monoethanolamine, 102, 156, 187, 205 Moody friction factor, 416, 417 moving bed gasifiers, 247 nanocasting, 191 natural gas auto thermal reforming, 367–8 CLC advantages and limitations, 364–6 power plants, 126 power stations, 98–9, 100 natural gas combined cycle, 231–3


Index power plants, 98, 100 natural gas-fired gas-steam combined cycles, 171 natural gas-fired gas turbines, 171 Near Zero Emission Coal, 14 NEMS, 31 NGCC see natural gas-fired gas-steam combined cycles NGGT see natural gas-fired gas turbines Ni-based oxygen carrier, 369 Ni/NiO system, 364, 369 Niagara Plant Group, 104 nitrogen, 323 nitrogen oxide emissions, 274–5 noise logs, 460 nominal pipe size, 38 non-porous membranes, 214 NORSK Hydro, 335, 339, 351 nuclear power, 103 plants, 100, 126, 128 stations, 98 ocean storage, 509 One-Step Reforming, 236 Ontario case study, 124–50 data, 125–35 existing power plants, 126–30 new power plants, 130–5 results, 135–50 coal power plants fuel-switched to natural gas, 138 economic analysis, 145–50 new construction, fuel switching, and CCS retrofit, 135 power allocation and electricity production, 139–45 Ontario Power Generation, 95 Oosterkamp and Ramsen quality specification, 411 OPEX, 392, 399 Ordinary Portland Cement, 471 OSM see oxygen selective membranes OSPAR CO2 guidelines, 82 overlaying, 419 oxidant, 249–50 oxycoal boilers, 307–8 Oxycoal–AC cycle, 307, 344, 345, 346 oxyfiring combustion, 359 oxyfuel combustion, 31, 320 advantages and limitations, 306–7 capture readiness, 312 oxycoal boilers, 307–8

535

oxyfuel combustion gas turbine cycles, 308–12 basic principles, 285 oxyfuel boiler with FGR fluid, 286 carbon dioxide capture in power plants, 283–315 circulating fluidised bed with CO2 recirculation, 292 future trends, 313–15 stack downwards, 284 technologies and potential applications, 287–307 boilers, 289–93 gas turbines, 296–300 gas turbines with internally generated oxygen, 300–4 IGCC, 305 other oxyfuel combustion cycles, 306–7 SOFC-CC, 305–6 steam turbine cycles with direct heating, 293–6 systems classification, 287–9 technologies classification, 288 tests in a coal-fired boiler, 289–93 water cycle power plant layout, 294 oxyfuel combustion power cycles, 205 oxygen, 205 oxygen carrying capacity, 361 oxygen combustion, 475–7 oxygen production systems advantages and limitations, 347, 351–2 air separation technologies, 322–6 ceramic autothermal recovery process, 325 chemical-looping, 324–5 cryogenic air separation, 323–4 demands for zero-emission power plants, 322–3 oxygen selective membranes, 325–6 pressure and vacuum swing adsorption, 324 future trends, 352 OSM technology for oxyfuel power plants, 326–31 performance comparison systems, 349–50 power generation systems integrated with OSM units, 331–47 full oxidation in membrane unit, 335–6


536

Index

partial oxidation in membrane unit, 336–9 partial oxidation power cycle, 338 post-oxidation in membrane unit, 346–7 without oxidation in membrane unit, 339–46 power plants with integrated carbon dioxide capture, 320–54 SOFC–GT post-oxidation power cycle, 348 oxygen selective membranes, 321, 322, 325–6 technology for oxyfuel power plants, 326–31 ABO3-d and A2BO4 + d structures, 327 damaged perovskite membrane during fabrication, 330 MIEC membrane, 328 SCFC oxygen flux, 331 sinter shrinkage, weight loss and heat evolution during sintering, 329 oxygen transport membrane, 309, 321 Peng–Robinson equation, 413, 417 periodic cement bond logs, 459 physical washes, 263 physisorption, 505 Pipeline Safety Regulations, 427 pipeline transport, 37–40, 385–6, 414–23 CO2 on recompression distance, pressure and temperature gradient along the pipe, 418 CO2 quality recommendations, 388 corrosion and hydrate formation, 422–3 costs for offshore injection, 43–4 design parameters, 38 flow equations, 415–17 fracture, 419–22 Battelle TCM for methane, 421 materials, 419 pipeline cost, 39–40 levelised cost for onshore pipeline transport, 41 offshore, 40 onshore, 39–40 onshore vs offshore, 41 pipeline design, 37–8 actual vs calculated diameters, 38 offshore pipeline design, 39

