Professional Documents
Culture Documents
Editor-in-chief
D.Yogi Goswami
ADVANCES IN
SOLAR ENERGY
An Annual Review of
Research and Development
volume
17
prelims
2/7/07
3:51 PM
Page i
prelims
2/7/07
3:51 PM
Page ii
Editor-in-Chief
D. Yogi Goswami
Clean Energy Research Center, University of South Florida, Tampa, FL
Associate Editors
Viacheslav M. Andreev
Ioff Physico-Technical Institute, St Petersburg, Russia
J. Douglas Balcomb
National Renewable Energy Laboratory, Golden, CO
Adolf Goetzberger
Fraunhofer Institut Solare Energiesysteme, Freiburg, Germany
Yoshihiro Hamakawa
Ritsumeikan University, Kusatsu Shiga, Japan
Lu Weide
China Rural Energy Industry Association, Beijing, China
Antonio Luque
Universidad Politcnica de Madrid, Madrid, Spain
Gerald R. Nix
National Renewable Energy Laboratory, Golden, CO
Morton Prince
Melrose Park, PA
Aldo Steinfeld
Swiss Federal Institute of Technology, Zurich, Switzerland
Gunnar Svedberg
Mid Sweden University, Sundsvall, Sweden
Steven Szokolay
University of Queensland, Kenmore, Australia
Lorin Vant-Hull
University of Houston, Houston, TX
Nejat Veziroglu
University of Miami, Coral Gables, FL
Carl Weinberg
Walnut Creek, CA
Roland Winston
University of Chicago, Chicago, IL
Emeritus Editor-in-Chief
Karl W. Ber
University of Delaware, Newark, DE
prelims
2/7/07
3:51 PM
Page iii
London Sterling, VA
American Solar Energy Society, Inc., Boulder, CO
prelims
2/7/07
3:51 PM
Page iv
0731-8618
978-1-84407-314-6
prelims
2/7/07
3:51 PM
Page v
FOREWORD
Global climate change continues to be a hot topic on the world stage. However, we have
now moved away from the debate of whether global warming exists to what strategies
are needed to combat the challenge. Another ongoing debate concerns the peaking and
eventual decline of world oil production and how this will affect other conventional energy
resources including coal, natural gas and uranium. It is becoming increasingly clear that
whether we are concerned about global climate change or the peaking and eventual
decline of conventional resources, we must ramp up the use of renewable energy
resources worldwide. Indeed, Schindler and Zittel argue in their chapter in this volume of
Advances in Solar Energy that by 2050 as much as 50 per cent of our energy will have to
come from renewable energy. In our continuing series of country- and region-specific
articles we focus on the Middle East and North Africa (MENA) region Alnaser, Treib and
Knies show that by 2050 as much as 50 per cent of the electrical power for the MENA
countries may come from concentrating solar power (CSP).
At present, commercial deployment of renewable energy technologies is increasing at
rates of 2040 per cent per year. However, in order to increase the deployment of
renewable energy technologies to the levels foreseen by Schindler and Zittel, we need
continued research and development to raise the efficiency and reduce the costs of these
technologies. The chapters in this volume inform us on progress in this direction, for
example advances in quantum dot solar cells and organic solar cells. Other topics in this
volume include such applications of solar energy as solar-hydrogen production, solar
photocatalytic detoxification of water, solar drying and solar industrial process heat.
In the first chapter, Schindler and Zittel question the assumptions made by the
International Energy Agency (IEA) in forecasting the future availability of energy resources,
and present an alternative world energy outlook and a possible path towards a sustainable
future. According to the authors the world will experience the peaking of oil first, followed
by natural gas. They show that the resulting energy supply gap cannot be filled by a rising
share of nuclear energy. This gap could be filled partially by coal; however, without CO2
sequestration this would lead to emissions at unacceptable levels. Their analysis shows
that a transition in our energy resource use to renewable energy will occur. However, they
question whether there is enough time left for the transition to occur smoothly. They
recommend that the earlier we start with the transition, the better we will be able to
manage it.
Chapters 2 and 3 deal with the new developments in photovoltaic cells. Chapter 2
deals with quantum well solar cells and the progress made in this area over the last 15
years. The author, Ekins-Daukes, writes, The ability to adjust the absorption profile of a
bulk semiconductor through the inclusion of quantum wells gives rise to a number of
practical, near-term applications for quantum solar cells. Some of these applications
include multi-junction solar cells and thermophotovoltaics. The author concludes that
even though to date only the band gap has been engineered with the application of
quantum wells, it is likely that in the future it will be possible to control other properties,
such as optical transitions and phonon modes in order to produce highly efficient devices.
Chapter 3 describes the recent progress in organic photovoltaic materials and devices.
Organic and polymeric materials have the potential to reduce the cost of solar cells
prelims
2/7/07
vi
3:51 PM
Page vi
making them cost competitive. At present the conversion efficiencies of organic solar
cells are relatively small (< 6 per cent), which can be attributed to a number of reasons,
such as, photon, exciton and carrier losses, poor material morphologies, and unoptimized
structures. However, there is plenty of room for improvement. The paper outlines the
approaches being followed in improving the performances.
Chapter 4 describes the progress in thermal and material characterization of
immersed heat exchangers used in solar domestic water heating systems. The authors
have made recommendations based on published literature and their own work over the
last 20 years. As manufacturers consider the use of polymers for heat exchangers, their
durability and reliability must also be considered in addition to thermal performance. The
authors also point out areas for additional research.
Chapter 5 deals with solar photocatalytic detoxification, in which there has been
tremendous research and development over the past ten years, although it is a relatively
recent environmental application of solar radiation. In this paper, the authors review the
scientific progress and applications of solar photocatalysis. Even though the process is
very attractive in remediating water pollution, the quantum efficiency is very low. The
authors describe the research and development (R&D) approaches to increase the
quantum yield, such as modifying the catalyst structure and composition and finding new
catalysts with band gaps that better overlap the solar spectrum. The authors also describe
the progress in using photo-Fenton treatment, and combined photocatalytic-biological
treatment.
At present there is considerable interest in the potential use of hydrogen for
transportation, to replace the eventually depleting supplies of oil. However, for hydrogen
to provide a viable solution, it must be produced economically and from a clean source,
such as solar energy. Solar photo-electrochemical generation of hydrogen is one such
method that has a viable potential. However, this method needs some fundamental R&D.
In Chapter 6, the authors systemically describe the solid-state chemistry-related issues
that must be considered in order to improve the viability of this method. The authors
explain the parameters such as band gap and Fermi level and the concepts of modifying
them.
