Industrial combustion teseting

INDUSTRIAL COMBUSTION TESTING © 2011 by Taylor and Francis Group, LLC INDUSTRIAL COMBUSTION TESTING Charles E. Baukal...

Author: Baukal C.E.

189 downloads 1678 Views 24MB 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

INDUSTRIAL COMBUSTION TESTING

© 2011 by Taylor and Francis Group, LLC

INDUSTRIAL COMBUSTION TESTING Charles E. Baukal, Jr.

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business


CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2011 by Taylor and Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number: 978-1-4200-8528-0 (Hardback) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Industrial combustion testing / edited by Charles E. Baukal, Jr. p. cm. “A CRC title.” Includes bibliographical references and index. ISBN 978-1-4200-8528-0 (alk. paper) 1. Furnaces--Testing. 2. Furnaces--Combustion. 3. Furnaces--Industrial applications. I. Baukal, Charles E. TH7140.I47 2010 621.402’3--dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com


2010014467

Contents Preface..........................................................................................................................................................................................ix Editor...........................................................................................................................................................................................xi Contributors............................................................................................................................................................................. xiii

Section I General 1. Introduction......................................................................................................................................................................... 3 Charles E. Baukal, Jr. 2. Testing Safety.................................................................................................................................................................... 41 Charles E. Baukal, Jr. 3. Experimental Design....................................................................................................................................................... 63 Joseph Colannino 4. Fluid Flow.......................................................................................................................................................................... 77 Wes Bussman and Joseph Colannino 5. Temperature....................................................................................................................................................................... 97 Charles E. Baukal, Jr. 6. Heat Flux............................................................................................................................................................................117 Charles E. Baukal, Jr. 7. Pollution Emissions........................................................................................................................................................141 Charles E. Baukal, Jr. 8. Combustion Noise.......................................................................................................................................................... 183 Mahmoud M. Fleifil, Carl-Christian Hantschk, and Edwin Schorer 9. Flame Impingement Measurements............................................................................................................................211 Charles E. Baukal, Jr. 10. Physical Modeling in Combustion Systems............................................................................................................. 241 Christopher Q. Jian 11. Virtual Testing................................................................................................................................................................ 251 Eddy Chui, Allan M. Runstedtler, and Adrian J. Majeski

Section I I Advanced Diagnostics 12. Laser Measurements...................................................................................................................................................... 269 Michele Marrocco and Guido Troiani 13. CARS Temperature Measurements in Flames in Industrial Burners................................................................. 289 Patrick M. Hughes, Thangam Parameswaran, and Richard J. Lacelle

v © 2011 by Taylor and Francis Group, LLC

vi

Contents

14. Diode Laser Temperature Measurements..................................................................................................................311 Thomas P. Jenkins and John L. Bergmans 15. Image-Based Techniques for the Monitoring of Flames........................................................................................ 337 Javier Ballester and Ricardo Hernández 16. High Temperature Cameras......................................................................................................................................... 355 William J. Lang 17. Liquid Fuel Atomization Testing................................................................................................................................ 369 Khaled A. Sallam

Section II I Burner Testing 18. Process Burners............................................................................................................................................................... 377 Jeffrey Lewallen, Thomas M. Korb, Jaime A. Erazo, Jr., and Erwin Platvoet 19. Commercial Boiler Burners.......................................................................................................................................... 395 Yaroslav Chudnovsky and Mikhail Gotovsky 20. Power Burners..................................................................................................................................................................411 Vit Kermes, Petr Beˇ lohradský, Petr Stehlík, and Pavel Skryja 21. Regenerative Combustion Using High Temperature Air Combustion Technology (HiTAC)........................ 429 Ashwani K. Gupta, Susumu Mochida, and Tsutomu Yasuda 22. Characterization of Ribbon Burners.......................................................................................................................... 449 Colleen Stroud Alexander and Melvyn C. Branch 23. Flameless Burners.......................................................................................................................................................... 471 Joachim G. Wünning and Ambrogio Milani 24. Radiant Tube Burners.................................................................................................................................................... 487 Michael Flamme, Ambrogio Milani, Joachim G. Wünning, Wlodzimierz Blasiak, Weihong Yang, Dariusz Szewczyk, Jun Sudo, and Susumu Mochida 25. Metallic Mat Gas Combustion..................................................................................................................................... 505 Giuseppe Toniato, Andrea Zambon, and Andrea D’Anna 26. Performance Prediction of Duct Burner Systems via Modeling and Testing.................................................... 517 Steve Londerville 27. Oxy-Fuel and Oxygen-Enhanced Burner Testing.................................................................................................... 529 Lawrence E. Bool, III, Nicolas Docquier, Chendhil Periasamy, and Lee J. Rosen

Section I V Flare Testing 28. Large-Scale Flare Testing.............................................................................................................................................. 553 Charles E. Baukal, Jr., Jianhui Hong, Roger Poe, and Robert Schwartz 29. Flare Experimental Modeling...................................................................................................................................... 571 Chendhil Periasamy and Subramanyam R. Gollahalli


Contents

vii

30. Flare Radiation................................................................................................................................................................ 595 Wes Bussman and Jianhui Hong

Section V Testing in Combustors 31. Cement Kilns....................................................................................................................................................................615 Eugen Dan Cristea and Givanni Cinti 32. Glass Furnaces................................................................................................................................................................. 671 R. Robert Hayes and Charles E. Baukal, Jr. 33. Thermal Oxidizer Testing............................................................................................................................................ 691 Bruce C. Johnson and Nate Petersen 34. Utility Boilers.................................................................................................................................................................. 705 Giuseppe Toniato and Silvio Rudi Stella Appendix A: F-Distribution (99%, 95%, & 90% Confidence)........................................................................................ 729 Appendix B: EPA Sample Methods.................................................................................................................................... 733 Appendix C: Common Conversions.................................................................................................................................. 735 Index......................................................................................................................................................................................... 737


Preface This book is intended to fill a gap in the literature for books on industrial combustion testing. It should be of interest to anyone working in or with the field of industrial combustion. This includes burner and furnace designers, researchers, end users, government regulators, and funding agencies. It can also serve as a reference work for those teaching and studying combustion. The book covers a wide range of testing techniques used in a broad array of applications in the metals, minerals, thermal oxidation, hydrocarbon/petrochemical, and power generation industries. There are 61 authors from 10 countries representing 33 prominent combustion organizations, and these authors have hundreds of years of combined experience with industrial combustion testing. The book contains 34 chapters divided into five sections. Section I is a general section with 11 chapters: Introduction, Testing Safety, Experimental Design, Fluid Flow, Temperature, Heat Flux, Pollution Emissions, Combustion Noise, Flame Impingement Measurements, Physical Modeling in Combustion Systems, and Virtual Testing. It is designed to provide some of the basic information referenced in succeeding chapters. Section II contains six chapters on advanced diagnostics: Laser Measurements, CARS Temperature Measurements in Flames in Industrial Burners, Diode Laser Temperature Measurements, Image-Based Techniques for the Moni toring of Flames, High Temperature Cameras, and Liquid Fuel Atomization Testing. Section III has ten chapters on burner testing: Process Burners, Commercial Boiler Burners, Power Burners, Regenerative Combustion Using High Temperature Air Combustion Technology (HiTAC), Characterization of Ribbon Burners, Flameless Burners, Radiant Tube Burners, Metallic Mat Gas Combustion, Performance Prediction of Duct Burner Systems Via Modeling and Testing, and Oxy-Fuel and Oxygen-Enhanced Burner Testing. Section IV has three chapters on flare testing: Large-Scale Flare Testing, Flare Experimental Modeling, and Flare Radiation. Section V has four chapters on testing in combustors: Cement Kilns, Glass Furnaces, Thermal Oxidizer Testing, and Utility Boilers. The purpose of this work is to compile testing techniques utilized in industrial combustion for use by practitioners. No such book currently exists, which means that those in this field must consult a range of sources such as journals, magazines, and conference proceedings to get this kind of information. This is generally impractical because those practicing in the field usually do not have the time or the resources to

extensively research this topic. While academics have access to the information, they generally do not work at the large scales associated with industrial combustion and therefore may not be familiar with how techniques are applied in production applications. This book is designed to help practitioners both in the field and in academics. Besides providing a single-source reference, this book also provides information for specific applications. This means that someone practicing in a particular area can immediately go to their application, without necessarily having to read through other chapters. They can determine for themselves what is useful for them. The reader can save time and more quickly use the information provided by experts in each area. Nearly 1300 references and over 800 figures, and 140 tables are provided for those that need further information on a particular topic. The book provides case studies and examples to show how to apply the information for particular applications. This includes identifying potential problems that could be very costly if not avoided. For example, failure to properly measure pollution emissions could lead to large fines from regulatory agencies. The book is designed to be more hands-on and less theoretical so the information can be easily applied to real situations in a variety of industries. This book tells the reader how to make measurements and conduct tests in industrial combustion systems including full-scale burners, furnaces, heaters, boilers, flares, and thermal oxidizers. There are some topics that are not covered and some that are not treated extensively. Since the majority of industrial applications use gaseous fuels, there is more treatment of that type of fuel, with less discussion of liquid and solid fuels. This book concerns atmospheric pressure combustion, which is the predominant type used in most industrial applications. There are some burner designs, combustors, and applications that are not considered. As with any book of this type, there are sure to be author preferences and biases, but the coverage is fairly extensive and comprehensive. There are also generous discussions of many common industrial applications to help the reader better understand the requirements for different types of tests. Particularly because of the increasing emphasis on the environment, most industrial tests include some type of pollution emission measurements. While industrial combustion testing is a dynamic area of continuing research, the principles considered here are expected to be applicable well into the foreseeable future. ix


Editor Charles E. Baukal, Jr., PhD, is the Director of the John Zink Institute for the John Zink Co., LLC (Tulsa, Oklahoma) where he has been since 1998. He has also been the Director of Research and Development and the Director of the Research and Development Test Center at Zink, which is a leading supplier of industrial combustion equipment to a variety of industries. Previously, Dr. Baukal worked for 13 years at Air Products and Chemicals, Inc. (Allentown, Pennsylvania) in the areas of oxygen-enhanced combustion and rapid gas quenching in the ferrous and nonferrous metals, minerals, and waste incineration industries. He worked for Marsden, Inc. (a burner supplier in Pennsauken, New Jersey) for five years in the paper, printing, and textile industries, and Selas Corp. (a burner supplier in Dresher, Pennsylvania) in the metals industry, both in the area of industrial combustion equipment. He has 30 years of experience in the fields of industrial combustion, pollution control, and heat transfer and has authored more than 100 publications in those areas. Dr. Baukal is an adjunct instructor for Oral Roberts University and the University of Tulsa, both in Tulsa, Oklahoma. He is the author or editor of seven books in the field

of industrial combustion including: Oxygen-Enhanced Combustion (1998), Heat Transfer in Industrial Combustion (2000), Computational Fluid Dynamics in Industrial Combustion (2001), The John Zink Combustion Handbook (2001), Industrial Combustion Pollution and Control (2004), Handbook of Industrial Burners (2004), and Heat Transfer from Flame Impingement Normal to a Plane Surface (2009). Dr. Baukal has a PhD in mechanical engineering from the University of Pennsylvania (Philadelphia, Pennsylvania) and is a licensed Professional Engineer in the state of Pennsylvania, a Board Certified Environ mental Engineer, and a Qualified Environmental Professional. He has served as an expert witness in the field of combustion, has 11 U.S. patents, and is a member of numerous honorary societies and Who’s Who compilations. He is a member of the American Society of Mechanical Engineers, the Air and Waste Management Association, the Combustion Institute, and the American Society for Engineering Education. He serves on several advisory boards, holds offices in the Air and Waste Management Association and the American Society for Engineering Education, and is a reviewer for combustion, heat transfer, environmental, and energy journals.

xi © 2011 by Taylor and Francis Group, LLC

Contributors Colleen Stroud Alexander, PhD, received her doctorate in mechanical engineering from the University of Colorado at Boulder. Her research focused on the heat transfer, fluid flow, and chemical kinetics involved in methane-air combustion flame treatment processes. Her PhD research efforts resulted in five technical publications in multiple peer-reviewed journals. She carried out her postdoctoral work as a guest researcher at the National Institute for Standards and Technology in Gaithersburg, Maryland, performing both experimental and numerical analysis in the study of controlled combustion reactions within reacting flows. She most recently worked as a research engineer studying the performance of advanced fuels in various combustion regimes at the National Renewable Energy Laboratory in Golden, Colorado. Prior to receiving her degree, Dr. Alexander also worked in the information technologies sector as a project manager and technical support engineer at CSG Systems (Englewood, Colorado). She also worked as a pneumatics engineer supporting the assembly of the Atlas Centaur Rocket at Lockheed Martin in Denver, Colorado. Javier Ballester, PhD, is currently a professor in fluid mechanics at the University of Zaragoza (Spain), where he has been since 1997. He received his degree in electrical engineering from the University of Zaragoza in 1992. Previously, he was hired as a researcher at the Laboratory of Research on Combustion Technologies by the Technological Institute of Aragon (1991–1992) and by the Spanish Council of Scientific Research (1992–1997). His areas of expertise are fluid mechanics and combustion, and his current research interests include the combustion of solid fuels, advanced monitoring and control of industrial flames, and combustion instabilities. He has three patents and has authored over 90 papers in international journals and conferences. He has participated, in most cases as principal investigator, in more than 90 research projects, including contracts with private companies and projects funded by the Spanish and European administrations. Petr Beˇlohradský, MS, is currently a postgraduate student and he works as a technician at the Institute of Process and Environmental Engineering at Brno University of Technology (Czech Republic). He holds his degree in mathematical engineering from Brno University of Technology. His work is directed toward the research of combustion on gaseous fuels with special focus on modeling by using statistical methods and computational fluid dynamic methods. He is an author or co-author of several papers related to combustion modeling presented at international conferences. John L. Bergmans, MEng, is the principal engineer and owner of Bergmans Mechatronics LLC (Newport Beach, California). Bergmans received his degree in mechanical engineering from Carleton University (Ottawa, Canada) in 1995. He founded Bergmans Mechatronics in 2003 and has since developed data acquisition and control systems for several rocket motor test stands and an oxyfuel combustion system. Bergmans is also active in the development and testing of tunable-diode, laser-based instrumentation for large-scale combustion applications. Prior to founding BML, Bergmans was employed for eight years by CFD Research Corp. (Huntsville, Alabama), where he developed closed-loop pressure controllers for solid-propellant rocket and air-breathing propulsion systems. Wlodzimierz Blasiak, PhD, is head and professor in the Division of Energy and Furnace Technology, Royal Institute of Technology, Sweden. He has his degree of applied thermodynamics from Technical University of Czestochowa (Poland). He has carried out research on heat and mass transfer processes in boilers and furnaces and published around 200 papers since 1993. For the last ten years the main research themes of his work are high performance industrial furnaces, high temperature air combustion-HiTAC/flameless combustion for gaseous and solid fuel, high temperature air/steam gasification of biomass and wastes-HTAG, oxyfuel, and flameless oxyfuel combustion. He also carried out and managed many research projects financed by Swedish and international agencies in cooperation with European and Japanese industry. He has four patents (three of them are PCT, and two of them are U.S. provisional-pending) on solid fuel thermal conversions. Lawrence E. Bool III, PhD, is a senior development associate in the combustion research and development group for Praxair, Inc. (Tonawanda, New York) where he has been since 1997. Dr. Bool received his doctorate in chemical engineering from the University of Arizona in 1993. His work focuses on using basic science to develop new xiii © 2011 by Taylor and Francis Group, LLC