pipeline hydraulics, 417–19 process P1, 396 ship to pipeline and pipeline to ship processes, 402–3 plant capital investment cost, 32–4 polyethyleneimine, 187 polygeneration, 261 poor sweep, 447 porous crystals, 186, 191 porous membranes, 214 Portland cement, 455 POSCO, 502 post-combustion capture technology, 31, 102, 204–5, 291, 310, 320–1, 474–5 power generation systems, 331–47 power plants adsorption processes and technology for CO2 capture, 183–98 advanced absorption processes and technology for CO2 capture, 155–79 as an example for a complete system, 270–3 emission rates and costs with and without CO2 capture, 36 power stations, 120 PR equation see Peng–Robinson equation Praxair system, 347 pre-calciner kiln, 472 pre-combustion capture, 102, 205, 321 pre-combustion decarbonisation systems, 31, 347 pre-reduction furnace, 494, 495 Prenflo technologies, 254–5 pressure swing adsorption, 184, 322, 505 pressurised heavy water reactor, 98 pressurised thermogravimetric analyser, 365 primary syngas cooling, 250–1 PRISM technology, 324 public opinion, 74–80 current status of research on CCS, 75–7 awareness, 75–6 research issues, 76–7 pulverised coal-fired power generation systems, 313 pumping, 392 pyroelectrolysis, 499 Rankine cycle, 287, 288, 303, 307 with indirect heating, 289


Index raw gas shift, 266–7, 272 raw meal, 470 reactor containment, 250 Rectisol, 265, 305, 392, 400 Redox Technologies, 236 refractory lining, 250 regenerable sorbents, 192–6 natural minerals, 192–5 improving calcium sorbent performance, 195 limestone deactivation mechanism, 194–5 synthetic sorbents, 195–6 regeneration energy, 161 regenerative adsorption columns, 391 Regional Carbon Sequestration Partnerships, 14 reservoirs, 17–18 characteristics, 52 resource mobilisation, 71 R.H Saunders hydro station, 130 RICHGAS, 161 Richter scale, 461 rock shear, 457 rule of thumb method, 50 safe injection pressure, 448 SAIL, 502–3 Salah–Algeria project, 439, 446 San Juan Basin, 55 saturation pressure, 421 SBA-15, 188 SCADA data, 417 SCF, 327 SCFC, 331 SCGP see Shell coal gasification process Selexol, 263, 305, 400 two-stage Selexol flowsheet, 264 semi-closed oxyfuel cycle, 287 semi-pressurised ship, 386, 423–4 sequestration, 18, 359 Sheep Mountain Pipeline, 419 Shell coal gasification process, 254 Shell partial quench technology, 255 Shell technologies, 252, 254–5 Shell’s Pernis refinery, 261 ship transport, 41–2, 423–5 advantage over pipelines, 424–5 CO2 quality recommendations, 388 costs, 42 offshore injection, 43–4 levelised cost, 43