Chapter 7 reviews the recent progress in using solar heat for industrial processes,
with specific examples from Europe. The authors review the technology and its potential
for use in low (< 60C), medium (60150C) and medium-high (150250C) temperature
industrial applications. They present case studies for many industries in Spain and
Portugal, arguing that as much as 7 per cent of total final energy demand in southern
European countries may be fulfilled by providing solar process heat at temperatures
below 250C. The authors state that the present costs of solar thermal systems range
from 250 to 1000/kW of thermal power, leading to average thermal energy costs in
southern Europe of 25/kWh for low temperature applications and 515/kWh for
medium temperature applications. However, according to the authors, these costs can be
reduced by 50 per cent by 2010 by mass production, reducing operation and maintenance
expenses and improving collector efficiency and collector design.
The final paper (Chapter 9) of this volume is focused on the Middle East and North
Africa (MENA) as a part of our series of country-specific articles. This article, however,
prelims
2/7/07
3:51 PM
Page vii
FOREWORD
vii
D. Yogi Goswami
Editor-in-Chief
prelims
2/7/07
3:51 PM
Page viii
prelims
2/7/07
3:51 PM
Page ix
Contents
Foreword
List of figures and tables
About the authors
1
v
xii
xxi
1
1
7
28
38
41
42
43
45
N. J. Ekins-Daukes
2.1 History
2.2 Quantum Well Electronic Structure
2.3 Basic Operation of the P-I-N Quantum Well Solar Cell
2.4 Near-term Applications for Quantum Well Solar Cells
2.5 Efficiency Limits
2.6 Conclusion
Acknowledgements
References
45
46
49
61
64
66
66
67
74
Sam-Shajing Sun
3.1 Introduction
3.2 Organic versus Inorganic Semiconductors
3.3 Organic/Polymeric Solar Cell Developments
3.4 Organic Solar Cell Fabrications
3.5 Organic Solar Cell Optimizations
3.6 Conclusions and Future Perspectives
Acknowledgements
References
74
75
77
84
87
94
95
95
99
99
102
prelims
2/7/07
3:51 PM
Page x
Photocatalytic Detoxification of
Water with Solar Energy
112
118
124
125
125
130
Introduction
Solar Collectors for Photochemistry
Fundamental Parameters in Solar Photocatalysis
Factors Affecting Solar Photocatalysis
Solar UV Photocatalytic Degradation of Contaminants
Evaluation of Solar UV Radiation
Installed Solar Photocatalytic Treatment Plants
Photocatalytic Detoxification of Water with
Solar Energy: Outlook for the Future
Acknowledgements
References
Solar-Hydrogen: A Solid-State
Chemistry Perspective
130
132
138
147
149
151
154
160
163
163
169
170
174
175
178
180
190
193
194
203
204
204
205
207
207
209
210
211
211
211
prelims
2/7/07
3:51 PM
Page xi
CONTENTS
xi
216
216
217
220
238
247
257
258
261
262
264
265
274
282
284
291
295
302
302
prelims
2/7/07
3:51 PM
Page xii
5
7
10
11
15
16
17
19
20
21
22
23
24
25
26
27
28
29
30
32
34
35
36
43
prelims
2/7/07
3:51 PM
Page xiii
xiii
43
44
47
48
50
51
52
53
54
56
57
58
59
62
75
76
prelims
2/7/07
xiv
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
3.11
3.12
3.13
3.14
3.15
3.16
4.1
4.2
4.3
4.4
3:51 PM
Page xiv
78
79
79
80
80
82
83
85
85
89
89
90
92
93
101
102
106
109
prelims
2/7/07
3:51 PM
Page xv
4.5
4.6
4.7
4.8
4.9
4.10
4.11
5.1
5.2
5.3
5.4
5.5
5.6
5.7
5.8
5.9
5.10
5.11
5.12
xv
110
111
115
116
117
121
122
133
135
138
139
140
142
143
143
153
154
157
159
prelims
2/7/07
xvi
5.13
5.14
6.1
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
6.10
6.11
6.12
6.13
6.14
6.15
6.16
3:51 PM
Page xvi
160
161
170
172
176
176
177
182
184
188
192
192
199
201
203
205
206
207
prelims
2/7/07
3:51 PM
Page xvii
6.17
7.1
7.2
7.3
7.4
7.5
7.6
7.7
7.8
7.9
7.10
7.11
7.12
7.13
7.14
7.15
7.16
7.17
7.18
7.19
7.20
7.21
7.22
7.23
7.24
7.25
7.26
7.27
xvii
210
219
219
220
221
222
223
224
225
226
226
228
229
238
240
241
242
244
244
247
248
249
250
250
252
252
253
254
prelims
2/7/07
xviii
7.28
7.29
7.30
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12
8.13
8.14
8.15
8.16
8.17
8.18
8.19
8.20
3:51 PM
Page xviii
255
256
256
262
268
269
270
271
272
273
274
275
278
279
280
286
287
288
289
290
291
292
293
prelims
2/7/07
3:51 PM
Page xix
8.21
8.22
8.23
8.24
8.25
8.26
8.27
8.28
8.29
8.30
xix
294
294
299
299
300
300
301
301
302
303
TABLES
1.1
1.2
2.1
4.1
6.1
7.1
7.2
7.3
7.4
7.5
32
35
63
120
202
218
220
227
231
243
prelims
2/7/07
xx
7.6
7.7
7.8
7.9
7.10
7.11
7.12
8.1
8.2
8.3
8.4
8.5
8.6
8.7
8.8
8.9
8.10
8.11
8.12
3:51 PM
Page xx
245
247
249
251
253
255
257
267
267
276
278
279
281
282
283
283
284
296
297
prelims
2/7/07
10:28 PM
Page xxi
prelims
2/7/07
xxii
3:51 PM
Page xxii
as both the head of the Solar Chemistry Department of Plataforma Solar de Almera (PSA),
and the Spanish National Representative in the Task II group of the International Energy
Agency SolarPACES (Solar Power and Chemical Energy Systems). Dr. Blanco has been
involved in many national and contractual research and development projects, as well as
work for the European Union related to the development of solar technologies as applied
to waste water treatment processes and techniques. Dr. Blanco is author of 1 book and
co-author of 3 books and 8 chapters in different books, some 40 publications in indexed
international journals and more than 90 contributions to international congress and
symposiums. He has been honoured with the Jurys Grand Prix at the European Grand Prix
for Innovation Awards, 2004, in Monaco (www.european-grandprix.com/index_en.htm).
Maria Joo Carvalho has worked in solar thermal energy since the 1980s and is
currently the senior researcher in the Renewable Energy Department of the National
Institute for Engineering, Technology and Innovation (INETI), a state laboratory in
Portugal. She is the director of the Solar Collector Testing Laboratory of INETI. Dr.
Carvalho is the author of 19 papers published in international technical journals and 38
papers published in conference proceedings. She has given invited lectures on
renewable energy for universities and also seminars on the subject of solar thermal
energy with special emphasis on collector and system testing. She is a member of the
Portuguese Section of ISES.