xiv

Contributors

oxyfuel applications for industry. Recent examples include a novel process to reduce pollutants from power plants and a process to produce activated carbon. Dr. Bool holds 20 patents and has authored several peer-reviewed publications. Melvyn C. Branch, MS, PhD, is the Joseph Negler Professor of Mechanical Engineering, Emeritus at the University of Colorado at Boulder. He received his degrees in mechanical engineering from the University of California at Berkeley. He has previously served as associate dean of Engineering for Research and Administration, associate dean of the graduate school, and Director of the Center for Combustion Research. He has taught graduate and undergraduate courses on combustion fundamentals, fluid mechanics, heat transfer, applied thermodynamics, and fuel technology. His research activity in these areas includes experimental and theoretical studies of combustion-generated air pollutants, fuel efficiency, flame processing, metal burning, and aircraft and rocket combustion. His recent consulting activity includes the 3M Company, the Combustion Research Division of Sandia National Laboratories, the Air Pollution Control Division of the State of Colorado and the U.S. Federal Trade Commission. Dr. Branch has served as a member and Chair of the Colorado Air Quality Control Commission, the state agency responsible for promulgating state regulations relating to air quality, and as a member of the Research Committee of the Health Effects Institute. He is a member of the Combustion Institute, Tau Beta Pi, Pi Tau Sigma, and a Fellow of the American Society of Mechanical Engineers. He is a past chairman of the Western States Section‑Combustion Institute. He has been honored with the Society of Automotive Engineers Ralph Teetor Award for engineering educators and the University of Colorado Teacher Recognition Award for outstanding teacher during the year. His research awards include the American Society of Mechanical Engineers Gustus L. Larson Award, the Fulbright Fellowship, the University of Colorado Faculty Fellowship, and the Associated Western Universities Faculty Fellowship. He has authored over 90 technical articles and supervised 15 students to completion of their PhD. Wes Bussman, PhD, is a senior research and development engineer for the John Zink Co., LLC (Tulsa, Oklahoma) where he has been since 1981. He received his degree in mechanical engineering from the University of Tulsa (Tulsa, Oklahoma). Dr. Bussman has 19 years of basic scientific research work, industrial technology research and development, and combustion design engineering. He holds ten patents, has authored several published articles and conference papers, and has been a contributing author to several combustion-related books. He has taught engineering courses at several universities and is a member of Kappa Mu Epsilon Mathematical Society and Sigma Xi Research Society. Yaroslav Chudnovsky, MS, PhD, is a senior staff member of research and development at the Gas Technology Institute (Des Plaines, Illinois). Dr. Chudnovsky received his degrees in 1982 and 1990, respectively, from Bauman Technical University (Moscow, Russia). He conducts research and development of advanced, low-emissions, highefficiency, and high heat transfer combustion systems and technologies for industrial applications. He has over 25 years of combined basic and applied research and development experience in engineering, design, and laboratory/ field evaluation of advanced energy exchange and combustion systems and technologies. Prior to joining the Gas Technology Institute in 1995, he worked as a head of the research laboratory at Power Machinery Research Institute (Moscow, Russia) where he developed solutions for energy, space, and military applications. His areas of interest include: heat transfer enhancement and waste heat recovery, convective heat transfer and heat exchangers, advanced combustion and environmental technologies, and smart thermal management. He has over 100 publications and six patents. He is the editor of the Heat Exchanger Design Handbook and the Journal of Enhanced Heat Transfer. Eddy Chui, PhD, is currently a senior research scientist with CanmetENERGY, the clean energy research and technology development centre of Natural Resources Canada of the Canadian Government (Ottawa, Canada). Dr. Chui received his degree in mechanical engineering from the University of Waterloo (Ontario, Canada) in 1990. Prior to joining CanmetENERGY in 1993, he had worked for Bechtel Canada in project engineering, University of Alberta in acoustic research, and Advanced Scientific Computing Ltd. (presently ANSYS Canada) in numerical modeling. At CanmetENERGY, he is responsible for directing and conducting research on various aspects of combustion modeling technology and utilizing the model to assist industries in practical applications. Past achievements include the development of new modeling strategies to predict NOx formation in coal flames and natural gas flames, determination of the sensitivity of combustion performance to coal blending, design of a new generation of furnace model for process simulation, evaluation of combustion performance in appliances using biomass, development of strategies for industrial processes to convert to a lower-carbon fuel, and the successful implementation of the model to resolve combustion-related problems in full-scale units like utility boilers, coke ovens, refinery furnaces, blast


Contributors

xv

furnaces, and industrial furnaces for metal processing. Current research efforts are focused on developing clean coal technologies: oxy-coal combustion and coal gasification for CO2 capture, modeling CO2 storage in subsurface environments, new computational fluid dynamic tools for nonexpert users, and assisting the power sector in China and Canada to burn coals more cleanly and efficiently through the use of simulation. Also, a new capability of microscale modeling has been developed under his supervision, presently being implemented on investigating solid oxide fuel cells. He has authored and co-authored over 150 publications in international journals, conference proceedings, industrial reports, and government departmental reports. Giovanni Cinti is the technology department manager for the Technical Centre of Italcementi Group (CTG), located in Bergamo (Italy). He received his certificate degree at Politecnico di Milano in 1973 as a chemical engineer. In 1975 he started his professional activity in Italcementi SpA in the central headquarters as a member of the combustion department, dealing with all the aspects of cement kiln burners and related combustion pollutants. He is a member of the International Flame Research Foundation, holding the chairmanship of the Italian Association for four years (2004–2008). He has represented the company in the Associazione Tecnico-Economica del Cemento (AITEC) and in the European Cement Association (Cembureau) and was the cement expert of the Italian delegation in the meeting for the definition of Best Available Technologies for Cement Manufacturing in 2001 and 2007. Joseph Colannino, BS, MS, is director of engineering for John Zink Co., LLC, where he has worked for the last 12 years. He received his degree in chemical engineering with minors in materials and chemistry from the California Polytechnic University at Pomona and a degree in knowledge management with emphasis in organizational dynamics from the University of Oklahoma. He is a registered professional engineer in the state of California. He has been engaged in combustion research for more than 20 years and has authored many papers and presentations. His book, Modeling of Combustion Systems: A Practical Approach (Taylor & Francis), was published in 2006. Besides the John Zink Handbook, Joseph has also contributed book chapters in other volumes including the Industrial Combustion Handbook (CRC Press), and the Air and Pollution Control Equipment Selection Guide, (Lewis). He is an adjunct faculty member at Tulsa University and Oral Roberts University (both in Tulsa, Oklahoma) teaching combustion and engineering courses. Colannino is listed in several Who’s Who compilations. Eugen Dan Cristea, PhD, MEng, has worked since 1987 in the cement and lime industry in the positions of technical director of Cimprogetti SpA, an engineering company located in Bergamo (Italy) and today as a process function manager of Italcementi Group in Bergamo. He received his doctorate degree in thermal sciences from the Politehnica University of Timisoara (Romania) and his engineering degree in power generation engineering from Politehnica University of Bucharest (Romania). He did postdoctoral work as a visiting adjunct assistant professor at Montana State University (Bozeman, Montana) performing numerical simulation combustion for MHD combustor fired on natural gas. He has served as head of Combustion and MHD laboratory of Scientific Research Division (formerly the Power Institute of Romania Academy) of the Institute of Scientific Research and Technological Engineering for Power Equipment in Bucharest. He conducted some fundamental and mainly applied research works in all areas of thermal sciences including combustion science and combustion engineering, heat and mass transfer, fluid mechanics, thermodynamics and chemical thermodynamics, and direct energy conversion, with particular emphasis on experimental as well as computational approaches. He is a member of the American Society of Mechanical Engineers, of the International Flame Research Foundation at Pisa (Italy), and has served on the Italian Flame Research Committee. Dr. Cristea has authored and co-authored two combustion-related books, over 20 journal articles, over 20 conference papers and holds four patents for novel burner development. He has delivered seminar lectures at the International Flame Research Foundation. Andrea D’Anna, PhD, is an associate professor of chemical engineering at Università “Federico II” di Napoli (Napoli, Italy) where he has been since 2001. He has a degree in chemical engineering from Università “Federico II” di Napoli (Napoli, Italy). He was a researcher at Istituto Ricerche Combustione, CNR and at Fertimont, Montedison SpA. His research interests include combustion chemistry, chemical kinetics, combustion-formed particles and their effects on health and climate, nano-material synthesis, characterization and modeling, transport properties of nano-materials, and filtration procedures. He is the author of over 100 technical publications. Nicolas Docquier, PhD, is the general manager of ACI (Atlanta, Georgia), a division of Air Liquide Advanced Technologies US, specializing in combustion equipment for steel, nonferrous, and glass industries. He has a doctoral degree in energy sciences from the Ecole Centrale de Paris (France), a degree in fluid mechanics from the von Karman


xvi

Contributors

Institute (Belgium), and an engineering degree from Université de Liège (Belgium). He has been a combustion specialist with Air Liquide since 2003, in France and in the United States. He also worked at IFP Powertrain Engineering (Paris, France) and for Rolls-Royce Industrial and Marine Gas Turbines (Coventry, United Kingdom). He has a strong industrial and research background in combustion, industrial heating and melting processes, fluid mechanics, heat transfer, fuels and emissions, and has developed several combustion test platforms. His experience includes oxyburners, furnaces and melting processes, safety systems and practices, emission measurements and sensors, optical diagnostics, internal combustion engines, and turbomachinery. He is the author or co-author of over 20 publications on these topics, has nine patents for novel burner and sensor development and has taught graduate courses on fluid mechanics. Jaime A. Erazo, Jr., MS, is a design/test engineer at the John Zink Co., LLC, Tulsa, Oklahoma. He has worked for the John Zink Company Process Burners group for one year. He graduated with a degree in mechanical engineering from the University of Oklahoma in 2008. He authored five combustion related technical publications and presentations during his time at Oklahoma University. Michael Flamme, PhD, is internationally known for his work on gas-fired technology over a period of more than 20 years with Gaswärme-Institut (Essen, Germany). In 1989 he received his degree from Bochum University, Germany for his work focused on high temperature processes using high preheated combustion air. He has particular expertise and knowledge of high temperature industrial processes, combustion technologies for gas turbines and boilers, and waste and biomass conversion to energy. He authored over 110 publications in national and international journals and conference proceedings. His scientific achievements were rewarded by the Wilhelm Jost Medal of the German Section of the Pittsburgh Combustion Institute in 1993. He currently manages his own independent energy consultancy (FlammeConsulting) in Essen, Germany. Mahmoud M. Fleifil, BS, MS, PhD, is a senior thermoacoustic and vibration engineer in the research and development department of John Zink Co., LLC (Tulsa, Oklahoma). He has been with the company since 1999. Dr. Fleifil graduated from Ain Shams University, Cairo, Egypt with his degrees in mechanical engineering, and his doctoral degree in mechanical engineering from a co-supervisory program between Ain Shams University and MIT. His areas of expertise are fluid dynamics, combustion instabilities, and noise control. He published eight journal articles and over 20 conference papers. He has over 13 years of experience in advanced techniques of acoustically driven combustion instability and noise control. He is a member of ASME and AIAA. He is an honored member of several Who’s Who compilations. Subramanyam R. Gollahalli, MASc, PhD, is a professor and holds the Lesch Centennial Chair in the School of Aerospace and Mechanical Engineering at the University of Oklahoma (Norman, Oklahoma) where he has been since 1976. He received his Master’s degree in 1970 and his doctoral degree in 1973 both in mechanical engineering from the University of Waterloo (Waterloo, Canada). He has held the positions of assistant professor, associate professor, professor, and Lesch Centennial professor, and academic director. He also worked as a research assistant at the University of Waterloo (Waterloo, Canada) and as a lecturer at the Indian Institute of Science (Bangalore, India). He has developed and taught courses in the combustion and energy areas at both undergraduate and graduate levels. He investigated the combustion of emulsified fuels and synthetic fuels, the publications based on his research, which are cited frequently. He served on the following technical committees: Combustion and Fuels Committee, Gas Turbine Division, ASME; Terrestrial Energy Committee, AIAA; Propellants and Combustion Committee-AIAA; Fuels and Combustion Technologies Committee, ASME; Emerging Energy Technologies Committee, ASME; and the Technical Program Committee, Combustion Institute. He has been recognized with the following awards: Regents Award for Superior Teaching at the University of Oklahoma, Energy Systems Award-AIAA, George Westinghouse Gold Medal-ASME, AIAA Sustained Service Award Robert Angus Medal, Engineering Institute of Canada, and Three Best Paper Awards, ASME Fellow, AIAA Associate Fellow. He was the associate editor of the Journal of Energy Resources Technology (1994–2000) and the associate editor of the Journal of Engineering for Gas Turbines and Power (2000–2006). Mikhail Gotovsky, MS, PhD, ScD, is a leading researcher at NPO CKTI (1963 to present) and a professor at GTURP (State University of Plant Polymer Technologies). He received his first degree in 1963 from St-Petersburg State Polytechnic University (Russia) and his other two degrees in 1970 and 2000, respectively, from Central Boiler-Turbine Institute (Russia). After graduating from St-Petersburg State Polytechnic University, Dr. Gotovsky


Contributors

xvii

joined the team of highly experienced and motivated professionals at NPO CKTI and grew from a junior engineer to the leading research and development professional in the area of heat and mass transfer for industrial applications. His areas of interest include: a variety of heat transfer problems (liquid metals, forced convection, high heat flux boiling), two-phase flow hydrodynamics and heat transfer, heat transfer enhancement, and thermal problems of nuclear waste transportation and storage. He has over 120 publications and 12 patents. Ashwani K. Gupta, PhD, DSc, is a distinguished university professor at the University of Maryland (College Park, Maryland) where he has been a professor of mechanical engineering. Prior to this he was a member of the research staff at MIT, and an independent research worker at the University of Sheffield, United Kingdom. His main research interests have been in the fields of combustion, air pollution, propulsion, high temperature air combustion, swirl flows, diagnostics, fuel sprays, fuel reforming, sensors, microscale combustion, and wastes to clean energy conversion. He has co-authored three books on swirl flows, and flowfield modeling and diagnostics, and high temperature air combustion: from energy conservation to pollution reduction. In addition he has authored nine book chapters and published over 450 archival papers in journals, refereed symposia, and conference proceedings. His honors and awards include: AIAA Energy Systems Award, AIAA Propellants and Combustion Award, ASME George Westinghouse Gold Medal, ASME James Harry Potter Gold Medal Award, ASME James N. Landis Medal Award, ASME Worcester Reed Warner Medal Award. Dr. Gupta received the University of Maryland President Kirwan Research Award and College of Engineering Research Award. He received eight Best Paper Awards from ASME and AIAA for his research contributions. He is the founding co-editor of the Energy Engineering and Environment Series published by CRC Publishers. He is an associate editor of the Journal of Propulsion and Power, Journal of Applied Energy, International Journal of Sprays and Combustion Dynamics, and International Journal of Reacting Systems. He has served as chair of AIAA Terrestrial Energy Systems Technical Committee, chair of Propellants and Combustion Technical Committee, deputy director of Energy Group, and director of Propulsion and Energy Group. At ASME he served as chair for Fuels and Combustion Technology Division, and Computers in Engineering Division. He is cited in Who’s Who in America, Engineering, Technology, American Education, and Aviation in the United States, and The Men of Achievement in the United Kingdom. Carl-Christian Hantschk, PhD, has been working as a consulting engineer in industrial acoustics for MüllerBBM GmbH (Munich, Germany) since 2001. He was promoted to managing director in 2009. He works on industrial acoustics in general, including theoretical and applied acoustics, environmental acoustics, aero-acoustics and numerical acoustics, with special focus on the interdisciplinary field between combustion and acoustics. He holds a diploma in mechanical engineering and received his doctorate in thermodynamics from the Technical University Munich, Germany. His research focused on combustion-driven acoustic oscillations in burners and combustionacoustic interactions. He gave lectures on chemical thermodynamics, thermal radiation, and heat transfer and acoustics at his university, international conferences, and for industrial clients. His work resulted in 30 publications and four invention disclosures. As one of his main research projects, he codeveloped an active acoustic feedback control for industrial combustion systems. R. Robert Hayes, MS, is the vice president of El Dorado Engineering, Inc. (Salt Lake City, Utah) where he has been since 2006. He also worked at the National Renewable Energy Laboratory and the John Zink Co., LLC. He received his degree in mechanical engineering from Brigham Young University. He has a strong technical background in combustion, heat transfer, fuels, and emissions formation/reduction, with experience in both industrial and research facilities. He led a world-class alternative fuels and advanced vehicle technology research facility at the National Renewable Energy Laboratory. His experience includes internal combustion systems, alternative fuels, furnaces, burners, air/exhaust handling systems, pollution control systems, emissions measurements, instrumentation, safety systems, and combustion of energetic materials (propellants and explosives). He is the author or co-author of over 20 publications on these topics. He has three patents for novel burner development and has taught professional courses on burners, formation and control of combustion emissions, heat transfer, and fluid mechanics. Ricardo Hernández, graduated in 2003 with a degree in physics at the University of Zaragoza (Zaragoza, Spain), is currently a PhD student there and works in the Laboratory of Research on Combustion Technologies. He works on research of advanced monitoring and control of industrial flames and combustion instabilities, has participated in some research projects and is co-author of several papers in international journals and conferences.