537

process S1, 397 ship to pipeline and pipeline to ship processes, 402–3 transport system design, 42 Siemens process, 255–6 Siemens SFG gasifier, 257 Siemens SGT5-4000F gas turbine, 337, 341 silica supported sorbents, 188–9 simple cycle gas turbines, 98, 99 sintering, 477 slag removal, 272 Sleipner–Norway project, 439, 440, 449, 452 Smelting Reduction Process, 492, 499 Snövhit pipeline, 428 Soave–Redlich–Kwong equation, 413 solid fuels chemical-looping process schematic for combustion, 372 CLC technology, 368–71 solid oxide fuel cell, 299, 305–6, 322 as combustion chamber in gas turbine cycle, 306 solvent flow rate, 164 Sonar Log, 460 Sorb KX35, 196 Sorb NX30, 196 sorbents functionalised, 186–92 alumina supported, 189–90 carbon supported, 187–8 glass fibre supported, 190 polymer and resin supported, 187 porous crystals, 191 silica supported, 188–9 templated, 191–2 xerogel supported, 190–1 zeolite supported, 190 regenerable, 192–6 natural minerals, 192–5 synthetic, 195–6 sound survey, 460 South African bituminous coal, 369 spiral wound module, 224 SRK equation see Soave–Redlich–Kwong equation stabilisation wedges, 9, 11 stage cut, 226, 227 steam turbine, 98 cycles with direct heating, 293–6 sted, 493


538

Index

steel, 492 steel mills, 513 Steinour formula, 483 STRIPPER, 161, 162 sublimation, 206 sulphur emissions, 274 sulphur recovery, 272–3 Summit Power 300 MWe, 256 supercritical CO2, 419, 455 supply-push case, 108 syngas, 99, 244, 305 CLC advantages and limitations, 364–6 cooling, 254 syngas-fired combustion turbines, 268–9 syngas treatment, 261–7 acid gas removal, 262–5 chemical and solvent processes, 263 physical washes, 263 Rectisol, 265 Selexol, 263 carbon monoxide shift, 265–7 clean gas shift, 267 raw gas shift, 266–7 COS hydrolysis, 267 mercury removal, 267 tanker trucks, 36 TCM see Battelle Two–Curve Model techno-vert case, 108, 113 technological innovation system, 65, 70–3 functions, 66, 72–3 structural elements, 70 temperature logs, 460 templated sorbents, 191–2 templating, 191 tetraethylenepentamine, 187 Texaco coal gasifier, 253 see also GE Energy TGR-BF process see top gas recycling blast furnace process thermal decomposition, 475 top gas recycling blast furnace process, 499

schematics, 500 transport economics, 425 tubular membranes, 328 two electrolysis variants, 499 ULCOLYSIS, 499 ULCORED process, 492, 495, 503, 508 schematics, 501 ULCOS Program see Ultra Low CO2 Steelmaking Program ULCOWIN, 499 Ultra Low CO2 Steelmaking Program, 498, 506 underground fluid injection, 436 Underground Injection Control Program, 446 vacuum pressure swing adsorption, 505 vacuum swing adsorption, 322 vapour–liquid separator drums, 391 well drilling costs, 47 Western Research Institute, 325 wet feed system, 248 wind power plants, 100 wuesite, 367 xerogel, 190–1 xerogel supported sorbents, 190–1 ZEITMOP see zero emission ion transport membrane oxygen power zeolite supported sorbents, 190 zeolite 13X, 190 zeolites, 185 ZEPP see zero emission power plants zero CO2 emission, 285 zero emission ion transport membrane oxygen power cycle, 341, 343, 344 zero emission power plants, 296–7


Developments and Innovation in Carbon Dioxide (CO2) Capture and Storage Technology: Volume 1: Carbon Dioxide (CO2) Capture, Transport and Industrial Applications (Woodhead Publishing Series in Energy)

Developments and Innovation in Carbon Dioxide (CO2) Capture and Storage Technology: Volume 2: Carbon Dioxide (CO2) Storage and Utilisation (Woodhead Publishing Series in Energy)

Carbon capture and storage

Oxy-Fuel Combustion for Power Generation and Carbon Dioxide (CO2) Capture

Capture and Utilization of Carbon Dioxide with Polyethylene Glycol

Carbon Dioxide Populations and Communities

Solubility in Supercritical Carbon Dioxide

Carbon Dioxide and Environmental Stress

Transport Energy and CO2 : Moving towards Sustainability

Carbon Dioxide Sequestration and Related Technologies

Carbon Dioxide and Terrestrial Ecosystems (Physiological Ecology)