Jane H. Davidson has been a professor in the Department of Mechanical Engineering at
the University of Minnesota since 1993. Before coming to Minnesota, she was a faculty
member at the University of Delaware and Colorado State University. Dr. Davidson is past
editor-in-chief of the Journal of Solar Energy Engineering. She was the 2004 recipient of
the ASME John I. Yellott Award for outstanding research in solar energy and is a fellow of
the American Society of Mechanical Engineering and the American Solar Energy Society
(ASES). Her publications include more than 150 papers in journals and conferences and 3
book chapters. She was the 2005 recipient of the University of Minnesotas Distinguished
Woman Scholar Award in Science and Engineering. Her research interests include solar
thermal systems and energy efficient buildings.
N. J. Ekins-Daukes is a lecturer at the University of Sydneys School of Physics, having
earned his Ph.D. from Imperial College, London in 2000 for his work on strain-balanced
quantum well solar cells. He later held a fellowship from the Japan Society for the
Promotion of Science (JSPS) at the Toyota Technological Institute in Nagoya, Japan. Dr.
Ekins-Daukes, present research involves the application of nano-structures to high
efficiency photovoltaic devices as well as contributing towards projects on the radiation
resistance of III-V solar cells and the application of III-V solar cells in concentrator systems
and the application of nano-structures to photovoltaic devices. Dr. Ekins-Daukes is a
member of the Institute of Electrical and Electronics Engineers (IEEE). He has published
papers on a range of topics covering quantum well solar cells, luminescent up-conversion,
radiation resistance and photovoltaic concentrator systems.
Joo A. Farinha Mendes is a mechanical engineer (thermodynamics), currently working
as a senior researcher and is the Head of the Solar Energy Unit in the Renewable Energy
prelims
2/7/07
3:51 PM
Page xxiii
xxiii
Department (DER) at the National Institute for Engineering, Technology and Innovation
(INETI), a state laboratory in Portugal. His main research activity has been devoted to solar
thermal applications, serving as national representative for several European projects,
leading to a number of papers on non-imaging optics, phase change heat transfer and
solar system design published in international technical journals and conference
proceedings. He is the Portuguese representative for the International Energy Agencys
Executive Committee on Solar Heating and Cooling Implementing Agreement and has
participated in the standardization work of solar collectors at the national, European and
international levels. He has given invited lectures on solar thermal energy for universities
and participated in numerous seminars in Portugal and other countries in Europe.
Pilar Fernndez Ibez earned her Ph.D. in applied physics from the University of
Granada in 2004, following specialization in environmental sciences and applied physics
at the University of Almera, Spain. Dr. Fernndez has ten years of work experience in
various research sectors, participating in ten European Union and two national research
and development projects related to disinfection and wastewater treatment with the help
of solar technologies development. She is co-author of three books and six chapters in
other collaborative books. She holds two patents, and is co-author of 30 international
technical journal articles and 50 contributions to different international congresses and
symposiums. She has tutored the scientific work of students on solar disinfection and
detoxification of wastewater. Dr. Fernndez has also been Invited Guest Associate Editor
for special issues of the indexed journals Catalysis Today and Solar Energy.
Lorraine F. Francis is a professor of chemical engineering and materials science at the
University of Minnesota. She joined the University in 1990 and currently serves as the
Director of Undergraduate Studies in Materials Science and Engineering. Dr. Francis
received the National Science Foundation (NSF) Young Investigator Award (19931998)
and the DuPont Young Professor Grant (19941997). Her research interests include
polymer and ceramic coatings, processing-microstructure-property relationships in
coatings and composites, interfacial studies and biomaterials. Dr. Francis is a member of
the American Ceramic Society (ACerS), the Materials Research Society (MRS) and ASM
International.
Wolfgang Gernjak earned his Ph.D. in land and water management from the University of
Natural Resources and Applied Life Sciences (BOKU) in Vienna, following specialization in
analytical and physical chemistry at the Vienna University of Technology. He has been
involved in five European Union research and development projects connected with
the application of solar photocatalysis and solar photo-Fenton for the detoxification of
industrial wastewater and disinfection of drinking water. He is author of 20 articles
in indexed international journals and 33 contributions to international congress and
symposiums.
Klaus Hennecke earned his Doctor of Engineering degree in aerospace technologies in
1978 at the University of the Federal German Forces, Munich, Germany. In 1989 he
became a manager of a series of research projects in the Solar Research Division of the
Institute of Technical Thermodynamics (ITT) within the German Aerospace Center (DLR) in
prelims
2/7/07
xxiv
3:51 PM
Page xxiv
Cologne, Germany. His research is particularly involved with the field of direct steam
generation in parabolic trough collectors and solar process heat applications. As head of
the applications subdivision Dr. Hennecke is currently responsible for the transfer of
research and development results into the market. Engaged in international cooperations
under the umbrella of the International Energy Agency (IEA), implementing the Solar
Power and Chemical Energy Systems (SolarPACES) agreement, Dr. Hennecke leads the
subtask System Integration and Demonstration of the new IEA SHC Task 33/SolarPACES
Task 4 on Solar Heat for Industrial Processes (SHIP).
Gerhard Knies holds a Ph.D. in physics and has worked in the fields of high energy
physics and elementary particle research at Deutsches Elektronen Synchrotron (DESY)
Hamburg, Germany; the European Centre for Nuclear Research (CERN) in Geneva,
Switzerland; and the Stanford Linear Accelerator Center (SLAC) in both Stanford
and Berkeley, California, US. Dr. Knies developed a road map towards energy, water
and climate security by use of solar energy. In 1981 he co-founded the Hamburg Scientists
Peace Initiative, a disarmament campaign against the threatening EastWest arms race,
and made studies on the vulnerability of industrial civilization. In 1995 he founded
the Hamburg Climate Protection Foundation, which in 2003 allied with other groups
to form the Trans-Mediterranean Renewable Energy Cooperation movement for
development, climate stabilization and good neighbourhoods. Dr. Knies has been retired
since 2001.
Dirk Krueger earned his B.Sc. in energy and environmental engineering at the University
of Applied Sciences in AachenJuelich, Germany. Since 1997 he has specialized in
investigating parabolic trough collector systems for process heat applications.
Dr. Kruegers major research fields are collector testing, simulations for output forecasts
and project assistance for companies engaged in parabolic trough technology.
Sixto Malato Rodrguez is a senior researcher of the Centro de Investigaciones
Energticas, Medioambientales y Tecnolgicas (CIEMAT) at Plataforma Solar de Almera
(PSA), Spain, and has 18 years of experience in various research sectors. Since 1990 his
work at the PSA-CIEMAT has involved all the European Union research and development
projects linked to the solar detoxification of water and solar wastewater treatment
technologies. He is author of 1 book and co-author of 5 books as well as 21 chapters in
others. Dr. Malato has also co-authored more than 80 publications in indexed international
journals, 21 articles in technical journals and more than 140 contributions to 68 different
international congresses and symposiums and holds 4 patents. In 2004 he received the
Jurys Grand Prix at the European Grand Prix for Innovation Awards in Monaco
(www.european-grandprix.com/index_en.htm).