xviii

Contributors

Jianhui Hong, BS, PhD, is a flare process engineer at John Zink Co., LLC (Tulsa, Oklahoma). He is the principal investigator and lead inventor of the ultra-stable WindProof™ pilot, the triple-redundancy InstaFireTM flare pilot, and the ultra-efficient, steam-assisted Steamizer-XP™ flare. He received a degree from Tsinghua University, Beijing, China and his doctorate from Brigham Young University, Provo, Utah, both in chemical engineering. He has other U.S. and foreign patents including low NOx incinerator apparatus and control method and air-assisted flare. He also worked as a research and development engineer at John Zink. His other areas of expertise include ground flare design and optimization, kinetic simulation involving NOx, SOx, and soot; global optimization of steel stack structure considering structural and process constraints; phased array of thermal radiometers for measuring the flame epicenter and radiant fraction of industrial flares; flare smoke control method; flare control for over-steaming/over-aeration prevention. He has authored and co-authored over 15 journal articles and book chapters. His personal interests include heli-plane design, aircraft emergency landing system, and personal aerial vehicle design. Patrick M. Hughes, MSc, is the group leader for measurement systems and combustion kinetics at Canmet ENERGY Ottawa (Ottawa, Canada). He has been a research scientist with CanmetENERGY since 1982. Before that he was a defense scientist for six years with National Defence Canada. He has his degree in mechanical engineering from the University of Waterloo (Waterloo, Canada). Throughout his career he has developed advanced measurement techniques to study combusting flow fields. His research with Natural Resources Canada has involved the use of laser-based and other optical techniques to study industrial burner technology. He is also involved in the development of techniques to characterize coal combustion kinetics and deposition in power boilers. His publications cover optical measurement techniques and their application to industrial burners, data packages for evaluation of computational fluid dynamic models, advanced characterization techniques for coal combustion and deposition, and rocket motor instabilities. He is currently the editor-in-chief of the Combustion Journal of the International Flame Research Foundation. Thomas P. Jenkins, PhD, is a senior scientist at MetroLaser, Inc., where he has worked since 2000. He received his degree in mechanical engineering from the University of California at Davis. He has been principal investigator on eight research programs for the DoE, Air Force, NASA, Army, and private industry to develop laser-based diagnostics for studying combustion and fluid mechanics. He has demonstrated several first-of-a-kind measurements, including quantitative nonintrusive measurements of soot concentration in an aircraft engine exhaust, nonintrusive temperature and H2O concentration in an industrial glass furnace, and a large area flow velocimetry system for studying parachutes. He is a member of the American Institute of Aeronautics and Astronautics (AIAA), and is an active member of the AIAA Aerodynamic Measurement Technologies committee. He has been the primary author on more than 30 journal articles and conference papers. Prior to coming to MetroLaser, Dr. Jenkins worked for three years as a research associate at Stanford University, where he developed soot diagnostics for advanced propulsion systems. Christopher Q. Jian, PhD, is the director of the Simulation Technology Solutions Group (STS) at John Zink Co., LLC (Tulsa, Oklahoma) where he has been since 2004. He received his degree in mechanical engineering from the University of Maryland at College Park (College Park, Maryland). Prior to joining John Zink, he was the business research manager at Owens Corning responsible for asset utilization and customer profitability analyses and mergers and acquisitions. He was the research and development manager at Vortec Corporation before he joined Owens Corning in 1995. He served as past chair of the production efficiency subcommittee of the Glass Manufacturing Industry Council and a member of its executive advisory committee. Dr. Jian’s research areas include fossil fuel combustion, glass melting and delivering, computational fluid dynamics and physical modeling, as well as low level radioactive material vitrification. He holds four U.S. patents and has authored and/or co-authored over 80 technical publications. Bruce C. Johnson, MSc., PE, is the technology development leader for the Thermal Oxidation Systems Group at the John Zink Co., LLC (Tulsa, Oklahoma). He received his degree in chemical engineering from the University of North Dakota under a Bureau of Mines research fellowship. He has spent much of his career in research and development and has been employed by the Calgon Corporation, Department of Energy, Combustion Engineering Co., and several thermal oxidation companies. His qualifications include process design, new product development, testing, and project management of governmental and corporate research and development groups. He has five patents and has authored numerous technical reports and papers.


Contributors

xix

Vit Kermes, PhD, currently works as a lecturer at Brno University of Technology (Czech Republic). He holds his degree in process engineering from Brno University. His work is directed at applied and industrial research of reduction of NOx emissions in combustion of gaseous fuels and industrial research of nonstandard liquid combustion such as liquid wastes and renewable liquid fuels. He is an author and co-author of about 20 papers related to combustion modeling presented at international journals and international conferences. Thomas M. Korb, PhD, PE, is a technical leader in the Process Burner Group at the John Zink Co., LLC (Tulsa, Oklahoma). He received his degree in mechanical engineering from Arizona State University (Tempe, Arizona). Dr. Korb has 15 years of experience in combustion and thermal sciences. His work has included design and testing of combustion equipment for the refining and petrochemical industries as well as failure analysis engineering of accidental fires and explosions. He has also worked in the development of both gas turbine and diesel engines. He is a registered professional engineer and is a member of Tau Beta Pi National Engineering Honor Society, American Society of Mechanical Engineers, Society of Automotive Engineering and the Experimental Aircraft Association. His research focuses on fundamental ignition mechanisms with a particular emphasis on hot surface ignition of hydrocarbon fuels and the impact of hot surface material and surface oxide structure. He is a recipient of the Darryl E. Metzger Scholarship and Dean’s Graduate Scholars Award at Arizona State University. Richard J. Lacelle, CET, LSO, is an electro-optics technologist working at CanmetENERGY, Natural Resources CANADA (Ottawa, Ontario). He is a graduate of Algonquin College of Applied Arts and Technologies (Ottawa, Ontario, Canada) in the field of electronics engineering technology and electronics engineering techniques. His initial research project was the commissioning of the coherent anti-Stokes Raman spectroscopy (CARS) system for the measurement of high temperature combustion flames in industrial burners. He is also the laser safety officer who oversees all laser operations at the CanmetENERGY Bells Corners Complex located west of Ottawa. Lacelle has been a member of the CARS team at CanmetENERGY since its inception. He is responsible for the development, assembly, bench testing, operation, and data acquisition of the CARS system. For the past 26 years he has been working on research projects like very high speed electronic triggering circuits for the various laser applications, Schlieren photography, high speed camera imaging, laser sheet visualization, infrared imaging and analysis, laser doppler velocimetry, Canmet flame identification control system, tuneable diode laser absorption spectroscopy, and laser induced breakdown spectroscopy. He has authored over 20 technical documents, as laser safety manuals based on ANSI Z136.1 standards for the safe use of lasers, operator’s manual, and standard operating procedures for the above technologies. William J. Lang, BS, is vice president and co-owners of Lenox Instrument Co., Inc. (Trevose, Pennsylvania), manufacturer of the FireSight high-temperature video camera system, along with a full line of other remote visual inspection equipment. A graduate of La Salle College, he began his career in the Lenox shop fabricating high-temperature lenses and optics, and he later pioneered the use of the portable FireSight camera system that is in wide use in fossil-fuel power plants. He has 42 years of application engineering experience visual inspection and process monitoring. He has written articles that have appeared in a variety of technical publications. One such article, “Furnace Cameras Assist in NOx Reduction” appeared in Power Engineering magazine in November 2002. His extensive background in applications include: installing a furnace camera in the combustion chamber of an operating gas turbine, monitoring nuclear waste encapsulation in glass, chemical and biological warfare incineration, and thousands of boiler and furnace installations around the world. He has experience with all fossil fuels. Jeffery Lewallen, BSME, PE, is the applications sales manager of the burner division of Callidus Technologies by Honeywell (Tulsa, Oklahoma). He is a University of Tulsa graduate with a degree in mechanical engineering and a professional engineer licensed in Oklahoma. He has over 17 years of combustion related experience including design engineering, production testing, technical field support, sales, and project management for global projects in the refining and petrochemical industry. He is a contributing author in the books Industrial Burners Handbook and the John Zink Combustion Handbook. Steve Londerville, BSME, is currently director of design engineering at Coen Company (Foster City, California). He received his degree from San Jose State University (San Jose, California) in 1977. Previous positions, since 1978, at Coen were chief technical officer research and development, vice president research and development, director research and development, and chief engineer. During the last 31 years he has been involved with all aspects of product


xx

Contributors

development at Coen. He holds seven patents and has authored 15 publications. He is a member of ASME, ACHIE, Combustion Institute, Tau Beta PI, and past officer and board of directors for Institute for Liquid Atomization and Spray Systems. He was recognized as Engineer of the Year by ASME, Santa Clara Valley section. He also received a Best Paper Award from ASME. Adrian J. Majeski, MSc, is a research engineer at Natural Resources Canada’s CanmetENERGY research center in Ottawa, Canada. He received his degree from the University of Alberta where he participated in the Flare Research Project. Since joining in 2001, he has worked on computational fluid dynamic models of both pilot- and industrialscale combustion systems including utility boilers and equipment related to metal processing and petroleum refining. His current research includes model development for clean-coal technologies, such as gasification and oxyfuel combustion. Before joining CanmetENERGY, Majeski worked on low-swirl burner technology at Lawrence Berkeley National Laboratory. Michele Marrocco, PhD, is a researcher in laser spectroscopy at ENEA (Rome, Italy) (1999 to present). He received his degree in physics from the University of Rome in 1994. He was employed as a postdoctorate at the Max-Planck Institute for Quantum Optics (Munich, Germany), as a researcher at the Quantum Optics Labs at the University of Rome (Rome, Italy), and as an optics researcher by the army. His research activities include: traditional and innovative spectroscopic techniques for diagnosis of combustion and nanoscopic systems studied by means of optical microscopy. The techniques used include: adsorption, laser induced fluorescence, spontaneous Raman, stimulated Raman gain, stimulated Raman loss, coherent anti-Stokes Raman, degenerate four wave mixing, polarization spectroscopy, laser induced breakdown, laser induced incandescence, laser induced thermal gratings. He has over 30 technical publications. Ambrogio Milani, DrIng, is a consultant for WS GmbH (Germany). He received a degree from the Politecnico di Milano (Milan, Italy) in 1965. He has 40 years of experience in combustion technology in energy intensive industrial sectors (steel and power generation). He was the head of the former CSM Experimental Station on combustion devoted to research and development for products and industrial processes. He works with the International Flame Research Foundation and the Combustion Institute. He is the manager of ECSC-funded research and development projects and of educational/training programs and courses. His interests include: combustion research (mainly iron and steel making), burner development, heat recovery, high efficiency, low emissions, and flameless oxidation. He has a number of technical publications including co-authoring the Handbook of Burner Technology for Industrial Furnaces. Susumu Mochida is the director and general manager of Technology & Engineering Division in Nippon Furnace Co., Ltd. (Yokohama, Japan) where he has been since 1982. He has participated in a number of projects and has been an active member of the high temperature air combustion (HiTAC) project team. The HiTAC technology has been widely adapted in industrial furnaces to save energy, reduce size of the equipment, and reduce pollution emission. He serves as chairman on the Japanese Flame Research Committee and a member of the Executive Committee of International Flame Research Foundation from 2006. He has authored over 25 technical publications, contributed to 20 patents, and made several presentations at meetings and conferences. He has been honored with AIAA Best Paper Award in 1999 and ASME George Westinghouse Silver Medal Award in 2001. Thangam Parameswaran, PhD, is a research scientist at CanmetENERGY, Natural Resources Canada (Ottawa, Canada). She obtained her degree from Northwestern University (Evanston, Illinois). Her early research was focused on the theoretical aspects of the optical properties of organic and transition metal complexes. During later years her research activities at Carleton University (Ottawa, Canada) involved theoretical and experimental aspects of laser Raman spectroscopy of transition metal compounds. Subsequently she worked for the National Research Council of Canada toward the development of coherent anti-Stokes Raman spectroscopy (CARS) for combustion diagnostics. Dr. Parameswaran has been a member of the CARS teams at NRC, Canada and CanmetENERGY for many years. During this period she was responsible for developing and applying theoretical calculations and analysis methods for retrieving information from CARS spectra. For the past 11 years she has been working as a research scientist at CanmetENERGY, Natural Resources Canada. She has also developed methods for applying flame emission spectroscopy for flame performance monitoring in industrial burners. Recently this approach was tested in an industrial boiler and has the potential to be implemented in a flame advisory system. Other optical methods she has initiated at CanmetENERGY are tunable diode laser absorption spectroscopy for stack gas measurements in pilot-scale facilities


Contributors

xxi

and laser induced breakdown spectroscopy for trace metal detection in combustion emissions. She has authored over 60 technical documents, as journal publications, conference proceedings, presentations, and contract reports in the fields of optical spectroscopy and combustion applications of spectroscopic methods. In 1993 she received a Joint Staff Performance Award as a member of the Advanced Combustion Diagnostics Technique (Coherent antiStokes Raman Spectroscopy) team, of the Institute for Chemical Processes and Environmental Technology, National Research Council of Canada. Chendhil Periasamy, PhD, BS, MS, a research scientist at Air Liquide Delaware Research and Technology Center (Newark, Delaware) since 2007. He received his doctorate in 2007 in mechanical engineering from the University of Oklahoma with Professor S. R. Gollahalli. He has degrees in mechanical engineering from the Indian Institute of Technology Madras (India) and Anna University (India). He specializes in developing and testing cleaner and energy-efficient oxy-combustion burners for glass, nonferrous, and steel industry applications. He has developed test platforms for evaluating oxy-burner performance and conducted several customer field trials. His research interests include oxy-combustion, industrial furnaces, energy systems, burner testing, combustion diagnostics, combustion in porous media, and oxygen safety. He is the author or co-author of over 25 peer-reviewed journal and conference publications in combustion and energy related topics. He received the Outstanding Graduate Student Award in 2007 for his porous media combustion research and undergraduate teaching activities. Nate Petersen, PE, BS, MS, is currently a process engineer at John Zink Co., LLC (Tulsa, Oklahoma) where he has been since 2005. He has degrees in chemistry and chemical engineering along with a degree in chemical engineering from the University of Utah (Salt Lake City, Utah). He has served in various engineering roles in the process burner group, flare group, and thermal oxidizer group consisting of process and mechanical design and equipment testing. He is a licensed professional engineer in the state of Oklahoma. Erwin Platvoet, MSc, is the director of process burner engineering at John Zink Co., LLC (Tulsa, Oklahoma) where he has been since 2009. He has a degree in chemical engineering from Twente University of Technology (Enschede, The Netherlands). He was a cracking furnace specialist at Total Petrochemicals (Feluy, Belgium) from 2004 to 2009. He was at ABB companies in The Netherlands, USA, and Switzerland from 1993 to 2009 and held a variety of positions including thermal engineer, principal development engineer, and research and development engineer at a variety of locations around the world. He worked at NRF Thermal Engineering (Uden, The Netherlands) from 1991 to 1993. He has authored a number of technical publications and has eight patents. Roger L. Poe, BS, is a research associate at John Zink Co., LLC (Tulsa, Oklahoma) where he has worked since 1999. He received a degree in mechanical engineering from Fairmont State University (Fairmont, West Virginia). Previously, he served as manager of the Callidus Technologies (Tulsa, Oklahoma) Test and Research Center from 1995 to 1999 where he was responsible for the design and testing of specialty burners, as well as the development of new burner equipment for the refinery and petrochemical industry. From 1989 to 1995 he managed the facilities and personnel for the Penn State University Energy and Fuels Research Center (State College, Pennsylvania). He served as a working manager and researcher for the Donlee Technologies Research and Development Group (York, Pennsylvania) from 1985 to 1989. During his career he has been involved with low NOx boiler burner technologies as they relate to both liquid and gaseous fuels, coal gasification in pilot-scale, fluidized bed reactors, and the development and testing of fluidized bed combustion units while working with Donlee Technologies. Further work, when on staff at Penn State University, was done with fluidized bed combustion, coal gasification, micronized and pulverized coal applications, coal slurry formulation and combustion, as well as low NOx gas and oil development. His most recent work has been concentrated in low NOx process type burners and large-scale flaring equipment while working with both the John Zink Co. and Callidus Technologies. He has published more than 24 articles and holds numerous patents relating to burners, flares, and pilots. His areas of interest are centered on fluid mechanics, combustion, thermodynamics, combustion testing, and manufacturing. Over the course of his career he has jointly worked with the Department of Energy, the Department of Defense, NORAD, the Institute of Gas Technology, the Gas Research Institute, Sandia National Labs, and the Natick Naval Test labs. Lee J. Rosen, DSc, is a senior manager of the combustion research and development group for Praxair, Inc. (Tonawanda, New York) where he has been since 2003. He received his doctorate in mechanical engineering from Washington University in St. Louis, Missouri. Dr. Rosen has 19 years of basic scientific research work, industrial technology research and development, and combustion design engineering. His experience includes oxyfuel