Acid Gas Injection and Carbon Dioxide Sequestration

The Phanerozoic Carbon Cycle: CO2 and O2

Carbon Dioxide and Environmental Stress (Physiological Ecology)

Carbon Dioxide as Chemical Feedstock

Carbon Dioxide Angiography: Principles, Techniques, and Practices

Carbon dioxide and climate: a second assessment

Cost and Performance of Carbon Dioxide Capture from Power Generation (IEA Energy Papers)

Carbon Dioxide as Chemical Feedstock

Supercritical Carbon Dioxide in Polymer Reaction Engineering

Carbon Capture: Sequestration and Storage (Issues in Environmental Science and Technology)

Caching the Carbon: The Politics and Policy of Carbon Capture and Storage

Natural Extracts Using Supercritical Carbon Dioxide

Carbon Dioxide Utilization for Global Sustainability

Natural Extracts Using Supercritical Carbon Dioxide

Acid Gas Injection and Carbon Dioxide Sequestration (Wiley-Scrivener)

Stand-alone and Hybrid Wind Energy Systems: Technology, Energy Storage and Applications (Woodhead Publishing Series in Energy)

Green Chemistry Using Liquid and Supercritical Carbon Dioxide

Capturing king coal deploying carbon capture and storage systems in the U.S. at scale

Capturing King Coal Deploying Carbon Capture and Storage Systems in the U.S. at Scale

Energy Economics: CO2 Emissions in China

Developments and Innovation in Carbon Dioxide (CO2) Capture and Storage Technology: Volume 1: Carbon Dioxide (CO2) Capture, Transport and Industrial Applications (Woodhead Publishing Series in Energy)

Developments and Innovation in Carbon Dioxide (CO2) Capture and Storage Technology: Volume 2: Carbon Dioxide (CO2) Storage and Utilisation (Woodhead Publishing Series in Energy)

Carbon capture and storage

Oxy-Fuel Combustion for Power Generation and Carbon Dioxide (CO2) Capture

Capture and Utilization of Carbon Dioxide with Polyethylene Glycol

Carbon Dioxide Populations and Communities

Solubility in Supercritical Carbon Dioxide

Carbon Dioxide and Environmental Stress

Transport Energy and CO2 : Moving towards Sustainability

Carbon Dioxide Sequestration and Related Technologies

Carbon Dioxide and Terrestrial Ecosystems (Physiological Ecology)

Acid Gas Injection and Carbon Dioxide Sequestration

The Phanerozoic Carbon Cycle: CO2 and O2

Carbon Dioxide and Environmental Stress (Physiological Ecology)

Carbon Dioxide as Chemical Feedstock

Carbon Dioxide Angiography: Principles, Techniques, and Practices

Carbon dioxide and climate: a second assessment

Cost and Performance of Carbon Dioxide Capture from Power Generation (IEA Energy Papers)

Carbon Dioxide as Chemical Feedstock

Supercritical Carbon Dioxide in Polymer Reaction Engineering

Carbon Capture: Sequestration and Storage (Issues in Environmental Science and Technology)

Caching the Carbon: The Politics and Policy of Carbon Capture and Storage

Natural Extracts Using Supercritical Carbon Dioxide

Carbon Dioxide Utilization for Global Sustainability

Natural Extracts Using Supercritical Carbon Dioxide

Acid Gas Injection and Carbon Dioxide Sequestration (Wiley-Scrivener)

Stand-alone and Hybrid Wind Energy Systems: Technology, Energy Storage and Applications (Woodhead Publishing Series in Energy)

Green Chemistry Using Liquid and Supercritical Carbon Dioxide

Capturing king coal deploying carbon capture and storage systems in the U.S. at scale

Capturing King Coal Deploying Carbon Capture and Storage Systems in the U.S. at Scale

Energy Economics: CO2 Emissions in China

Recommend Documents