Manuel I. Maldonado Rubio earned his Ph.D. from the University of Almera
in 2001, following specialization in environmental sciences and chemistry at the University
of Granada and the Instituto de Investigaciones Ecolgicas in Mlaga, Spain.
Dr. Maldonado has worked for the the Centro de Investigaciones Energticas,
Medioambientales y Tecnolgicas (CIEMAT) at Plataforma Solar de Almera, Spain, since
2002, and has been involved in seven European Union research projects, three national
prelims
2/7/07
3:51 PM
Page xxv
xxv
research projects and three research and development contracts related to the
development of solar technologies applied to wastewater treatment. He has authored 29
peer-reviewed articles for international publications, co-authored 1 book and 5 chapters in
others, and given presentations at 52 international and national congresses.
Susan C. Mantell is a professor of mechanical engineering at the University of Minnesota.
Dr. Mantell has received several awards for excellence in teaching and research including
a National Science Foundation (NSF) Young Investigator Award (19941999) and a
McKnight Land-Grant Professorship (19951997). She is a leading expert in the field of
composite materials and manufacturing and most recently has been working on
development of polymer heat exchangers for solar energy applications and for use in the
transportation industry. Dr. Mantell has served as an associate editor of the Journal of
Composite Materials since 2000 and the Journal of Materials Processing and
Manufacturing Science from 1998 to 2003. She has published over 75 papers and 2 book
chapters on composite materials and polymer processing.
Janusz Nowotny received his Ph.D. in solid-state chemistry from the Institute of Physical
Chemistry, Polish Academy of Science, in 1967, and following specialized research in
diffusion in nonstoichiometric compounds earned a D.Sc. in materials science from the
Academy of Mining and Metallurgy, Cracow, Poland in 1974. Dr. Nowotny is Director of
the Centre for Materials Research in Energy Conversion, University of New South Wales,
Sydney, Australia. He has been a visiting professor at several universities in France,
Germany and Japan. Dr. Nowotny s expertise includes photo-electrochemistry, solid-state
electrochemistry, defect chemistry and the science of materials interfaces. His research
includes defect chemistry of oxide semiconductors, diffusion in ionic solids, segregation
in nonstoichiometric compounds and photo-electrochemistry. He has published over 400
refereed papers, and edited 17 books.
Jrg Schindler is a business economist and since 1984 has worked for the LudwigBlkow-Systemtechnik GmbH, becoming its managing director in 1992. His major
activities include the technical and economic analyses of market introduction of
photovoltaics and other renewable energy sources, clean drive systems for road
transport, clean fuels produced from renewable energy sources (like hydrogen), and the
future availability of fossil fuels. Mr. Schindler is currently a member of the board of
e5 The European Business Council for a Sustainable Energy Future and from 1999 to
2003 was a member of the Enquete Commission of the Bavarian House of
Representatives entitled New Energy for the New Millennium. Enquete Commissions
serve the national government by collecting scientific state-of-the-art information and then
transmitting it through publicly discussed booklets. He is also a member of the board of
the Global Challenges Network.
Hans Schweiger is director of the thermo-energetical systems and renewable energies
company energyXperts.BCN, Barcelona, Spain. Since 1991, Dr. Schweiger has been active
in the field of solar thermal energy. He has worked in collaboration with several national
and international research projects in the fields of solar thermal energy (solar collector
development, transparent insulation, building simulation and absorption cooling) and
prelims
2/7/07
xxvi
3:51 PM
Page xxvi
numerical methods for heat transfer and fluid dynamics (software development for CFD
and radiation heat transfer). He served as project coordinator of the European Union
research project on solar industrial process heat, POSHIP (NNE5-1999-0308).
Dr. Schweiger also teaches masters and post-graduate courses on solar thermal energy
at the Catedra UNESCO (Polytechnical University of Catalonia). He was co-founder of the
engineering cooperative Aiguasol Engineering (Barcelona, Spain), where he was active
from 1999 to 2005. He earned his degrees from the Politechnical University of Catalonia
(UPC, Terrassa, Barcelona, Spain), specializing in the optimization of solar thermal
absorber elements with transparent insulation.
Leigh Sheppard is a postdoctoral fellow at the Centre for Materials Research in Energy
Conversion, School of Materials Science and Engineering, University of New South Wales,
Sydney, Australia. In 2006 he received the Deutsche Akademische Austausch Dients
award of the Hahn-Meitner Institute to work with Professor H. Tributsch.
Dr. Sheppards research interests include photo-electrochemistry of TiO2. He has
published 16 refereed papers and 2 book chapters.
Charles C. Sorrell is Professor of Ceramic Engineering, School of Materials Science
Engineering, University of New South Wales, having earned his Ph.D. in ceramic
engineering at the University of New South Wales in 1987. His main research interest areas
include ceramic processing, phase equilibria, microstructure and performance. Professor
Sorrells other research activities are focused on titania-based materials for
environmentally friendly applications, bioceramics and microwave heating, glass-ceramics,
refractories, high temperature superconductors, graphite, silicon nitride, zirconia, and
gemstone heat treatment. Professor Sorrell is a fellow of the Australian Institute of Energy
(AIE) and the Institution of Engineers/Australia (IEAUST). He has been editor of the Journal
of the Australasian Ceramic Society, Advances in Applied Ceramics, Metals and Materials
International, and Ceramics International. Professor Sorrell has published over 350 papers
and supervised 75 B.Sc. projects, as well as 35 postgraduate students.
Sam-Shajing Sun obtained his Ph.D. degree in polymer/materials chemistry from the
University of Southern California in 1996. After a postdoctoral appointment at the Loker
Hydrocarbon Research Institute, Dr. Sun joined the Norfolk State University in early 1998.
He is currently leading an organic and polymeric materials research programme at the
Center for Materials Research (CMR) at Norfolk State University. Dr. Sun is also directing
a Center for Research and Education in Advanced Materials sponsored by NASA. Dr.
Suns research interests include the moulding, design, synthesis, processing and
characterization of novel organic and polymeric electronic and photonic materials.
Current research projects of his lab focus on the development of novel organic/polymeric
photovoltaic materials for potential plastic solar cells and photodetector applications.