xxii

Contributors

combustion, flame stability, pulsed combustion, and flame synthesis of ultrafine particles. He holds three U.S. patents and has authored several published articles and conference papers. Silvio Rudy Stella, PhD, is the marketing director of Reway SrL (Possagno, Italy) that manufactures cogeneration systems where he has been since 2006. He has a degree in electrical engineering from the Ministry of Scientific Research and Technology (Rome, Italy). He worked in a variety of roles at Riello Energy Group for Life from 2002 to 2005, at Thermital SpA from 2000 to 2002, at Calortecnica SpA from 1992 to 2000, at Gemmo SpA from 1991 to 1992 and at Padova Ricerche from 1989 to 1990. He has participated in a number of organizations and associations, held offices in some of those organizations including president of the Italian cogeneration federated ANIMA called ItalCogen, has given many presentations, has been an expert witness, and is an inventor on five patents. He won a prize in memory of Antonio Sarpi (University of Padua) as the best graduate of the Faculty of Engineering in 1989. Allan M. Runstedtler, MASc, is a research scientist with CanmetENERGY, the energy research and technology centre of Natural Resources Canada of the Canadian Government (Ottawa, Canada). He received his degree in mechanical engineering from the University of Waterloo (Waterloo, Canada) in 2000. He has been the technical lead on the investigation of industrial problems in refinery heaters and metal ore processing. Other work has led to the development of a simple boiler model for process modeling of utility boilers and the conceptual design of an ultra-low NOx burner for heat recovery from microturbines. He is interested in fundamental physical, chemical, and mathematical issues related to energy systems and has authored or co-authored journal papers on radiant properties of combustion gases and diffusion in micro-scale pores. Among his current interests are a fluid dynamic theory of turbulence and the use of density functional theory to study the relationship of material properties and reactivity to atomic structure. Khaled A. Sallam, PhD, is an associate professor at Oklahoma State University (Stillwater, Oklahoma) since 2003. He received his degree in aerospace engineering from the University of Michigan (Ann Arbor, Michigan) where he worked in the spray dynamics lab in 2002. From 2003 to 2009 he worked as an assistant professor of mechanical and aerospace engineering at Oklahoma State University. In the summer of 2009, he was tenured and promoted to associate professor. In 2008 he was selected as a summer faculty fellow for the Air Force Summer Fellowship Program at Wright-Patterson Air Force Base. He was awarded the 2007 Halliburton Excellent Young Teacher Award from Oklahoma State University and the 2006 W.R. Marshall Award from ILASS-Americas—Institute for Liquid Atomization and Spray Systems, North and South America. He is a member of the American Society of Mechanical Engineers, the American Institute of Aeronautics and Astronautics (and a member of AIAA Fluid Dynamics Technical Committee), the American Physical Society, and the Institute for Liquid Atomization and Spray Systems. He published 12 journal articles and 29 conference papers and supervised two PhD students and seven master students. Edwin Schorer, PhD, has been working as a consulting engineer in industrial acoustics for Müller-BBM GmbH (Munich, Germany) since 1989. He received his degree in electrical engineering and his doctorate in psychoacoustics from the Technical University Munich, Germany. He was promoted to managing director in 2006. He works in industrial acoustics in general, including theoretical and applied acoustics, with special focus on noise predictions for flare noise and fan noise, fluid mechanics, ship acoustics, and acoustic optimization of postal automation systems. His research work resulted in 15 publications on psychoacoustics as well as industrial and technical acoustics. Dr. Schorer is a member of the German Institute for Standardization, the Noise Control and Vibration Engineering Standards Committee, and the German Acoustical Society. His research focuses on a functional schematic of just noticeable frequency and amplitude variations. He worked as temporary academic counsel at his university, lecturing electroacoustics and technical acoustics. He acted as supervising tutor for the student’s diploma theses and practical trainings. Robert E. Schwartz, PE, BS, MS, is a senior technical specialist at John Zink Co., LLC (Tulsa, Oklahoma). He has received his degrees in mechanical engineering from the University of Missouri. He has worked in the fields of combustion, flares, pressure relieving systems, fluid flow, and heat transfer for more than 40 years including 42 years with Zink where he has provided technical and business leadership in all product areas and has extensive international experience. He has 51 U.S. patents for inventions of apparatus and methods that are in use throughout the John Zink Co. He is the associate editor of The John Zink Combustion Handbook. His areas of technical expertise include: development, design, fabrication, and operation of combustion equipment including flares, incinerators, process


Contributors

xxiii

burners, boiler burners, and vapor control; reduction of NOx and other emissions from combustion processes; fluid flow and heat transfer in process and combustion equipment; noise elimination and control; vapor emissions control using recovery processes; hazardous waste site remediation; and permitting and operation of hazardous waste storage and disposal sites. His professional organizations and awards include: member of the American Society of Mechanical Engineers, the American Institute of Chemical Engineers and Sigma Xi, the Scientific Research Society; Registered Professional Engineer in the state of Oklahoma; recipient of the University of Missouri Honor Award for Distinguished Service in Engineering and election to the University of Tulsa Engineering Hall of Fame. Pavel Skryja, MS, currently works as a consultant for industrial power burners and combustion equipment. He holds a degree in process engineering from Brno University of Technology. He has more than seven years of experience as a project manager and designer in research and development of industrial power burners designed for refineries and special installations. He cooperates with Brno University of Technology (Czech Republic) on research of renewable liquid fuels and liquid waste combustion. Petr Stehlík is a professor and director of the Institute of Process and Environmental Engineering at Brno University of Technology in the Czech Republic. He currently holds a position of vice president of the Czech Society of Chemical Engineers. He has several years of experience in engineering industrial practice before joining the university, and at present he is also a director of the research and development team for a certified engineering and contracting company with activities focusing on waste and biomass processing. Some of his main activities include: executive editor of Heat Transfer Engineering and guest editor of international journals, coordinator or contractor of international research projects, author or co-author of more than 200 papers in journals and proceedings, and plenary or keynote speaker at various international conferences. His research and development as well as application activities involve waste and biomass processing, waste to energy systems, applied heat transfer, energy saving, and environmental protection. Jun Sudo is a consultant for Nippon Furnace Co., Ltd. (Japan). He has 40 years of experience in combustion technology in the fields of steel, petro-refinery, boiler, and cement industries. He has taken a leading role for the development of regenerative combustion system in Nippon Furnace. He has numerous technical publications and several patents. He co-received the Technical Award from the Combustion Society of Japan in 1995 for The Regenerative Combustion System. Dariusz Szewczyk, PhD, specializes in high temperature combustion and innovative methods of combustion related to traditional fuels, biofuels, industrial, and waste gases. He received his degree at the Poznan University of Technology (Poznan, Poland). His particular interests are in oxygen deficient combustion technologies centered on the idea of lowering fuel consumption and pollutants emission. He worked at the Royal Institute of Technology (Stockholm, Sweden) where he published, either as author or co-author, a number of papers concerning high temperature air combustion as well as high temperature air gasification. From 2004 to 2007, Dr. Szewczyk worked for VTS AB, a Swedish engineering company working in the field of industrial combustion systems. Currently he is a general manager and co-owner of ICS Industrial Combustion System Sp z o.o., Poland, an engineering company working in the field of industrial combustion systems. In VTS as well as in ICS he is responsible for projects utilizing NFK HRS/HTB/HiTAC technology in Europe. Giuseppe Toniato, DrIng, is business innovation manager of Riello Group (Legnago, Italy) where he has been since 1999. He received his degree in mechanical engineering from the University of Padova (Italy) in 1991. He has been director of engineering for Riello Burner Division for eight years. Prior to this he was projects manager in Magneti Marelli Engine Control Division (Fiat Group). He has been engaged in combustion research for more than 14 years. He holds ten patents and has authored or co-authored over ten publications on metallic mat combustion, catalytic combustion, and burner controls. Guido Troiani, PhD, is a researcher in combustion and fluid-mechanics at ENEA (Rome, Italy) where he has been since 2006 as a postdoctorate. He received his degree in fluid mechanics from the Faculty of Engineering of the University of Rome “La Sapienza” (2004). He began his postdoctoral research at the University of Rome “La Sapienza” in cooperation with the Italian Ship Model Basin (INSEAN), performing experiments and theoretical analysis on free-surface turbulence and on the transition to turbulence of laminar flows. Successively he was employed as postdoctorate at the ENEA research center in the field of turbulent combustion. His main topics


xxiv

Contributors

of interest are the interactions between turbulence and combustion, flame chemiluminescence emissions, fractal aspects of flame fronts and heavy particle dynamics from a theoretical and experimental point of view. He carried out his experiments by means of intrusive devices such as Pitot and hot-wire probes, acoustic doppler anemometers, and also spectroscopical techniques: LDA, PDA, PIV, and laser induced fluorescence. He is the author of five peer-reviewed publications and of 15 conference proceedings. Joachim G. Wünning, DrIng, is the general manager and co-owner of WS GmbH (Germany). He received his degree from the RWTH Aachen (Technical University of Aachen, Germany). His special interests include: combustion research, burner development, heat recovery, high efficiency, low emissions, flameless oxidation. He has numerous technical publications and several patents. Weihong Yang, PhD, is an associate professor in the Division of Energy and Furnace Technology, Royal Institute of Technology, Sweden, where he has been since 2000. He has a degree in thermal energy from Central South University of Technology (China). His major research areas include: high-temperature air combustion (flameless, or MILD combustion), flameless oxyfuel combustion, gasification with high-temperature air/steam, improvement of combustion systems in boilers and incinerators and rotary kiln in the process industries. He has published about 40 papers in international journals, and presented over 60 papers at international conferences. He also carried out and managed many research projects financed by the Swedish nation, EU, and industries from Sweden, United States, Japan, China, France, Poland. Tsutomu Yasuda, BS, is auditor (Supervisory Board Member) of NFK-Holdings Co., Ltd., and auditor of Nippon Furnace Co., Ltd (Yokohama, Japan; wholly owned subsidiary of above company). He received a degree of electric and electronic engineering, from Tokyo Institute of Technology (Tokyo, Japan). He has been engaged as director in research and development of high temperature air combustion technology since he joined the company in 1993. He moved in October 1998 to Japan Industrial Furnace Manufactures Association, which is a nonprofit organization attached to the Japanese Ministry of Economy, Trade and Industries. He assumed director of Japan Industrial Furnace in charge of HiTAC industrial furnaces and took responsibility of administrating the national project funded by Japanese Ministry of Economy, Trade and Industries. In March 2000, he returned to Nippon Furnace Kogyo Co. and assumed director of project promotion mainly related to high temperature air combustion. He has 20 patents and authored 15 papers all regarding HiTAC and high temperature steam gasification. He has presented the papers in Germany, France, Italy, Sweden, Poland, Indonesia, Malaysia, Thailand, Vietnam, Taiwan, and P.R. China. Andrea Zambon, PhD, is an associate professor at the University of Padova (Padova, Italy) where he has been a researcher since 1990 and as an associate professor in the Faculty of Engineering since 1998. He received his degree in metallurgical engineering from the Polythecnic of Turin (Italy) in 1986. He has taught courses in metallic materials, technology of the metallic materials, science and technology of composite materials, materials science and metallurgy, and of materials selection and design. He is also a professor in the PhD course of mechatronics and industrial systems of the PhD School in Industrial Engineering by the University of Padova. He is a member of the Steering Committee at the university. He is a member of the Italian Association of Metallurgy (AIM-Milan) and of both the Technical Committee of the Center for Physical Metallurgy and Materials Science and of the Technical Committee of the Center for Welding and Permanent Joints of the Italian Association of Metallurgy. He has authored over 100 scientific or technical publications on: development of mathematical models for the analysis and the forecast of thermal fields in laser welding and processing of metallic materials; application of nonconventional techniques (plasma torches, laser, high velocity oxygen-fuel) in the production of coatings and in the modification of the surfaces for the improvement of the wear resistance; methods of mechanical characterization and study of the tribological behavior of metal matrix composites; and powder metallurgy: processes of production by gas atomization, analysis of the cooling regimes, production of bulk spray-formed and sintered samples, microstructural and mechanical characterization of the spray-formed or sintered products.


Section I

General


1 Introduction Charles E. Baukal, Jr. Contents 1.1 Introduction........................................................................................................................................................................ 4 1.2 Industrial Combustion Applications.............................................................................................................................. 8 1.2.1 Metals Production.................................................................................................................................................. 8 1.2.2 Minerals Production.............................................................................................................................................. 9 1.2.3 Chemicals Production......................................................................................................................................... 10 1.2.4 Thermal Oxidation.............................................................................................................................................. 11 1.2.5 Industrial Boilers and Power Generation......................................................................................................... 11 1.2.6 Drying................................................................................................................................................................... 12 1.3 Combustion System Components................................................................................................................................. 13 1.3.1 Burners.................................................................................................................................................................. 13 1.3.1.1 Burner Design Factors.......................................................................................................................... 13 1.3.1.2 General Burner Types........................................................................................................................... 17 1.3.1.3 Burner Components.............................................................................................................................. 22 1.3.2 Combustors........................................................................................................................................................... 23 1.3.2.1 Design Considerations......................................................................................................................... 23 1.3.2.2 General Classifications......................................................................................................................... 23 1.3.3 Heat Load.............................................................................................................................................................. 25 1.3.3.1 Opaque Materials.................................................................................................................................. 26 1.3.3.2 Transparent Materials.......................................................................................................................... 26 1.3.4 Heat Recovery Devices........................................................................................................................................ 26 1.3.4.1 Recuperators.......................................................................................................................................... 26 1.3.4.2 Regenerators.......................................................................................................................................... 27 1.4 Testing............................................................................................................................................................................... 27 1.4.1 Purposes................................................................................................................................................................ 27 1.4.2 Probe Types........................................................................................................................................................... 29 1.4.3 Configurations...................................................................................................................................................... 30 1.4.3.1 Bench Scale Testing............................................................................................................................... 30 1.4.3.2 Pilot Scale Testing................................................................................................................................. 31 1.4.3.3 Large-Scale Testing .............................................................................................................................. 31 1.4.3.4 Field Testing........................................................................................................................................... 31 1.4.4 Important Input Parameters............................................................................................................................... 31 1.4.4.1 Fuel Composition.................................................................................................................................. 32 1.4.4.2 Air/Fuel Ratio........................................................................................................................................ 32 1.4.4.3 Geometry................................................................................................................................................ 32 1.4.4.4 Furnace Temperature and Pressure................................................................................................... 32 1.5 Experimental Errors........................................................................................................................................................ 33 1.5.1 Sources of Error.................................................................................................................................................... 33 1.5.2 Minimizing Errors............................................................................................................................................... 33 1.5.3 Uncertainty Analysis........................................................................................................................................... 35 1.6 Combustion Testing Resources...................................................................................................................................... 36 1.6.1 General References.............................................................................................................................................. 36

3 © 2011 by Taylor and Francis Group, LLC

4

Industrial Combustion Testing

1.6.2 Test Facilities......................................................................................................................................................... 37 1.6.2.1 Academic................................................................................................................................................ 37 1.6.2.2 Institutional............................................................................................................................................ 38 1.7 Future................................................................................................................................................................................ 38 References................................................................................................................................................................................... 38

1.1 Introduction Combustion systems are among the most challenging technologies to study. There are not only high temperatures, but usually very high temperature gradients ranging from the incoming reactants at ambient temperature up to flame temperatures. The fluid flow is typically turbulent and may include swirl. The heat transfer includes conduction, convection, and radiation. The radiation is further complicated by the spectral nature of gaseous combustion products. The chemistry is extremely complicated, where the combustion of a relatively simple fuel like methane can involve hundreds of chemical reactions and dozens of species. Trace combustion products such as carbon monoxide and nitrogen oxides are critical because they are typically regulated. For liquid and solid fuels, multiple phases are present. The fuel composition can vary widely and may contain multiple components, waste products, and sometimes multiple phases, depending on the process. The length scales in industrial combustion processes may vary by orders of magnitude, ranging from millimeters for fuel injection ports up to meters for the combustor itself. The combustion system may also include heat recuperation equipment such as air preheaters and waste heat boilers, and pollution control equipment such as scrubbers and catalytic treatment reactors. The materials being heated may be solids, liquids, or gases and have a wide range of properties. For example, molten aluminum can be highly reflective, molten glass is spectrally absorptive, and cement is highly absorptive. The field of industrial combustion is very broad and touches, directly or indirectly, nearly all aspects of our lives. The electronic devices we use are generally powered by fossil fuel fired power plants. The cars we drive use internal combustion engines. The planes we fly in use jet fuel powered turbine engines. Most of the materials we use have been made through some type of heating process. While this book is concerned specifically with industrial combustion, all of the above combustion processes have many features in common. Industrial combustion is complicated by many factors. First, the science of combustion is still developing and has a long way to go until we completely understand it so it can be better applied and controlled. While fire has been with us since the beginning of time, much remains to be learned about it. As the science of combustion combines heat transfer, thermodynamics, chemical


kinetics, and multiphase turbulent fluid flow to name a few areas of physics, the study of industrial combustion is interdisciplinary by necessity. Combustion has been the foundation of worldwide industrial development for the past 200 years [1]. Industry relies heavily on the combustion process as shown in Table 1.1. The major uses for combustion in industry are shown in Table 1.2. Hewitt et al. (1994) have listed some of the common heating applications used in industry, as shown in Table 1.3 [2]. As can be seen in Figure 1.1, the worldwide demand for energy continues to increase. Most of the energy (86%) is produced by the combustion of fossil fuels like petroleum, natural gas, and coal (see Figure 1.2). According to the U.S. Department of Energy, the demand in the industrial sector is projected to increase by 0.8% per year to the year 2020 [3]. Figure 1.3 shows that the industrial sector is one of the largest energy consumers in the United States. Figure 1.4 shows the projected energy source and end use for the United States in 2008. This again highlights the importance of industrial combustion. The combustion community has identified a number of driving forces that will shape industrial combustion for the foreseeable future: markets and economics; environmental quality and greenhouse gases; process improvement; policy and politics; fuel/oxidant choices; energy efficiency; enabling technologies; health, safety, and reliability; and research and education [1]. Many of these are complex issues that dynamically change over time. For example, pollution regulations vary by Table 1.1 The Importance of Combustion to Industry % Total Energy from (at the Point of Use) Industry Petroleum refining Forest products Steel Chemicals Glass Metal casting Aluminum

Steam

Heat

Combustion

29.6 84.4 22.6 49.9 4.8 2.4 1.3

62.6 6.0 67.0 32.7 75.2 67.2 17.6

92.2 90.4 89.6 82.6 80.0 69.6 18.9

Source: U.S. Department of Energy (DOE). Industrial Combustion Vision: A Vision by and for the Industrial Combustion Community. Washington, DC: U.S. DOE, 1998.