Dr. Sun has published more then 50 refereed papers and presented more then 40 papers
and lectures relevant to organic/polymeric optoelectronic materials at international
scientific communities and academic/governmental organizations. He has recently
edited a new CRC Press book titled Organic Photovoltaics: Mechanisms, Materials and
Devices, a comprehensive book covering the state of the art of the field. Dr. Sun is a
member and technical referee of a number of major scientific organizations.
prelims
2/7/07
3:51 PM
Page xxvii
xxvii
Franz Trieb has worked in the field of renewable energies since 1983. After the
implementation of hydrogen storage for an autonomous renewable energy system at the
University of Oldenburg, Germany, he obtained specialized training in renewable energy
at the National University of Tacna, Peru. Since 1994, Dr. Trieb has worked as project
manager at the Institute of Technical Thermodynamics of the German Aerospace Center
(DLR), working on solar energy resource assessment by satellite remote sensing, market
strategies for concentrating solar power and renewable energy scenarios.
Two of Dr. Triebs recent studies look at the future demand and the sustainable supply
of electricity using the potential of renewable energy sources in the countries of the
Mediterranean Region and Europe (www.dlr.de/tt/med-csp and www.dlr.de/tt/trans-csp).
Werner Zittel obtained his Ph.D. in physics in 1987 from the Max-Planck Institute for
Quantum Optics, Garching, and at the Technical University of Darmstadt, both in Germany.
He has worked as a scientific collaborator for the German Aerospace Center (DLR) and at
the Institute for Technical Physics of the German Research Center for Aircraft and Space
Technology (DFVLR, today DLR) in Stuttgart and at the Fraunhofer Institute for Solid State
Technology, in Munich, Germany. Since 1989, he has worked for Ludwig-BlkowSystemtechnik GmbH, Ottobrunn, Germany.
Dr. Zittel actively researches climate change effects on the trace gas methane
(especially in Russia); the external and social costs of energy use in the public transport
sector; the energetic amortization of renewable energy generation technologies;
advanced hydrogen storage technologies; the use of hydrogen in existing natural gas
grids; hydrogen production from biomass; analysis of the existing hydrogen markets;
introduction scenarios for hydrogen into the energy and transport economy; fundamental
questions of the power industry; analysis of fossil resource availability; and advanced
hydrogen storage systems for automotive use.
He is a member of the Association for the Study of Peak Oil, Germanwatch e.V., and
Global Challenges Network e.V.
prelims
2/7/07
3:51 PM
Page xxviii
ch-01
2/6/07
7:36 PM
Page 1
Abstract
This paper sketches a possible path towards the future energy supply in the light of
foreseeable challenges and opportunities. First, the shortcomings of the World Energy
Outlook, which is biennially published by the International Energy Agency, are revealed.
The deficiencies of these reports are the reason why a more progressive approach and
alternative scenarios are needed. An alternative approach must address both the
challenges as well as promising pathways towards a sustainable energy future. Second,
we briefly address the challenges and discuss those non-sustainable approaches which
represent todays conventional wisdom in the energy community. Finally, and most
important, we want to describe the real potential of renewable energies in a scenario
called Alternative World Energy Outlook 2006.
It is not the intention of this paper to draw a complete and exact picture of what our
energy future will look like, as every attempt to do so is certain to fail. However, the aim
is to describe the limits of what is possible and what is impossible simply due to
geological and physical restrictions. Most relevant in this respect is to take into account
the peaking of oil production in the near future and the peaking of natural gas production
in the mid-term future.
Keywords World Energy Outlook; renewable energy; climate change; economic principles; logistic
growth
ch-01
2/6/07
7:36 PM
Page 2
R/P
SO2
SOT
t
toe
Tcf
UK
USA
WEO
yr
ch-01
2/6/07
7:36 PM
Page 3
the public. The following scenarios try to figure out if the present resource depletion
problem eventually could be mastered just by increasing the extraction rate of other
conventional energy sources.
This transition period probably has its own rules which are valid only during this phase.
Things might happen which we have never experienced before and which we may never
experience again once this transition period is over. Specifically our way of dealing with
energy topics may change completely, which will have consequences for our economic
systems.
The International Energy Agency (IEA) denies that such a fundamental change of our
energy supply is likely to happen in the near or medium-term future and therefore does
not give a warning that our economic system is in danger. Therefore the views of the IEA
are briefly analysed. It will be shown that the business as usual approach of the biennially
published World Energy Outlook does not describe our energy future appropriately.
After that the main drivers determining the transition to a changing energy future are
discussed. It is important to understand that we are entering such a transition period,
leading to fundamental changes. The acknowledgement of this development and the
resulting mindset are necessary preconditions for coping with this situation in an
appropriate way. This will lead to completely different behaviour on the part of individuals,
companies and governments. The imminent transition is not a voluntary act in which
people might or might not engage because changing boundary conditions will force us to
adapt our energy system and also our way of life.
projections, energy consumption will rise tremendously until 2030. There will be no
sign of rising supply problems or an imminent peak of oil production.
Fossil fuel use will dominate the supply. Renewable energy sources (apart from
hydropower or traditional biomass use) will contribute only a minor share and will
only rise marginally over the next 25 years. Even by 2030 the contribution will still be
only about 2 per cent.
A more detailed analysis reveals that the secured supply of this rising demand over the
next 30 years relies on a few basic assumptions which are not questioned by the IEA.2, 3
The main assumptions are:
On the one hand it is claimed that oil resources are sufficient to supply the projected
rise of consumption. At the same time, the IEA recognizes that proved reserves are
not sufficient and new reserves must be discovered. In addition, it is stated that the
uncertainty and poor quality of data is of growing concern to all involved in the oil
industry. Even with this background, the IEA concludes that growth of oil
ch-01
2/6/07
7:36 PM
Page 4
ch-01
2/6/07
7:36 PM
Page 5
Mtoe
6000
oil
5000
2005
WEO 2004
WEO 2005
WEO 2005-altern.
WEO 2005-low invest
gas
4000
coal
3000
2000
biomass
1000
Nuclear (x3)
0
1965
Other-WEO2005-alt
Other-WEO2005
Other-WEO2004
hydro
1985
2005
2025
Year
Source: Historical data BP Statistical Review of World Energy Outlook International Energy Agency 2004/2005
FIGURE 1.1 World energy demand 19712030 as given in the IEAs World Energy Outlook 2004 and 2005
contribution, with about 1 per cent is derived from all other renewables such as wind, direct
solar energy, geothermal energy and wave power. The contribution from these sources might
rise to a share of 2 per cent in 2030, according to the WEO projection.
The graph tells us that oil will remain the most important fuel. Its consumption and
production levels will increase over the next 25 years by the same amount as they did
during the last 40 years. The message to consumers is that in the next two decades we
will see no major changes. Not shown in the graph, the price of oil is assumed to rise to
US$27/bbl by 2030 in WEO 2004 and to US$39/bbl in WEO 2005.