5

Introduction

Table 1.2 Major Process Heating Operations Metal melting • Steel making • Iron and steel melting • Nonferrous melting Metal heating • Steel soaking, reheat, ladle preheating • Forging • Nonferrous heating Metal heat treating • Annealing • Stress relief • Tempering • Solution heat treating • Aging • Precipitation hardening Curing and forming • Glass annealing, tempering, forming • Plastics fabrication • Gypsum production Fluid heating • Oil and natural gas production • Chemical/petroleum feedstock preheating • Distillation, visbreaking, hydrotreating, hydrocracking, delayed coking

Bonding • Sintering, brazing Drying • Surface film drying • Rubber, plastic, wood, glass products drying • Coal drying • Food processing • Animal food processing Calcining • Cement, lime, soda ash • Alumina, gypsum Clay firing • Structural products • Refractories Agglomeration • Iron, lead, zinc Smelting • Iron, copper, lead Nonmetallic materials melting • Glass Other heating • Ore roasting • Textile manufacturing • Food production • Aluminum anode baking

Source: U.S. Department of Energy (DOE). Industrial Combustion Vision: A Vision by and for the Industrial Combustion Community. Washington, DC: U.S. DOE, 1998.

Table 1.3 Examples of Processes in the Process Industries Requiring Industrial Combustion Process Industry Steel making Chemicals Nonmetallic minerals (bricks, glass, cement, and other refractories) Metal manufacture (iron and steel, and nonferrous metals) Paper and printing

Examples of Processes Using Heat Smelting of ores, melting, annealing Chemical reactions, pyrolysis, drying Firing, kilning, drying, calcining, melting, forming Blast furnaces and cupolas, soaking and heat treatment, melting, sintering, annealing Drying

Source: Adapted from U.S. Department of Energy, Energy Information Administration. Annual Energy Outlook 2008, report DOE/EIA-0384, 2008, Washington, DC, released June 26, 2009.

location and continue to get more rigorous. Technology continues to improve as the emission requirements get more stringent. The U.S. Department of Energy sponsored a workshop to develop a roadmap for industrial combustion technology. The resulting report [4] identified the following priority research and development (R&D) needs: • Through a system approach, create burner designs that effectively transfer heat to the load. • Improve heat recovery processes to capture waste gases cost-effectively. • Create innovative new furnace designs through better fundamental understanding of combustion and scale-up.


• Improve boiler technology and combustion cycles through fundamental design innovations such as gasification and fluidized beds. • Determine advanced methods to maintain a stable flame and achieve low emissions while using different fuels. • Develop a burner capable of adjusting operating parameters in real time. • Develop real-time sensors and process controls that are more reliable and robust in harsh environments. • Develop computational tools that are more accurate in a wide variety of applications through the collection of physical data and model validation.

6

Other Nuclear electric power Hydroelectric power Coal

2006

2004

2002

2000

1998

1996

1994

1992

1990

1988

1986

1984

Natural gas Petroleum

1982

500 450 400 350 300 250 200 150 100 50 0

1980

Quadrillion Btu


Year Figure 1.1 Historical and projected (2008) world energy consumption. (Energy Information Administration).

Natural gas, 23%

Coal, 27%

Hydroelectric power, 6% Nuclear electric power, 6% Petroleum, 36% Other, 2%

Figure 1.2 2006 world energy consumption. (Energy Information Administration).

120 Transportation Industrial Commercial Residential

Quadrillion Btu

100 80 60 40

Year Figure 1.3 U.S. energy consumption by industry sector. (Energy Information Administration).


2009

2006

2003

2000

1997

1994

1991

1988

1985

1982

1979

1976

1973

1970

1967

1964

1961

1958

1955

1952

0

1949

20

7

Introduction

Stock change and other6 0.18 Coal 23.86

Exports 7.06

Natural gas 21.15

Fossil fuels 57.94

Crude oil1 10.52

Coal 22.42 Domestic production 73.71

Petroleum4 27.56

8

Supply 106.55

NGPL22.41 Nuclear electric power 8.46

Renewable energy3 7.32

Petroleum 3.77 Other exports7 3.30

Natural gas 23.84

Fossil fuels10 83.44

Petroleum9 37.14

Imports 32.84

Nuclear electric power 8.46 3

Renewable energy 7.30

Residential12 21.64 Commercial12 18.54

Consumption11 99.30

Industrial12 31.21

Transportation 27.92

Other imports5 5.28

Figure 1.4 2008 projected energy flow (quadrillion Btu) by source and end use in the United States. (Energy Information Administration.)

• Create a pathway to demonstrate and commercialize new technologies and enable information sharing throughout industry about new technologies. • Develop robust design tools that are more userfriendly and accurate, especially with complex phenomena such as turbulence and create a unified code to allow sharing of information more easily and speed development. • Improve integration of boiler systems with the rest of the plant process with new “smart” control systems and designs that capture unused waste heat. Table 1.4 lists the priority R&D needs that were identified for burners, boilers, and furnaces. Essentially all of these needs require some amount of testing. These needs were generated from the following end-use requirements: increased system efficiency; reduced NOx, CO, CO2, and particulate emissions; increased fuel flexibility; more robust and flexible process control and operations; better safety, reliability, and maintenance; lower capital and operational costs; faster, low-cost technology development; and enhanced system integration. Coupled with these needs are some barriers to improvement: financial risk, inability to accurately predict the performance of new systems, lack of industry standards, and the wide gap that often exists between the research done at a small scale that needs to be applied to industrial-scale systems. Testing is often required to address some of these barriers. As shown in Figure 1.5, three elements are required to sustain combustion processes: fuel, oxidizer, and an


ignition source usually in the form of heat. Industrial combustion is defined here as the rapid oxidation of hydrocarbon fuels to generate large quantities of energy for use in industrial heating and melting processes. Industrial fuels may be solids (e.g., coal), liquids (e.g., oil), or gases (e.g., natural gas). The fuels are commonly oxidized by atmospheric air (which is approximately 21% O2 by volume) although it is possible in certain applications to have an oxidizer (sometime referred to as an “oxidant” or “comburent”) containing less than 21% O2 (e.g., turbine exhaust gas [TEG] [5]) or more than 21% O2 (e.g., oxy/fuel combustion [6]). The fuel and oxidizer are typically mixed in a device referred to as a burner that is discussed in more detail below. An industrial heating process may have one or many burners depending on the specific application and heating requirements. Many theoretical books have been written on the subject of combustion, but they have little if any discussion of industrial combustion processes [7–12]. Edwards (1974) has a brief chapter on applications including both stationary (boilers and incinerators primarily) and mobile sources (primarily internal combustion engines) [13]. Barnard and Bradley (1985) have a brief chapter on industrial applications, but have little on pollution from those processes [14]. A book by Turns (2000), which is designed for undergraduate and graduate combustion courses, contains more discussions of practical combustion equipment than most similar books [15]. There have also been many books written on the more practical aspects of combustion. Griswold’s (1946) book has a substantial treatment of the theory of combustion,

8


Table 1.4 Priority Research and Development Needs Burners Burners capable of adjusting operating parameters in real time Sensors and controls for emissions and heat distribution Fuel and oxidant physical and chemical characteristics Fuel and oxidant mixing within the burner Process interactions within the burner system Robust design tools/unified code Fundamental understanding of physical and chemical science Burner system design models (using fundamentals) Performance targets for models More effective heat transfer to load Radiative heat transfer from the flame to the load Heat delivery to the load Physical modeling Sensors and controls Advanced combustion stabilization methods Use of multiple fuels and their characteristics New flame stabilization techniques Boilers New boiler technologies Sensors and controls Alternatives to burners for solid and liquid fuels Fuel and oxidant flexibility Boiler system efficiency improvement

System integration Sensors and controls Application of combined heat and power systems (CHP) Condensate systems integration Boiler personnel training Technology transfer Test and demonstration programs Speed of the commercialization process Furnaces New furnace designs Fundamental knowledge of physical and chemical science Materials of construction Model validation and technology transfer End user product quality Process characterization and development Advanced sensors and process control Best practices (from other industries) Advanced materials of construction Application of sensors and controls Cost-effective heat recovery Design, application, and performance of heat recovery systems Advanced materials of construction Flue gas characterization Integrated computational design tools Design tool characteristics Technology transfer

Source: Energetics, Inc. Industrial Combustion Technology Roadmap: A Technology Roadmap by and for the Industrial Combustion Community, October 2002, www1.eere.energy.gov/industry/combustion/pdfs/combustion_roadmap2002.pdf, 2002.

Fuel

Oxygen

Source of ignition Figure 1.5 Combustion triangle. (From Baukal, C. E., ed., Oxygen-Enhanced Combustion, Boca Raton, FL: CRC Press LLC, 1998; courtesy of CRC Press.)

but is also very practically oriented and includes chapters on gas burners, oil burners, stokers and pulverized coal burners, heat transfer, furnace refractories, tube heaters, process furnaces, and kilns [16]. Stambuleanu’s (1976) book on industrial combustion has information on actual furnaces and on aerospace applications, particularly rockets [17]. There is much data in the book on flame lengths, flame shapes, velocity profiles, species concentrations, liquid, and solid fuel combustion. A book on industrial combustion has significant discussions on flame chemistry, although little on pollution from flames [18]. A book by Borman and Ragland (1998) attempts to bridge the gap between the theoretical and practical


books on combustion [19]. Deshmukh (2005) discusses a range of topics of interest in industrial combustion including fuels, burners, and refractories, where the emphasis is more on metal treatment [20]. However, the book has little discussion about the types of industrial applications considered here. Keating’s (2007) book on applied combustion is aimed at engines and has no treatment of industrial combustion processes [21]. Mullinger and Jenkins (2008) have written a very practical book that discusses a wide range of industrial furnaces and processes [22]. Surprisingly, many handbooks on combustion applications have little if anything on industrial combustion systems [23–27].

1.2 Industrial Combustion Applications Some of the most common industrial applications are briefly discussed next. Many of them are discussed in more detail in other chapters in this book. 1.2.1 Metals Production Metals are used in nearly all aspects of our lives and play a very important role in society. The use of metals

9

Introduction

has been around for thousands of years. There are two predominant classifications of metals: ferrous (ironbearing) and nonferrous (e.g., aluminum, copper, and lead). Ferrous metal production is often high temperature because of higher metal melting points compared to nonferrous metals. Many metals production processes are done in batch, compared to most other industrial combustion processes considered here (e.g., glass production) that are typically continuous. Another fairly unique aspect of metal production is the very high use of recycled materials. This often lends itself to batch production because of the somewhat unknown composition of the incoming scrap materials that may contain trace impurities that could be very detrimental to the final product if not removed. The metals are typically melted in some type of vessel and then sampled to determine the chemistry so that the appropriate chemicals can be either added or removed to achieve the desired grade of material. Another unique aspect of the metals industry is that transfer vessels are preheated prior to the introduction of molten metal into the vessel to minimize the thermal shock to the refractory. Figure 1.6 shows an example of preheating a transfer ladle used to move molten metal around a plant. Since metals melt at higher temperatures, higher intensity burners are often used in these applications. This includes, for example, oxygen-enhanced combustion (OEC) [28,29] (see Chapter 27) and air preheating (see Chapter 21) to increase the flame temperatures and metal melting capability. These higher intensity burners have the potential to produce high pollutant emissions

Firewall

Ladle seal Ladle

Figure 1.6 Ladle preheater. (From Baukal, C. E., ed., Industrial Burners Handbook, Boca Raton, FL: CRC Press, 2004.)


so the burner design is important to minimize these emissions. Another somewhat unique aspect of metals production is that supplemental heating may be required to reheat the metals for further processing. For example, ingots may be produced in one location and then transported to another location to be made into the desired shape (e.g., wheel castings are often made from remelting aluminum ingots or sows). While this process may be economically efficient, it is energy and pollutant inefficient due to the additional heating. Burners are used in the original melting process as well as in the reheating process (see Chapter 9). This is something that has begun to attract more attention in recent years where the entire life cycle of a product is considered rather than just its unit cost and initial energy requirements. For example, aluminum has a low life cycle cost compared to many other metals because of its high recycle ratio. While the energy consumption to make aluminum from raw ore is fairly high, remelting scrap aluminum takes only a fraction of that energy that also means less overall pollution as well. Burners commonly used in the metals industry include high velocity burners, regenerative burners (Chapter 21), radiant tube burners (Chapter 24), air-oxy/ fuel burners and oxy/fuel burners (Chapter 27). 1.2.2 Minerals Production Some common minerals processes include the production of glass, cement, bricks, refractories, and ceramics. These are typically high temperature heating and melting applications that require a significant amount of energy per unit of production. They also tend to have fairly high pollutant emissions as a result of the high temperatures and unit energy requirements. Most of the minerals applications are continuous processes, but there is a wide range of combustors. Large glass furnaces are typically rectangularly shaped and have multiple burners (see Chapter 32). On the other hand, cement kilns are long refractory-lined rotating cylinders that are slightly inclined so that the materials flow gradually down hill (see Figure 1.7; see Chapter 31). A typical cement plant is shown in Figure 1.8. Many of the minerals applications employ some type of heat recovery in the form of air preheating to improve To off gas treatment Feed materials

Burner

Processed clinker

Figure 1.7 Countercurrent rotary cement kiln schematic. (From Baukal, C. E., ed., Industrial Burners Handbook, Boca Raton, FL: CRC Press, 2004.)

10


Figure 1.8 Cement plant. (From Baukal, C. E., ed., Industrial Burners Handbook, Boca Raton, FL: CRC Press, 2004.)

Figure 1.9 Refinery. (From Baukal, C. E., ed., Industrial Burners Handbook, Boca Raton, FL: CRC Press, 2004.)

energy efficiency. However, the heat recovery typically significantly increases NOx emissions. While recycling of used glass (referred to as cullet) is practiced in some applications, there is generally must less recycling in the minerals industry compared to the metals industry. 1.2.3 Chemicals Production This is a very broad classification that encompasses many different types of production processes that have been loosely subcategorized into chemicals (organic and inorganic) and petrochemicals (organic) applications. A typical refinery is shown in Figure 1.9. There is some overlap in terms of the types of heating equipment used where many of the incoming feed materials are in liquid form (e.g., crude oil) that are processed in heaters with tubes running inside them. These are generally lower temperature applications ( 3 or so. To accommodate fewer experimental points, fractional factorial designs have been developed. We shall discuss each in turn. A two-level full factorial is an experimental design that investigates all possible factor combinations at two levels: high and low. These levels are coded to ±1 using the transforms presented in Equation 3.2. Since the designs use ±1, generally, only the sign of the factor is reported. To construct a full factorial in f factors, prepare f columns of design points, with minus and plus signs alternating in blocks of 2f−k where k is the factor subscript. For example, for f = 3 there will be 2f = 8 design points. Then x1 alternates in blocks of four (23−1), x2 alternates in blocks of two (23−2), and x3 alternates in blocks of one (23−3). Table 3.1 gives the design. Once the design is selected, the experiments are run, the responses are measured, and the coefficients calculated. As an example, the reader should duplicate the least squares solutions calculated earlier. However, something very important is missing from the analysis—an analysis of experimental error.

If the first four experimental points were run before the lunch break where cooler temperatures prevailed and the second four experimental points were run in the afternoon where warmer temperatures prevailed, then x1 would be confounded with ambient temperature. If a1 were judged to be significant, it would be impossible to know whether the effect was due to variations in x1 or variations in the temperature. Even worse, since we are unaware that ambient temperature affects the results, we would ascribe the results to a1 even though there may be no such association! Therefore, randomizing the run order is an important prophylactic against serial correlation. The presumption in Equation 3.1 is that there is only one error structure to worry about. That is, it presumes that all experiments are run in such a way that they differ only with respect to changes associated with various levels of the factors. This would be likely with randomized runs occurring over a short span of time but unlikely if, for example, half the experiments were run in January and the other half in October. Unknown factor variations in feedstock, ambient conditions, differences in equipment used, and so on could all affect the results. If, however, we can presume that Equation 3.1 is an accurate model, then we can estimate the statistical significance of each coefficient. To do this, let us consider the example given in Table 3.2. We postulate the following model

3.4 About Experimental Error

Cognizance of experimental error is critical for understanding the meaning of the coefficients and conducting the experiments themselves. As a first principle, factorial experimental designs should be run in fully randomized order. This is necessary in order to break any serial correlations that may exist. For example, suppose the design of Table 3.1 were run in the order presented. Further suppose that ambient temperature is an important but unrealized factor affecting the responses.

We have truncated the design at second-order. This is equivalent to presuming that the x1x2x3 interaction is insignificant. Presuming higher order interactions are negligible is usually a good assumption. However, if we would like an independent error assessment, we can duplicate some points. We discuss this later. For now, we shall continue on the basis of our assumption. Equation 3.16 shows the normal equation and the solution to the

3


y = a0 +

∑ k =1

2

ak x k +

3

∑∑ a x x + ε jk

j k

(3.15)

j< k k =1

67

Experimental Design

coefficients. Table 3.2 results in the following normal equation and coefficient estimates.  318.5  8  120.9      82.6    57.0  =      38.9       − 3.1    0.9  

8 8 8 8 8

  a0   a0   39.8   a   a   15.1  1   1      a2   a2   10.3   a  ;  a  =  7.1 .  3   3      a12   a12   4.9         a13   a13   −0.4 8  a23   a23   0.1

(3.16)

The problem now is to decide if the model as a whole is significant, and if so, which coefficients are significant. This is most easily done in an analysis of variance table (ANOVA).

Therefore, any of the quantities in Equation 3.19 may be used interchangeably for yˆ T yˆ in Equation 3.18. Note also the following equalities that come in handy when constructing the ANOVA. SSM = ( yˆ T yˆ − y T y ) = a 1T… ( X T y )1 ,

SSR = ( y T y − yˆ T yˆ ) = y T y − a T ( X T y ) ,

(3.20)

SST = ( y T y − y T y ) = y T y − a T0 ( X T y )0 where T is the coefficient row vector, starting from 1 a 1 (the second element) rather than 0 (the first element), T ( X y )1 is the XTy column vector starting from 1 (the second element) rather 0 (the first element), a T0 is a0, the first element of the coefficient vector, ( X T y )0 is the first element of the XTy column vector (the one associated with a0).

3.5 Sums of Squares ANOVA is constructed using the following equation:

∑ ( yˆ k

− y) + 2

k

∑(y

k

2 − yˆ ) =

k

∑(y

− y ) , 2

k

(3.17)

k

where k indexes the observations, yk is the kth observation, y is the average of all observations, ∑kyk/n, and, yˆk is the kth predicted observation (least squares solution). The first term is known as the sum of squares, model (SSM). The second term is known as the sum of squares, residual (SSR), and the final term is known as the sum of squares, total (SST). Equation 3.16 is true for any least squares solution whatsoever. However, SSM can be apportioned by component (i.e., a0, a1, a2, …) only for so-called orthogonal models or data sets—that is, those that generate diagonal XTX matrices (as is the case for factorial designs). Equation 3.16 also has a matrix formulation:

( yˆ T yˆ − y T y ) + ( y T y − yˆ T yˆ ) = ( y T y − y T y ) .

(3.18)

Also, since y = Xa, by the rules of matrix algebra, Equation 3.19 also holds

yˆ T yˆ = a T ( X T X ) a = a T ( X T yˆ ) = a T ( X T y )


(3.19)

3.6 Degrees of Freedom The sums of squares are an important component of the ANOVA table that we shall construct. Another important component is known as the degrees of freedom (DF). The total degrees of freedom (DFT) is given by n – 1, because if we know y and n – 1 other of the yk values, we can calculate the remaining yk value. Likewise, there are m – 1 degrees of freedom for the model (DFM) because yˆk is determined from m parameters (i.e., the model coefficients) and we must subtract 1 DF for y. Finally, there are n – m degrees of freedom in the residual (DFR) because the last term comprise n values of yk and m values to determine yˆk. We are now ready to build our ANOVA (Table 3.3).

3.7 The ANOVA Making use of these terms, the general ANOVA table is organized as shown in Table 3.3. Applying this to the data of Table 3.2 and making use of Equation 3.15 gives the ANOVA shown in Table 3.4. We shall take this opportunity to point some salient features of the ANOVA table. The first three columns of cell entries are defined by appending the left-hand column entry (M, R, or T) to the top row

68


Table 3.3 The General ANOVA for Model and Residual Terms Parameter M

F

SS

DF

MS

2 ∑ ( yˆ k − y )

m−1

2 ∑ ( yˆ k − y )

k

k

2 ∑ ( yˆ k − y )

 m−1  k  2  ∑ ( y k − yˆ )  N − m 

m−1

centered at zero by letting z = y – μ/σ. Unfortunately, we typically do not know µ and σ precisely. However, if our sample is sufficiently large, we can estimate them with good confidence. In that case, we use y to estimate the mean and s to estimate the standard error. These are given by the following formulas:

k

R

2 ∑ ( y k − yˆ )

∑ ( y k − yˆ )

n−m

k

k

∑ ( yk − y )

T

2

2

N−m

y=

n−1

k

Table 3.4

ANOVA for Table 3.2 and Equation 3.15 Parameter

SS

M R T

DF

3275.7 5.3 3281.0

6 1 7

F

MS 546.0 5.3

103.4

entry (SS, DF, or MS). For example, SSR = 5.7, DFM = 6, and MSM = 546.0. The mean square is MS and F is Snedecor’s ratio. We shall have more to say about both of these presently. But first, we digress to a discussion of error.

3.8 Error If the model accounts for all of the major influences in the response then the residual cumulates all the minor errors. The central limit theorem states that these errors will be distributed as a normal distribution, characterized by a mean (µ) and a standard error (σ) according to Equation 3.21.

PN ( y ) ∼

1 e 2 πσ 2

1  y−µ  −  2  σ 

2

, PN ( z) ∼

y−µ and z = σ

1 e 2π

z2 − 2

=

, 2 2 πe ( z ) (3.21)

where PN(y) is the probability that the normal distribution will have the value at y. In Equation 3.21, the parameter µ indicates where the distribution will be centered (central tendency) and the parameter σ indicates the dispersion (how narrow or broadly the distribution will spread with y). This is the familiar bell-shaped curve shown in so many statistics books. Generally the curve is normalized in terms of standard error units (σ) and


N

∑(y k

k

k

k

,

(3.22)

− y)

2

.

n−1

(3.23)

Note that in Equation 3.23, the denominator is reduced by 1 in order to subtract the degree of freedom used to estimate the mean, as discussed previously. Equation 3.22 presumes that each observation differs only by random error, ε. Since the long run average of ε is 0, then averaging with a sufficiently large sample will give a good estimate for µ by way of y. However, an arithmetic average may not be used to estimate σ because the result will always tend to zero. Rather, we shall use a root mean square (RMS) procedure to derive the dispersion parameter. If the various yk differ 2 only by experimental error, then ( y k − y ) is a measure of σ2, known as the variance. The long run average of these measures will tend toward σ2. However, a short run average will be biased if we divide by n. This is because we have used the data themselves to estimate y. Therefore, we must subtract one degree of freedom in order not to double count one data point and unfairly shrink our estimate of σ2. The true variance is estimated by s2:

1

s=

∑y

s2 =

∑(y k

k

− y)

n−1

2

.

(3.24)

Returning to the ANOVA, let us consider what would happen if the model were not statistically significant. In that case there would be no difference between yˆ and y, except for experimental error; that is, y − yˆ = ε. In that case, both MSM and MSR would estimate σ2. In such a case, the ratio MSM/MSR ~ 1. Indeed, there would be no reason to use two separate estimates of the same quantity and we would simply pool both estimates. By inspection, the reader may see that adding SSM + SSR and dividing by DFM + DFR, we arrive at Equation 3.24. If, on the other hand, the model were statistically significant, it would be improper to pool the estimates and Equation 3.24

69

Experimental Design

would not be appropriate for estimating σ2. In such a case, MSR would be the appropriate estimate for σ2 because it alone averages the random error. Moreover, MSM/MSR >> 1 in such a case. The question now arises. How much g reater than 1 should MSM/MSR be in order to decide that the model is significant? To decide, we must know the probability distribution for MSM/MSR. One thing is certain, it will not be distributed as PN because the range of PN is –∞  1. Tabulations exist for critical values (Fcrit) of the F distributions. These are shown in Appendix A. To determine Fcrit from the table, we need to know three things. First, how many degrees of freedom are in the model (DFM)? Second, how many degrees of freedom are in the residual (DFM)? Third, since we can never be perfectly confident that the values do not differ by some chance event, what probability of a false positive (α) are we willing to tolerate? Generally α is set to 0.05 (sometimes noted as p = 0.05; i.e., 95% confidence that if F exceeds Fcrit, the model accounts for significantly more variance than could occur merely by chance). From Appendix A we see that Fcrit (6, 1, 0.05) = 234.0; however, F = 103.4. Since F fails to equal or exceed Fcrit, the model is not statistically significant as a whole. However, perhaps some parameters are statistically significant. An advantage of factorial designs is that they are orthogonal and therefore permit separate assessment of each model parameter. We can reconstruct the ANOVA table (see Table 3.5) using the kth sum of squares as ak(XT y)k and noting that Table 3.5 ANOVA with an Accounting of Individual Parameters Parameter

F

SS

DF

MS

a1

a1(XTy)1

1

a1(XTy)1/1

a1 (X T y)1 / ∑ ( y k − yˆ ) / 1

a2

a2(XTy)2

1

a2(XTy)2/1

a2 (X T y)2 / ∑ ( y k − yˆ ) / 1

a3

a3(XTy)3

1

a3(XTy)3/1

2 a3 (X T y)3 / ∑ ( y k − yˆ ) / 1

a12

a12(X y)12

1

a12(X y)12/1

2 a12 (X y)12 / ∑ ( y k − yˆ ) / 1

T

T

2

k

2

k

k

1

a13(XTy)13/1

a13 (X y)13 / ∑ ( y k − yˆ ) / 1

a23

a23(XTy)23

1

a23(XTy)23/1

a23 (X T y)23 / ∑ ( y k − yˆ ) / 1

R

∑ ( y k − yˆ )

1

∑ ( y k − yˆ ) / 1

T

3281.0

2

k

7


Parameter a0 = 39.8 a1 = 15.1 a2 = 10.3 a3 = 7.1 a12 = 4.9 a13 = −0.4 a23 = 0.1 R T

SS

DF

MS

(presumed significant) 1827.2 1 1827.2 851.9 1 851.9 406.3 1 406.3 189.0 1 189.0 1.2 1 1.2 0.1 1 0.1 5.3 1 5.3 3281.0 7

F * 345.9* 161.3* 76.9* 35.8* 0.2 0.0

* Significant effects.

Table 3.7 ANOVA for Table 3.2 and Equation 3.15 with a Reduced Model: y = a0 + a1x1 + a2x2 + a3x3 + a12x1x2 + ε Parameter M R T

SS

DF

MS

F

3274.5 6.6 3281.0

3 3 6

1091.5 2.2

498.5*

* Significant effects.

SSM = ∑ mk =1 ak ( X T y )k ; that is, excluding a 0(XT y)0. Note we are asking the following statistical question. Do any parameters besides a 0 add some significant advantage to the model? Even if there are good theoretic reasons to exclude a 0 in our model, we shall always compare the final model with a model comprising a single parameter—the mean—which for factorial designs is a 0. Therefore, a 0 is excluded from the analysis. Table 3.6 shows the values. From Table 3.6 we see that a1 through a12 are statistically significant as indicated by an asterisk in the F column; a 0, in the absence of other parameters, is equivalent to y the hypothesis against which we are testing all other effects. However, a13 and a23 are not statistically significant. As such, they can be pooled into the residual. Since a1 through a12 are significant, they could be pooled into the model. If this were done, the reduced model would now show itself to be significant (see Table 3.7).

k

a13(XTy)13

2

ANOVA for Table 3.2 and Equation 3.15 with an Accounting of Individual Parameters

T

a13

k

Table 3.6

2

T

k

2

k

3.9 Using Center Points and Replicates The residual error may be divided into two components—random error (also called pure error, the

70


accumulation of manifold minor sources that act in no organized way) and bias error (a systematic error that is not random). Bias error is also called lack of fit. Now if we replicate a point exactly and randomize the run order, any difference in responses to the replicate points should be due to random error alone. If we subtract this from the residual, then what remains is an estimate for bias or lack of fit. We can test the bias component in the same way as any other and see whether it is indeed significant. If it is significant, we can augment the model with additional points to eliminate any bias caused by an incorrect model specification. If the bias component is not significant then we will have additional confidence that our model is adequate and its terms are significant. We want to add replicate points in a way that will not disturb the orthogonal properties of the XTX matrix. One way to do this is to replicate all points in a design. Another way is to add several center points. Replicating center points helps distinguish bias error in two ways—replicating them allows for an independent assessment of pure error and their location allows us to detect any curvature or systematic bias in our model. Additionally, replicating center points is typically less expensive experimentally as we do not need to double or triple the existing number of runs. Table 3.8 shows the design augmented with three additional points. We must now add two additional rows (four additional entries) to our ANOVA. We need a row for bias comprising the appropriate entries for sums of squares, bias (SSB) and the associated degrees of freedom (DFB). Additionally, we need the pure error terms (SSE

Table 3.8 23 Factorial Design Plus 3 Center Points (Hypothetical Data) Oxygen x1

Temperature x2

H2 in Fuel x3

NOx, ppm y

1 2 3 4 5 6 7 8

− − − − + + + +

− − + + − − + +

− + − + − + − +

12.7 25.8 21.7 38.6 32.3 47.2 64.1 76.1

9 10 11

0 0 0

0 0 0

0 0 0

39.5 41.4 39.3

Point


and DFE). Regarding the bias, we have the following formulations: n

SSB =

∑(y

k

2 − yˆ k ) = y k T y k − yˆ T yˆ , DFB = u − m (3.25)

k =1

where k indexes each response, y k is the mean of all the runs performed at the design coordinate associated with yk. In the case of nonreplicated points, y k = yk because a single value is its own mean. The number of unique points is u (i.e., number of coordinates or unique positions in factor space), and m is the number of model parameters (includes a0). Regarding the pure error, we may account for it thus: n

SSE =

∑(y

− y k ) = y T y − y k T y k , DFB = n − u (3.26) 2

k

k =1

The total number of points is n, and u is the number of unique points (i.e., number of coordinates or unique positions in factor space). Constructing the ANOVA from Table 3.8 and Equation 3.15 we note that n = 11; that is, there are a total of 11 runs, u = 9; that is, there are nine unique coordinate points (8 factorial coordinates and 1 center point coordinate); therefore, u = 9, and m = 7; that is, there are seven model parameters—a 0, a1, a2, a3, a12, a13, a23. If we chose to fit the reduced model—a 0, a1, a2, a3, a12—then m = 4. Table 3.9 shows the generic ANOVA table for individual model parameters and with residual split into bias and pure error components. Table 3.10 gives the actual numbers. For convenience, we have appended the values of F to the ANOVA table and added an asterisk for statistically significant effects. From the table in Appendix A, we have Fcrit(0.05, 1, 1) = 18.5 and Fcrit(0.05, 2, 2) = 19.0. As before, a0 through a12 are significant; but notably, the bias is not significant since 2.7  5. Moreover, in principle, one needs only as many unique data points as there are coefficients to fit, plus additional replicates as necessary. In such cases, full factorials may have many more runs than are necessary. For example, consider that we wish to investigate performance in response to the following factors: • ξ1: oxygen concentration in the furnace, from 1 to 3% • ξ2: hydrogen concentration in the fuel, from 0 to 25% • ξ3: propane concentration in the fuel, from 0 to 25% • ξ4: furnace temperature, from 1400 to 1800°F • ξ5: fuel injection angle, from 5 to 15°

* Statistically significant, α = 0.05.