The latest WEO 2005 report also discusses a low investment scenario and an
alternative policy scenario. The projected oil consumption for these two alternative
scenarios is also shown in Figure 1.1. According to the report, the low investment
scenario would result in an oil price rise of up to US$52/bbl by 2030. In contrast to these
scenarios, however, oil was traded at US$6070/bbl for several months in 2005 and early
in 2006 and has risen above US$75/bbl.
Another IEA message is that natural gas and coal supplies will compete for the second
most important position behind oil, with natural gas supplies probably developing more
rapidly than coal: the supply of natural gas is expected to more than double within the
next 25 years.
At present, North America and Europe are the biggest natural gas markets, consuming
more than half of the worlds natural gas production. In both regions natural gas
production has already peaked. North America is experiencing a decline of its domestic
supply. Only very small amounts of natural gas are imported from foreign sources due to
limited import capacities. There is no doubt that large quantities of natural gas are still in
the ground. However, it is doubtful that future global extraction rates will exceed past
ch-01
2/6/07
7:36 PM
Page 6
levels when production in North America and Europe is already in decline. Any new natural
gas supplies must first close this gap before there can be a net increase. This may turn
out to be a big challenge!
A few comments can also be made on the other energy sources. The growth of biomass
use is supposed to continue as in the past. On nuclear energy the report spends just 1 out
of more than 400 pages, claiming without any further justification that total production will
remain almost at present levels. New renewable energies like wind or solar will almost triple
their contribution within the next 25 years. This is based on an expected average growth rate
of about 4.5 per cent per year. However, this rate is far below the actual experience: during
the last 15 years the average growth rate of wind energy was about 3035 per cent per year,
for solar thermal energy about 1015 per cent and for solar electricity production 1520 per
cent and strongly accelerating in the last few years.
It is obvious that the IEA systematically overestimated the future role of fossil energy
sources and underestimated that of renewable energy sources.
ch-01
2/6/07
7:36 PM
Page 7
14
12
other
10
8
6
coal
gas
oil
1920
1940
1960
1980
Emission reduction:
ecological welfare
2000
2020
2040
Year
FIGURE 1.2 The basic dilemma Business as usual or climate policy; growing fossil
energy consumption is tantamount to growing air pollution and reinforced greenhouse effect
to decline. Within a few years it will be followed by the decline of natural gas production
and after that by the geological restrictions of uranium and coal supply. The question is
not whether we like this or not depletion will occur even if one tries to ignore it. The
limited resource base will be detailed in the following section.
The third major driver is more positive. During recent years we have seen lots of
improvements and technical innovations in promoting the use of renewable energy
sources. This positive driver will make it easier to manage the transition from the
unsustainable present to a more sustainable future. However, this is not an automatism.
How fast the major economies of the world especially will move in such a direction
depends on many players these are the political leaders, the industry leaders, and of
course the consumers and voters. The extent and speed of action will be strongly
influenced by the assessment of the present situation.
ch-01
2/6/07
7:36 PM
Page 8
(mostly economists), rising prices will induce a fast increase of oil exploration and
production which in turn will lead to a relaxation in the oil market in the near future. In
contrast, the pessimists (mostly influenced by geological considerations) expect that it
will become increasingly difficult to balance the increase in demand by a sufficient rise in
supply. As a consequence production will not be able to keep up with demand and, after
a short phase of stagnation, will inevitably decline.
Many events during the last five years vindicate the theses of the pessimists. But it
would be much more appropriate to leave these misleading categories behind and to
speak only about realistic and unrealistic views. Facts and not beliefs will decide the
issue.
In the following sections the present state of the worlds oil supply will be outlined.
(For a more detailed analysis see Notes 4, 5 and 6.)
ch-01
2/6/07
7:36 PM
Page 9
the more existing spare capacities are reduced, the closer the production profile will follow
the described pattern.
In the history of oil production, which is now extending over more than 150 years, we
can identify some fundamental trends:
the worlds largest oil fields were all discovered more than 50 years ago;
since the 1960s annual oil discoveries have decreased tangentially;
since 1980 annual consumption has exceeded annual new discoveries;
till now more than 42,000 oil fields have been found, but the 400 largest oil fields
(1 per cent) contain more than 75 per cent of all oil ever discovered; and
the historical maximum of oil discoveries must be followed by a maximum of oil
production (the peak).
How close to the peak have we already got? How steep will the decline be after the peak?
These are the crucial questions remaining.
ch-01
2/6/07
7:36 PM
10
Page 10
Mb/day
history
40
35
30
Ecuador 99
Colombia 99
Argentina 98
25
20
15
10
5
Mexico 04
Norway 01
Oman 01
Australia 2000
United Kingdom 99
Malaysia 97
Gabon 97
Syria 95
India 95
Egypt 93
Indonesia 77
NGL
Romania 76
Alaska
Canada (conv.) 74
Rest-USA 7 1
Germany 67
Austria 55
Texas 71
1900 10
20
30
40
50
60
70
80
Yemen
Neutral Zone
Brazil
Angola
China
89
GOM
90 2000 10
Source: IHS 2003, BP Stat Rev 2005: Denmark, UK, Canada, Norway, Alaska, USA, Mexico, Brazil, Argentina:
2005 estimate based on Jan-Nov data from government statistics, Analyses and Forecast LBST
FIGURE 1.3 Oil production of countries outside OPEC and the FSU
ch-01
2/6/07
7:36 PM
Page 11
11
Mb/day
5
history
4
Bitumen
(4Gb)
3
Synthetic
crude oil
(6.7Gb)
Sasketchwan
Alberta (5.5Gb)
1
0
Total conv
New Foundland
oil production (18Gb) (1.4Gb)
1960
1970
1980
1990
2000
2010
2020
2030
Tarsands
(<10API)
Heavy
oil (1017API)
NGL-Alberta
Convcrude
oil production
Source: 19752004 data National Energy Board, Canada; 19601974 data USDoEEnergy Information Administration 2005:
Estimate by NEB, January 2006; 20062015 Forecast, tar sands based on NEB-study, May 2004, conventional and heavy oil
based on LBST estimate
FIGURE 1.4 Canadian oil production 19602030. The rising contribution of non-conventional oil production
(heavy oil below 17API and tar sands below 10API) substitutes declining conventional supply
collapse during the economic breakdown turned out to be much steeper than expected.
After the liberalization of the oil market, Russian companies were able to stop this decline
and to increase production levels again at double-digit rates in some years during the
last five years with the help of international cooperation and investments. However, this
fast recovery has now come to an end as the easily accessible fields have been developed
and the financial and technological backlog has caught up.
The double-digit growth rates in Russia contributed to compensating for the inescapable
production decline in other regions of the world. But despite this strong revival of Russian
production, oil prices have remained under pressure and have been rising slowly but
continuously and have even exceeded US$75 per barrel, a sixfold increase since 1999.
The two other important oil regions of the FSU are Azerbaijan and Kazakhstan.
Azerbaijan is the oldest industrial oil region of the world. Its highest production rates were
reached 40 years ago. Today, we can expect an expansion of production only in the offshore
areas. There, especially, the field complex Azeri-Chirag-Guneshli, has to be mentioned. Once
fully developed, this field probably will reach its maximum in 2008 or 2009, with a production
rate of 1Mb/day. Soon thereafter the production rate will decline very fast to almost negligible
amounts within 1015 years. The total production of this region, however, will increase by a
smaller amount as 150,000bbl/day are already produced from Azeri-Chirag-Guneshli today
and as the production from other fields will drop noticeably in coming years.
For some years Kazakhstan was considered to be a potential counterbalance for Saudi
Arabia. Today we know that these hopes were exaggerated. They were nurtured by
ch-01
2/6/07
12
7:36 PM
Page 12
speculations of the US Energy Information Administration (EIA), which estimated the oil
and gas reserves in the Caspian Sea region to be up to 300Gb of oil equivalent. Realistically,
only about 45Gb of oil are likely to be recoverable, about half of this amount located in
already developed fields.
High expectations regarding their future production potential concentrate on three
fields: Tengiz, Kamchagarak and Kashagan. Tengiz and Kamchagarak have been
producing oil for some years. All three fields contain oil with a high sulphur content, the
development of which jeopardizes the environment and is very expensive. In Tengiz
alone, more than 4500 tons of sulphur are separated from the produced oil each day and
stored in the surrounding area, polluting the environment. Plans for an extension of
production are delayed due to high development costs and difficult geological
conditions. In 2000 the third big oil field, Kashagan, was found. It is assumed that
production can be increased considerably from 2006 on; however, there are big doubts
whether this will be possible. The high sulphur content, a high deposit pressure of more
than 1000 bars and an unfavourable geographical location far away from any
infrastructure make it difficult and expensive to develop. It is certainly no coincidence
that two of the big companies involved in the discovery of the field (BP and Statoil) have
withdrawn from the consortium developing the field. After an analysis of the first
exploration drilling, it was communicated that the companies internal criteria for
development were not fulfilled.
Azerbaijan and Kazakhstan will probably be able to double their production rate by
2010 from 1.3Mb/day to 2.52.6Mb/day but to expect more seems unrealistic.
According to this assessment, the whole region may be able to increase its production
in the coming years, but the very big expansion expected by many people will not occur. A
production increase of 23Mb/day is probably already on the high side.
ch-01
2/6/07
7:36 PM
Page 13
13
now been producing oil for more than 50 years. It is a fact that more water is pumped into
the field than oil is extracted, and it seems quite possible that the production rate will
decline in the near future. Anyway, it is certain that Ghawar cannot contribute to an
expansion of the Saudi Arabian production.
There is an ongoing debate on whether Saudi Arabia will be able to increase its
production significantly. This debate was initiated in early 2004 by Matthew R. Simmons, an
American investment banker from Houston.7 Simmons very much doubts the possibility of a
significant growth of production. His assessment is based on a comprehensive in-depth
analysis of technical papers in the public domain addressing the problems of oil production
in Saudi Arabia and on a great number of interviews with engineers working on site and also
a visit to the oil fields in Saudi Arabia.8
Simmons has provoked comments by Abdul-Baqi and Nansen Saleri, senior
executives of the state-owned company Saudi Aramco.9 But their comments have rather
fuelled existing fears instead of assuring the world. First, it was admitted that the big old
oil fields are in decline, and that already the Abqaiq field has been depleted by 73 per cent
and Ghawar by 48 per cent. Moreover it was indirectly confirmed that the proved reserves
do not amount to 262Gb, as is widely assumed. The proved reserves amount to only
130Gb while another 130Gb have been counted as reserves already because it is regarded
probable that they can be developed eventually. If one were to apply the same criteria
which are common practice with Western companies, then Saudi Aramcos statement of
proved reserves should be devalued by 50 per cent. This was confirmed indirectly by
another Saudi Aramco executive.10
Furthermore, Saudi Aramco executives tried to counter Simmons fears by stating
that a production of 10Mb/day could be upheld until 2042. In doing this they had to
assume that the above mentioned reserves of 260Gb are proved reserves (which they
definitely are not). Saudi Aramco went on to state that in case of a more aggressive
development of the remaining reserves, production could be increased to 12Mb/day by
2016 and then could be maintained constant until 2033.9 But even this scenario put
forward by the Saudis is hardly reassuring in view of the projections of the IEA, which
assume that in the longer term an additional 20Mb/day are supposed to come from
those regions.
The analyses of Simmons and others (for example Bakhtiari11) make the point that
Saudi Arabias potential to increase production will soon reach its limits. The world is
nearing the moment of truth. The next few years will reveal whether those who believe
that OPEC has no more spare capacity left are right, or whether the peak of world oil
production can be delayed for a few more years.
Should world oil production be increased, it will be taken by many as evidence that oil
production can be increased for many more years to come. However, in reality if the oil
production increases, 1) the remaining oil will be consumed that much faster and 2) an
increase of production in the short term will in effect further increase the consumption of oil,
resulting in a steeper decline than otherwise necessary.
Recent developments are in obvious contrast to the assertions of the optimists which
do not foresee any problems in the availability of oil within the next 2030 years. But at
least they now acknowledge that price increases might be possible.
ch-01
2/6/07
14
7:36 PM
Page 14
Two oil supply scenarios are defined to be used later, these characterized by different
production profiles:
1 the high fossil scenario based on the Association for the Study of Peak Oil (ASPO)
production profile, with a peak before 2010 and a moderate decline rate of 23 per
cent per year;12 and
2 the low fossil scenario based on the fears that future global decline rates will be
higher than in the old mature oil regions (for example that decline rates in Alaska and
UK are 510 per cent versus 3 per cent in the lower 48 states of the USA).
Accordingly, after a peak in 2008 and plateau until 2010 a decline rate of 5 per cent
per year is assumed for 20102020, 3 per cent for 20202040, 2 per cent until 2050
and 1 per cent thereafter.
ch-01
2/6/07
7:36 PM
Page 15
15
Production forecast
(growth rate in 2010: 1.9 %)
Cumulative
discoveries
4000
2000
1920
1940
1960
1980
2000
2020
2040 2060
Year
2080
FIGURE 1.5 Cumulative worldwide gas discoveries and production. The extrapolation of the data supports
the forecast that in total about 12,000Tcf will be discovered. The peak of production can be expected from this
rough topdown analysis when about half of the total volume is produced, that is around 2025
Extrapolating the declining rate of new discoveries results in the assumption that in
total about 12,000Tcf (~325,000 billion m3) might be discovered until 2100. Proved gas
reserves are comparable in size to the proved oil reserves (~160Gtoe), but the already
produced share is smaller (one third for gas against half for oil). Therefore the global
production peak can be expected to happen later than for oil. A rough topdown approach
suggests that gas production will peak around 2025.