Table 3.11 ANOVA for Factorial with Replicate Points with Table 3.8 and Equation 3.15 as the Basis Parameter a0 = 39.9* a1 = 15.1* a2 = 10.3* a3 = 7.1* a12 = 4.9* R T

SS

DF

MS

(presumed significant) 1827.2 1 1827.2 851.9 1 851.9 406.3 1 406.3 189.0 1 189.0 9.3 6 2.7 3283.8 10

* Statistically significant, α = 0.05.


F

Fcrit

* 1827.2* 851.9* 406.3* 189.0*

6.0 6.0 6.0 6.0

Suppose further that we are only interested in the first-order terms per Equation 3.27: 5

yˆ = a0 +

∑a x

k k

k =1

.

(3.27)

72


The formula has 1 zero-order coefficient (a0), 5 first-order coefficients (a1 → a5), and needs at least one degree of freedom to estimate of error. Then the minimum number of runs required is 7. However, a full factorial would have 25 = 32 experimental points and be able to fit the following 32 coefficients associated with the respective factors and interactions.

a0 a5 a23 a45 a135 a345 a2345

a1 a12 a24 a123 a145 a1234 a12345

a2 a13 a25 a124 a234 a1235

a3 a14 a34 a125 a235 a1245

a4 a15 a35 a134 a245 a1345

This seems excessive. Can we reduce the number of required experiments? The answer is “Yes”, if we are willing to give up on higher-order effects. Fractional factorial coefficients offer one method for doing this and require 1/2 k · 2 f or 2f−k runs where 1/2k represents the fraction. For example, if we run the half fraction (k = 1), this will require only 16 runs. We do this by first preparing a 24 factor design and then we add a fifth column. We shall describe presently how the sign order is determined in this last column. For now, please examine Table 3.12. This design has all the desirable orthogonal properties of the full factorial. However, we have not gotten something for nothing. Although we will be able to determine up to 16 coefficients, they will all be confounded with others. We show the confounding pattern presently. Rather than list the coefficients, we merely list

Table 3.12 A ½ 25 Fractional Factorial Point

x1

x2

x3

x4

x5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

− − − − − − − − + + + + + + + +

− − − − + + + + − − − − + + + +

− − + + − − + + − − + + − − + +

− + − + − + − + − + − + − + − +

+ − − + − + + − − + + − + − − +


the factor numbers. For example, the x1x2x5 interaction requiring determination of a125 coefficient will be represented merely by the factor list 125. 0 = 12345 1 = 2345 2 = 1345 3 = 1245 4 = 1235 5 = 1234 12 = 345 13 = 245 14 = 235 15 = 234 23 = 145 24 = 135 25 = 134 34 = 125 35 = 124 45 = 123 From this list, we see that every coefficient is confounded with another. For example, a 34 is confounded with a125. This means though it will be possible to determine the a 34 coefficient (or the a125 coefficient— both are identical), it will not be possible to separately know the influence of x34 from x125 or the individually associated coefficients. Both effects will be merged. Notwithstanding, third- and higher-order effects are rarely important. Therefore, this is not a significant loss, and this is what makes fractional factorials so powerful. In this design, we see that no two-factor interaction is confounded with another two-factor interaction. Why not fractionate further? Let us examine the confounding pattern for the ¼ 25 factorial (see Table 3.13). Confounding pattern for Table 3.13: 0 = 145 = 235 = 1234 1 = 45 = 234 = 1235 2 = 35 = 134 = 1245 3 = 25 = 124 = 1345 4 = 15 = 123 = 2345 5 = 14 = 23 = 12345 12 = 34 = 135 = 245 13 = 24 = 125 = 345 We see from the confounding pattern that main and two-factor effects are confounded with other two-factor effects. Since two-factor interactions are often significant, highly fractionated designs like this are typically used for screening. In other words, we believe that firstorder effects will likely show up if significant; however, their values are subject to error. The reason for using highly fractionated designs is to weed out less important factors and reduce the design to fewer factors that can be investigated in greater detail, perhaps with a full factorial. Table 3.13 A ¼ 25 Fractional Factorial Point

x1

x2

x3

x4

x5

1 2 3 4 5 6 7 8

− − − − + + + +

− − + + − − + +

− + − + − + − +

− + + − + − − +

+ − − + + − − +

73

Experimental Design

y02

3.11 Generating the Fractional Design So how do we generate the fractional factorial? As stated before, we first prepare a 2f−k and determine the k remaining columns by selecting k defining contrasts. A defining contrast is any factor or factor interaction that occurs in the full factorial (e.g., 123, 1234, 245, etc.). We choose the defining contrasts to be as large as possible without significant overlap. For example, the ½ 25 factorial has k = 1. So we need to select 1 defining contrast; we do not need to worry about a second defining contrast or its overlap with the first. In such a case, we shall use 12345, the largest possible contrast in the 25 factorial. That is how column 5 in Table 3.13 was derived: by multiplying 1234 (that is x5 = x1x2x3x4). For example, the value for x5 in row 12 is found by x5 = (+)(−)(+)(+) = (−). This is equivalent to setting 12345 = + 1 in all rows. (We could have just as easily set 12345 to −1 as well. In that case, all the signs in column x5 would be reversed.)

3.12 Calculating the Confounding Pattern The method for calculating the confounding pattern also involves the defining contrast. The rule is that we shall multiply the contrast with the effect of interest and eliminate the squared components. For example (125)(12345) = 12223452 = 34. Therefore 125 = 34, and the reader may verify that multiplying either the x1x2x5 or x3x4 columns together gives an identical factor pattern. We repeat this for every possible factor combination so the entire confounding pattern may be found. In order to find the ¼ fraction, k = 2, and f – k = 5 − 2 = 3, we start with a 23 full factorial and we choose two defining contrasts. We attempt to choose them with as little overlap as possible because with multiple contrasts we need to worry not only about the defining contrasts but the implicit contrasts created by their interaction. For example, if we choose 1234 and 235 we also have, tacitly, 1223245 = 145, and this is how Table 3.12 was generated. Other choices fare no better (124 and 235 give 1345) and some fare much worse (e.g., 124 and 125 give 5, meaning that we will be confounding a5 with a0 – a very bad idea).

3.13 Central Composite Designs Generally, for large experimental design problems it is useful to tackle the job sequentially. For example, for 11


y22

y12

y0,0

y10

y20

y11

y21

y01 Figure 3.2 A central composite design. In two dimensions, the design comprises eight points and as many centerpoint replicates as required.

factors, one uses a highly fractionated factorial. Suppose such a screening design identifies two factors as important. Then we investigate them with a full factorial having a few center points. But when we analyze the data via ANOVA, we find significant lack of fit (bias). In such a case we may wish to augment the design with additional points that will allow us to fit a full second-order design to include pure quadratics. A very good design for doing this is known as a central composite. It comprises a full or fractional factorial augmented with axial points. Figure 3.2 shows a central composite in two factors. The design comprises four factorial points (connected in the figure for illustration by a dotted square), four axial points, and as many center point replicates as necessary (typically three to six). The central composite design allows one to fit a full second-order model of the following kind: f −1

f

y = a0 +

∑ k =1

ak x k +

f

∑∑ j< k k =1

f

a jk x j xk +

∑a

x + ε. (3.28)

2 kk k

k =1

Note the last summand indexes pure quadratic terms; that is, a11x12 + a22 x22 + . Central composites interrogate the factor space at five levels along each axis. The question becomes what axial levels should we use? We have two main choices.

1. Make the design strictly rotatable (e.g., circular, spherical, or hyperspherical) by using α = f ,

74


where all axial points lie along an axis ±α from the center. 2. Make the design strictly orthogonal by choosing α= nn f − n f / 2 , where n is the total number of points in the design, and nf is the number of factorial design points.

Typically the first option is chosen and the quadratics are assessed as a single orthogonal entry by fitting the model: f −1

f

yˆ = a0 +

∑

ak x k +

k =1

f

∑∑

f

a jk x j xk +

j< k k =1

∑a

kk

( xk2 − qk ) (3.29)

k =1

where qk = ∑ nk =1 xk2 /n. For example, consider the central composite design of Table 3.14. For this design, we shall fit the model of Equation 3.30: yˆ = a0 + a1 x1 + a2 x2 + a12 x1 x2 + [ a11 ( x12 − q1 ) + a22 ( x22 − q2 )]

(3.30)

where q1 = q2 = 2/3. Subtracting the mean square from x1 and x2 keeps them orthogonal to a0 and the rest of the model terms. Although the bracketed terms are orthogonal to all other model terms, a11 is not orthogonal with respect to a22. Therefore, we shall assess the entire bracketed quantity separately in the ANOVA. If it is significant we shall not know if one or both quadratic terms are significant. There is also a small chance that both terms could be significant but of exactly opposite effect so that they do not appear significant when pooled in the ANOVA. This is not very likely but it is possible.

Table 3.14

Table 3.14 generates the coefficients and ANOVA given in Table 3.15. From the ANOVA, all the parameters are significant including the entry for the quadratics. However, the quadratic coefficients are widely divergent with a11 = 0.2 and a22 = 3.9. It appears that a22 could be the only reason that the quadratic entry is significant. It is also apparent that there is no significant bias, meaning the model fits the data adequately. Therefore, we can pool the bias and pure error into a single residual, MSR. This gives Table 3.16. In this case MSR = 0.08 and s = √0.08 = 0.29. To assess the quadratic effects separately, we can compare each parameter with its standard error. The standard error is defined by Equation 3.31:

sk = s 2 ( X T X ) k , k −1

(3.31)

where sk is the standard error for the kth effect, s is the standard error given by the best estimate of error (√MSR if the bias and error can be pooled or √MSE if it cannot), −1 and ( X T X )k , k is the kth diagonal element of the inverse matrix.

Table 3.15 ANOVA with Coefficients for Table 3.14 Data Parameter a1 = 30.1* a2 = 15.0* a12 = 5.1* a11 = 0.2* + a22 = 3.9 B E T

SS

DF

MS

F

Fcrit

1800.12 799.62 103.02 100.85 0.06 0.43 2805.79

1 1 1 2 3 3 11

1800.12 799.62 103.02 50.43 0.02 0.14

12,559* 5579* 719* 352* 0.1

10.13 10.13 10.13 9.55 9.28

* Significant at p = 0.05.

A Central Composite Design in Two Factors Point

x1

x2

Y

1 2 3 4 5 6 7 8

− − + + 0 0

−

−√2 √2

+ −√2 √2 0 0

11.6 21.5 31.4 61.6 21.0 49.2 6.4 48.9

9 10 11 12

0 0 0 0

0 0 0 0

27.0 27.8 27.2 27.0


+ −

Table 3.16 ANOVA with Coefficients for Table 3.14 Data Having a Pooled Residual Parameter a1 = 30.1* a2 = 15.0* a12 = 5.1* a11 = 0.2* + a22 = 3.9 R T * Significant at p = 0.05.

SS

DF

MS

F

Fcrit

1800.12 799.62 103.02 100.85 0.49 2805.79

1 1 1 2 6 11

1800.12 799.62 103.02 50.43 0.08

12,559* 5579* 719* 352*

5.99 5.99 5.99 5.14

75

Experimental Design

Table 3.17

Table 3.18

Analysis of Standard Error of Effects

A 24 Factorial Design in Two Blocks

Coefficient a0 = 30.1* a1 = 15.0* a2 = 10.0* a12 = 5.1* a11 = 0.2 a22 = 4.0*

sk 0.08 0.10 0.10 0.14 0.11 0.11

t Ratio

tcrit

Point

x1

x2

x3

x4

Block

365.68* 149.04* 99.34* 35.66* 2.11 35.21*

2.45 2.45 2.45 2.45 2.45 2.45

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

− − − − − − − − + + + + + + + +

− − − − + + + + − − − − + + + +

− − + + − − + + − − + + − − + +

− + − + − + − + − + − + − + − +

II I I II I II II I I II II I II I I II

* Significant at p = 0.05.

Equation 3.32 shows the inverse matrix, (XTX)−1. Note the matrix is not quite diagonal with the quadratic entries mutually biasing one another:  1 12    −1 T (X X ) =     

18 18

14 5 32 1 32

    .  1 32   5 32 (3.32)

Table 3.17 presents the results from an analysis of the standard errors of each effect. Dividing the coefficient value by the standard error for each effect gives the t ratio. The critical t ratio is given by tcrit = √Fcrit(p, 1, DFR) ratio. Here we see that a22 exceeds tcrit while a11 does not. This method is not foolproof since a11 and a22 are mutually biased. However, for central composite designs the degree of bias is often slight.

3.14 Blocking Designs having large numbers of experimental runs or experiments run sequentially often cannot be run under homogeneous conditions. This is because during intermittent periods, important factor changes may occur. These include ambient factors, time-based factors such as settling or aging of raw materials, and batch effects. If we presume that such effects will alter the mean response but not dramatically affect other coefficient values, we can eliminate bias in the analysis by a technique known as blocking. In general one should “block whenever possible and randomize the remainder.” For example, suppose we wish to test a burner using a full factorial design in four factors. However, we only


have a limited time to run the tests. Perhaps the burner is only available for two days separated by some intervening gap of days or weeks. During the intervening time many things could change. Perhaps the oil fuel quality changes over time, or unbeknown to us, some ambient conditions are important and these conditions have changed. Anything that is not specifically accounted for in the declared model will accumulate in the error term, inflating it. That is, in a fully randomized experiment, the error term will contain not only random error but also systematic bias of which we are unaware. This kind of bias cannot be measured because we do not even know that it exists. Notwithstanding, our ignorance will not prevent this influential step change from affecting our results. Table 3.18 shows a suggested way of blocking the design. In the first set of tests we run points 2, 3, 5, 8, 9, 12, 14, 15 in randomized order. In the second set of tests we run the remainder, also in a randomized order. Before describing how we arrived at such a blocking arrangement, let us examine a very poor blocking arrangement—namely, Block I = Points 1–7, Block II = Points 8–16. If we use this blocking arrangement, and we fail to randomize runs, we will have committed two sins and completely confounded any block effect with the factor effect, x1. The block effect will be ascribed falsely to x1 because x1 is run at its low level in tests 1–7 and at its high level in runs 8–16. In short, x1 changes when the block changes and the effects are hopelessly entangled. What we desire is a block effect that will be orthogonal to the effects of interest. However, we tacitly learned how to block orthogonally when we learned how to fractionate a design. We

76


Table 3.19

Table 3.20

A 2 Full Factorial Design in Four Blocks with Eight Center Points

t Tests for Factor Effects for Table 19 Data

4

Point

x1

x2

x3

x4

b1

b2

1 2 3 4 5 6

+ − + − 0 0

+ − − + 0 0

+ + − − 0 0

− − + + 0 0

− − − − 0 0

− − − − 0 0

7 8 9 10 11 12

− + + − 0 0

+ − + − 0 0

+ + − − 0 0

+ + − − 0 0

− − − − 0 0

+ + + + 0 0

13 14 15 16 17 18

− + − + 0 0

− − + + 0 0

− + + − 0 0

+ − − + 0 0

+ + + + 0 0

− − − − 0 0

19 20 21 22 23 24

+ + − − 0 0

+ − + − 0 0

+ − − + 0 0

+ − − + 0 0

+ + + + 0 0

+ + + + 0 0

Coefficient

Block

y

I

19.5 16.7 26.4 12.1 17.5 18.1

II

27.2 41.0 7.0 4.5 20.7 19.4

III

16.5 31.9 15.6 21.1 20.5 20.2

IV

35.3 17.5 1.5 30.3 20.2 21.0

should assign our blocks in the same way: use a defining contrast. Since there is only one block, we use the largest possible contrast—1234B where B stands for the block effect. For this reason, the blocking pattern takes on a similar pattern to x5 in the fractional factorial of Table 3.12. We can also add some center points to each block and randomize them along with the rest of the runs in order to assess pure error within each block. The axial and factorial points of a central composite design are also orthogonal to one another, so one can run those designs sequentially and subtract the block effects from the error, thus making the F tests more sensitive. What if we need four blocks instead of two? Then, we would add two blocking factors (b1, b2) in the same way that we would add two additional factors in a fractional factorial. We would then assign the four combinations to a block. For example, (− −) = I, (− +) = II, + −) = III, and (+ +) = IV. Table 3.19 gives an example with defining contrasts being 124b1, 34b2, and 123b1b2. Therefore, 124, 34, and 123 are confounded with block effects. Let us presume that we can tolerate no more than six runs at any time and that there is a possibility


sk

Value

a0* a1* a2* a3* a4* a12* a13 a14 a23 a24* a34 a123 a124 a134 a234 a1234 Block I Block II Block III Block IV

20.07 4.71 −2.84 6.93 5.98 −1.39 0.03 0.01 0.06 0.53 −0.34 0.16 0.18 −0.04 0.06 0.13 −1.90 0.35 0.65 0.90

0.16 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.19 0.34 0.34 0.34 0.19 0.19 0.19 0.48 0.48 0.48 0.48

t 126.1* 24.1* −14.6* 35.6* 30.7* −7.1* 0.2 0.0 0.3 2.7 −1.0 0.5 0.5 −0.2 0.3 0.7 −4.0 0.7 1.4 1.9

* Significant at p = 0.05, tcrit = 2.57.

Table 3.21 ANOVA for Table 3.19 Data Coefficient

DF

SS

MS

F

Fcrit

a1 a2 a3 a4 a12 a13 a14 a23 a24 a34 a123 a124 a134 a234 a1234 Blocks R T

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 3 5 23

354.38 129.39 768.68 572.41 31.08 0.02 0.00 0.05 4.52 0.63 0.13 0.18 0.03 0.05 0.28 9.93 3.04 1874.78

354.38 129.39 768.68 572.41 31.08 0.02 0.00 0.05 4.52 0.63 0.13 0.18 0.03 0.05 0.28 3.31 0.61

582.82 212.80 1264.18 941.39 51.12 0.03 0.00 0.08 7.43 1.04 0.21 0.29 0.05 0.08 0.45 5.44 0.09

6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 6.61 5.41

that significant changes that affect our response may occur between blocks. Table 3.20 gives the factor effects. Tables 3.20 and 3.21 show the associated t and F tests. For these data, lack of fit turns out not to be significant. Therefore, we may use the pooled residual to perform our tests.

4 Fluid Flow Wes Bussman and Joseph Colannino Contents 4.1 Introduction...................................................................................................................................................................... 77 4.2 Gas Properties.................................................................................................................................................................. 78 4.2.1 Density.................................................................................................................................................................. 78 4.2.2 Ratio of Specific Heats and Related Concepts................................................................................................. 78 4.2.3 Viscosity................................................................................................................................................................ 78 4.2.4 Fuel Heating Value.............................................................................................................................................. 78 4.3 Basic Fluid Dynamic Concepts...................................................................................................................................... 79 4.3.1 Definition and Units of Pressure....................................................................................................................... 79 4.3.2 Gauge and Absolute Pressure............................................................................................................................ 79 4.3.3 Draft....................................................................................................................................................................... 79 4.3.4 Compressible Flow through an Orifice............................................................................................................ 80 4.3.5 Pressure Drop in Pipes and Fittings................................................................................................................. 81 4.4 Pressure Measurement Techniques.............................................................................................................................. 82 4.4.1 Manometer............................................................................................................................................................ 82 4.4.1.1 U-Tube Manometer............................................................................................................................... 82 4.4.1.2 Inclined Manometer............................................................................................................................. 82 4.4.2 Bourdon Tube Gauge........................................................................................................................................... 83 4.4.2.1 Design of the Bourdon Tube Gauge................................................................................................... 83 4.4.2.2 Common Failure Mechanisms............................................................................................................ 84 4.4.2.3 Calibration of Pressure Gauges........................................................................................................... 85 4.4.2.4 Selection.................................................................................................................................................. 85 4.4.2.5 Installation............................................................................................................................................. 85 4.5 Differential Producing Flow Meters............................................................................................................................. 86 4.5.1 Orifice Meter......................................................................................................................................................... 86 4.5.2 Venturi Meter....................................................................................................................................................... 89 4.5.3 Turbine Flow Meter............................................................................................................................................. 89 4.5.4 Vortex Flow Meter............................................................................................................................................... 89 4.5.5 Magnetic Flow Meter.......................................................................................................................................... 90 4.5.6 Ultrasonic Flow Meter......................................................................................................................................... 90 4.5.7 Thermal Mass Meter........................................................................................................................................... 91 4.5.8 Positive Displacement Meter.............................................................................................................................. 91 4.5.9 Pitot Tube.............................................................................................................................................................. 92 4.5.10 Annubar................................................................................................................................................................ 94 References................................................................................................................................................................................... 95

4.1 Introduction The hydrocarbon and petrochemical industries employ the flow of many fluids. For example, combustion equipment requires the flow of air and fuel. Then there are the process streams themselves comprising hydrocarbon

liquids and gases flowing under pressure, through pipes and fittings, to their delivery point. Calculation of fluid flow, therefore, is of fundamental importance. The purpose of this chapter is to give the reader a fundamental understanding of some of the fluid dynamic concepts and equipment that are important in testing combustion systems. 77


78


4.2.3 Viscosity

4.2 Gas Properties 4.2.1 Density Density (ρ) is defined as mass (m) divided by volume (V). Accordingly, units of density are g/cm3, kg/m3, lbm/ft3, and so forth. When the volume is macroscopic then, technically, the density is an average density. The general way to define the density at a point is to use the limit at infinitesimal volume.

ρ=

dm ; dV

(4.1)

however, this definition breaks down as we near the molecular level. Therefore, we shall presume that in all cases that matter this may be considered a continuum.

Shear stress (τ) is the force per unit area exerted due to friction of a flowing fluid. If a stationary wall is exposed to a flowing fluid, the velocity at the wall will be zero and the velocity far from the wall will be the free stream velocity. Between these two points a velocity gradient exists. The viscosity (µ) is the constant of proportionality between the velocity gradient (dv/dy) and the shear stress. Figure 4.1 shows the coordinate directions and the variation in velocity from zero at the plate to the free stream velocity, vs. The velocity is flowing in the x direction and the velocity gradient varies in the y direction. The shear stress for ideal flowing gases is given by the following equation and gauges the resistance to establishing a rate of strain (dislocation of the fluid) τ=µ

4.2.2 Ratio of Specific Heats and Related Concepts The specific heat represents the amount of energy required to raise a unit mass one unit in temperature. For gases, the specific heat differs depending on whether the gas is allowed to do work by expanding against an atmosphere (constant pressure definition, cp) or is constrained within a volume (constant volume definition, cv). Examples of units of cp and cv are J/g K or Btu/lbm °F. If we wish to express this on a molar basis we shall use the upper case: that is, Cp or Cv having units of J/mol K or Btu/lbmol °F. The universal gas constant (R) is defined for an ideal gas as follows:

R = PV/nT,

(4.2)

where P is the pressure, V is the volume, n is the number of moles, and T is the absolute temperature.

Cp − Cv = R.

The fuel heating value is the total amount of energy that can be liberated by combustion of a fuel. The heating value may be normalized by mass, moles, or volume; accordingly, it will have units of kJ/kg, J/mol, or J/Nm3. Customary units include Btu/lbm, Btu/lbmol, and Btu/scf. In the case of volume, some standard temperature and pressure must be chosen, such as 25°C and 1 atm. Two examples of unit volumes referenced to such conditions are the normal meter cubed (Nm3) and the standard cubic foot (scf). For hydrocarbons, combustion always produces water and carbon dioxide according to the following equation: CxHy + (x + y/4)O2 = xCO2 + y/2 H2O.

k=

c p Cp = . c v Cv

vs y

(4.3)


(4.4)

(4.6)

For combustion reactions producing water, we may consider two different standard states for calculating the heating value: the standard state may be taken to be either a liquid or vapor at the standard temperature and pressure. If the standard state considers water as a vapor, the heating value so determined is called the

x

It also turns out that the dimensionless ratio of heat capacities (k) is an important thermodynamic parameter:

(4.5)

4.2.4 Fuel Heating Value

By ideal gas we mean a gas whose atomic and molecular volumes and attraction entities exert no significant influence upon the bulk volume or pressure. All real gases deviate from ideality at very low temperatures and very high pressures. In many cases (and especially so for combustion) the ideal gas law is sufficiently accurate for engineering purposes. Remarkably, the universal gas constant and heat capacities are related:

dy . dx

Figure 4.1 Fluid flow shear stress.

79

Fluid Flow

lower heating value (LHV) or net heating value. If the standard state considers water as a liquid, then the heating value so determined is called the higher heating value (HHV) or gross heating value. It is important to understand that this determination is set by convention and has nothing to do with the actual state of the water. The difference between the higher and LHVs is the standard enthalpy of vaporization for water. Some combustion produces no water (e.g., combustion of CO) and therefore such higher and LHVs are identical.

Bernoulli’s equation, as written above, considers no losses due to friction or microscopic recirculation. In order to calculate the total draft for a furnace of a given height (h) and temperature, one may equate the potential energy of the gas owing to its mass in a gravitational field with that of the ambient fluid outside the furnace; that is, air in the usual case. This gives ΔP/Δρ = gh or ΔP = Δρgh where Δρ is the difference in densities between the two fluids. Noting that Δρ = PM/ RΔT where Δρ is the density, P is the ambient pressure, M is the molecular weight, R is the universal gas constant, and ∆T is the absolute temperature difference, we may recast the equation as

4.3 Basic Fluid Dynamic Concepts

∆P =

4.3.1 Definition and Units of Pressure Pressure (P) is defined as a normal force per unit area. For example, the pressure of a standard atmosphere at sea level is ~14.7 lbf/in2 (psi). Other pressure units include kilopascals (kPa), bar (101.3 kPa), and atmospheres (atm).

Problem statement: What is the maximum draft pressure at the floor of a 10-m furnace operating at 800°C? Presume that the control system adjusts the pressure at the top of the furnace to a draft of 3 mm w.c. via a stack damper. Solution: From Equation 2.110, at 1 atm of pressure (101 kPa) we have   g  101[kPa ]29    mol   ∆P =   3  8.314  m ⋅ Pa       mol ⋅ K   1 1   m   (800 + 273)[K ] − ( 25 + 273)[K ] 9.8  s 2  10 [m]  

4.3.3 Draft

∆P v 2 = . 2 ρ


(4.7)

(4.8)

Example 4.1: Calculation of Draft Pressure

Although the pressure at sea level is 14.7 psi, pressure gauges are calibrated to read 0 at 1 atmospheric pressure. Therefore, when one measures pressure in an automobile tire, say 32 psig, one is measuring the pressure above atmospheric. In this example, the total pressure (also called the absolute pressure) in the tire is actually 46.7 psia because the pressure at sea level is 14.7 psia. In order to distinguish between gauge (g) and absolute (a) pressure, the respective letter is appended to the psi designation.

  ,

where T∞ is the absolute ambient temperature and Tf is the absolute furnace temperature.

4.3.2 Gauge and Absolute Pressure

Draft is a differential pressure between the outside atmosphere and the internal pressure in a furnace. Differential pressure such as units of draft are sometimes referenced as psid. However, draft pressures are ordinarily so low that smaller units are needed. Two common units are mm w.c. or in w.c. (also w.c. or i.w.c.); that is, millimeters, water column or inches, water column, respectively. For example, in order to displace a fluid from a glass to one’s mouth three inches above the liquid surface via a straw, one must cause a differential pressure of 3 in. w.c. Draft pressures are often smaller. For example, a typical natural draft refinery furnace operates anywhere from 0.25 in. w.c. to 0.50 in. w.c. This seemingly small pressure difference can cause great volumes of air flow. Gas pressure and velocity are related by Bernoulli’s equation:

PMgh  1 1 − R  T∞ T f

= −54.6[Pa ][mm w.c.]

The negative sign reminds us that the pressure is below atmospheric. Adding the −3 mm w.c. pressure (draft) we have at the bridgewall gives a total of −8.6 mm w.c. pressure at the floor (or 8.6 mm w.c. draft).

We can also write an equation for the mass flow (m ˙ ) as a function of velocity and density:

 = ρAv , m

(4.9)

where A is the flow area. Coupling these two equations one arrives at mass flow as a function of draft pres = A 2ρ ∆P . For purposes of the sure through an area: m calculation, the absolute value of the pressure is used with the understanding that the flow direction is always defined as flow from the higher to the lower pressure. However, as previously noted, Bernoulli’s equation does not consider friction loss. We can account for friction by premultiplying by a loss coefficient, Co. In that case, the

80


mass flow rate as a function of pressure and density becomes  = Co A 2ρ ∆P . m

(4.10)

For British units, one should note that 32.174 lbm.ft/lbf−s2 = 1 = gc, where lbm is the pound mass unit and lbf is the pound force unit. Force and mass are related by Newton’s second law, F = mg, (4.11) where F is the force, m is the mass, and g is the gravitational acceleration (9.8 m/s2 or 32.174 ft/s2). By definition, 1 lbm at 1 g of gravity exerts 1 lbf. Therefore, 1 lbf = 32.174 lbm ft/s2. Since gc = 1 by definition, the incompressible flow equation is often written as  = Co A 2 gcρ ∆P , m

The incompressible flow equation used here is accurate for natural draft devices where the resulting air velocity is much lower than the speed of sound. However, it is not a good assumption for fuel flow from a fuel orifice as fuel pressures are normally 25 psig or so and the exit velocity is sonic. For these cases, we must make use of a more inclusive form of the Bernoulli equation.

(4.12)

when using British units. This would not be necessary except that engineers before Newton did not recognize the difference between mass and force and developed an inconsistent set of units. SI units, or any other consistent set of units, do not require the use of gc . Example 4.2: Calculation of Mass Flow Through Hole in a Furnace Problem Statement: A furnace has a leak equating to a total flow area of 4 in2 at an elevation in the heater corresponding to −0.25 in. w.c. of draft. Presuming an orifice coefficient of 0.80, what is the mass flow of air that this represents? Presume that the ambient air density is 0.075 lbm/ft3.  = Co A 2gc ρ| ∆P | with Solution: We use the equation m the following information: Co = 0.80 A = 4 in2 ρ = 0.075 lbm/ft3 ∆P = −0.25″ w.c.

4.3.4 Compressible Flow through an Orifice We shall find it convenient to define a pressure ratio (ro) that is a function of the absolute pressure upstream of the orifice (Po) and the absolute pressure surrounding the orifice into which the fluid is exhausting (P∞):

ro =

Po . P∞

(4.13)

As the pressure increases, the velocity through the rifice also increases until it reaches a natural limit o known as the speed of sound. This limit occurs at a critical pressure ratio (rc). This critical ratio is given by the following equation:

( )

k+1 rc = 2

k k −1

.

Considering compressible flow through an orifice the following equation gives the mass flow. An exact derivation is available at http://www.combustion-modeling.com

( )

1− k   2k  k  r 1 − o   M k−1    2 C A P β o o o 2  RT  roγ − β 4  m  =  2k  M k+1  Co Ao Poβ 2  RT k + 1 k −k 1  − β4 2 

( )

( )

( )

Therefore, we have

lbm lbm ft   = (0.80 )( 4 )[in2 ] 2(32.174 )  m 0.075  3  0.25[in w.c.].  ft   lbf s 2 

We also note that there are 144 in2 per 1 ft 2 and 27.69 in w.c. per 1 lbf/in2. With these conversions we have lbf 1 2  2  ft lbm 1 in2 ft lbm [ ]  in   0.075   0.25[in w.c.]  = (0.80 )( 4 )[in2 ] 144  2  m 2(32.174 )   ft 3  27.69[in w.c.] 144 [in2 ]  ft   lbf s 2  = 0.056 lbm/s (or 200.4 lbm/h).


(4.14)

ro < rc , (4.15) ro ≥ rc

81

Fluid Flow

where β is the ratio of the orifice diameter to the upstream pipe or duct diameter, Ao is the area of the orifice, and T is the upstream temperature. For monoatomic gases such as helium or argon, k ≈ 5/3. For diatomic gases such as air, oxygen, or nitrogen, k ≈ 7/5. For polyatomic gases k approaches 1 as the number of atoms increases to infinity; however, for common polyatomic gases having 3 to 5 atoms, 9/7 

Industrial combustion teseting

Industrial Burners Handbook (Industrial Combustion)

The John Zink Combustion Handbook (Industrial Combustion)

HEAT TRANSFER IN INDUSTRIAL COMBUSTION

Industrial Combustion Pollution and Control

HEAT TRANSFER IN INDUSTRIAL COMBUSTION

Heat Transfer in Industrial Combustion

Industrial Combustion Pollution and Control

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Spontaneous Combustion

Combustion Phenomena

Bear Combustion

Bear Combustion

Industrial Combustion Pollution and Control (Environmental Science & Pollution)

Turbulent Combustion

Combustion Theory

Combustion Celular

Industrial combustion teseting

Industrial Burners Handbook (Industrial Combustion)

The John Zink Combustion Handbook (Industrial Combustion)

HEAT TRANSFER IN INDUSTRIAL COMBUSTION

Industrial Combustion Pollution and Control

HEAT TRANSFER IN INDUSTRIAL COMBUSTION

Heat Transfer in Industrial Combustion

Industrial Combustion Pollution and Control

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Combustion

Spontaneous Combustion

Combustion Phenomena

Bear Combustion

Bear Combustion

Industrial Combustion Pollution and Control (Environmental Science & Pollution)

Turbulent Combustion

Combustion Theory

Combustion Celular

Recommend Documents