A bell-shaped production profile which fits the historical production pattern until 2004
and assumes the estimated total of 12,000Tcf, results in a smoothly diminishing annual
production growth rate (which is currently at 2.5 per cent and is expected to be at 1.9 per
cent in 2010). If future growth rates are larger, the expected production peak would
happen sooner and vice versa.
This extrapolation assumes that about 75 per cent of world gas reserves have already
been discovered (compared to 90 per cent for oil). Future reserve reassessments and
upward revisions of (older) producing gas fields will only marginally influence this pattern
as they usually do not result in changes of the production profile of these fields.
This topdown analysis does not take into account the quality of the gas fields. Especially
it does not make allowances for the so-called stranded gas situated far away from existing
transport infrastructures or for low quality gas with a high content of CO2 or SO2. Today,
stranded gas is not used for economic reasons. However, this could change with increasing
gas prices. But this would also imply much higher costs for the development of these fields
ch-01
2/6/07
16
7:36 PM
Page 16
1940
1960
1980
2000
2020
2040
2060
2080
Year
FIGURE 1.6 Annual gas production 19202004 and extrapolation based on a bell-shaped profile
and an estimated ultimate recovery of 12,000Tcf. Currently 3000Tcf have been consumed and proved
reserves are at 6300Tcf16, 17
and for the conditioning of the produced gas, especially for liquefaction or upgrading into
other liquid hydrocarbon fuels. The conditioning, transport and upgrading of this gas would
consume between 2050 per cent of its energy content, depending on whether the gas is
liquefied or transformed into synthetic crude oil, ammonia, methanol or hydrogen. Of course,
this conditioning would reduce the available gas reserves correspondingly.
These additional problems and the lead times for the construction of the necessary
infrastructure make it probable that the calculated production rate in this scenario
provides an upper (optimistic) limit for the gas extraction curve, shown in Figure 1.6.
Probably the peak will be sooner, followed by a plateau lasting several years.
Natural gas liquefies at temperatures of below -160C and ambient pressure. At ambient
conditions it is gaseous and can be transported best in pipelines. Therefore pipelines are the
backbone of the established gas markets, which have developed over time and connect the
major consumers with their supply regions. This regionalization is the main difference to
crude oil. Oil can be transported very easily, which helps to equalize regional imbalances.
The importance of the pipeline infrastructure leading to separate regional markets
makes it questionable whether the above sketched topdown approach provides the
basis for a meaningful interpretation. Probably this scenario will never be realized.
Regional supply scenarios have a much higher importance as they reflect regional supply
problems. Though liquefaction is possible and will be expanded in the future, a much higher
effort is necessary regarding energy, materials, investments and lead times. Today, about 7
per cent of the traded natural gas is liquefied; about 90 per cent is transported via pipelines.
In the following the most important regional markets are sketched North America and
Europe. Both regions together produce about 45 per cent of the worlds gas and consume
ch-01
2/6/07
7:36 PM
Page 17
17
about 55 per cent. In both markets the supply situation has dramatically deteriorated during
the last few years, coming as a complete surprise for economics-oriented gas market
analysts. The analysis of these regional markets provides a better understanding of
possible future supply restrictions than the optimistic topdown approach sketched above.
Tcf/year
25
Imports
20
15
US production
Louisiana
10
Oklahoma
Texas
0
1935 40 45 50 55 60 65 70 75 80 85 90 95 2000 05
Year
FIGURE 1.7 Gas supply in the US. Since 2001 production has been in decline; even imports from
Canada cannot substitute declining domestic supplies as production in Canada is also past its peak19
ch-01
2/6/07
18
7:36 PM
Page 18
conditioning exists. On average, in summer about 40 per cent less gas is consumed than
in winter.
The gas production, however, is almost independent of the season. Therefore huge
amounts of gas are stored during the summer. The weekly storage additions are used by
commodity market analysts as a measure of a tight or relaxed supply situation.
Over the last three decades gas has been traded on the commodity markets at
between US$1 and US$2 per million BTU (this corresponds to 3.77.4 cents/m3). As late
as the summer of 2002 the EIA forecast that gas demand and supply will rise by 50 per
cent until 2020.20 There was no hint at possible future supply problems by the agency.
According to this forecast, the gas price was expected to rise to about US$2 per million
BTU until 2020. But critical observers could foresee mounting problems already in the
winter of 2000/2001. Due to first supply restrictions well-head gas prices then climbed by
a factor of 3.5 up to a monthly average of US$6.7 per million BTU in January 2001.18
This sudden price rise had grave effects for some consumers: small and medium
enterprises with a high gas consumption (for example gardeners with large greenhouses)
went into bankruptcy; companies with long-term contracts (for example ammonia and
methanol producers21) stopped the production of their products and instead resold the
gas with a larger profit margin. The unexpected combination of high oil and gas prices
triggered the economic recession in 2001 which spread over to Europe and many other
countries. Since that time gas prices have stabilized at a level of US$68 per million BTU
during summer with spikes far above US$10 per million BTU during winter. At the
beginning of 2006 the price stood around US$10 per million BTU (~30 cents/m3).18
The declining domestic oil and gas production, in combination with limited import
capacities (new liquefied natural gas, LNG, terminals are planned but have long lead times
to construct), will probably increase the supply problems in the years to come. Over the
next years the domestic production will continue to decline. Canada will need more gas
for the extraction and upgrading of tar sands, which presently consumes about 5 per cent
of the domestic gas production.19
Hidden from the general public, many market observers and politicians, an energy
crisis looms over the horizon with grave consequences for the North American energy
markets, the whole economy and probably even for other regions.
ch-01
2/6/07
7:36 PM
Page 19
19
FIGURE 1.8 UK gas production 19702030. Each area represents the contribution from all new
fields which are developed within one year.23 The dashed line provides a possible future production
profile if expected future discoveries and remaining untapped reserves are connected in time
Since new field developments are rare, the future production profile can be estimated
with a high level of confidence. This is outlined in Figure 1.8. By 2010 the total production
will be 50 per cent below the level of 2000 and by 2015 another 3050 per cent below the
level of 2010, depending on the success of finding new fields.
Very soon the gas supply of the UK will depend on imported natural gas. A major
possible source is the Norwegian gas field Ormen-Lange, which will be directly connected
with the UK via pipeline. However, this field will supply gas at the earliest in late 2007. In
view of several downward revisions of the size of this field by the developing companies
and the retreat of BP from the consortium it is doubtful whether the initially envisaged
production can be achieved.24
The gas flow in the five-year-old pipeline connecting the UK with the continent has
been reversed. Originally the pipeline was built to export gas but is now used for
importing gas from the continent. Yet its capacity is rather limited. A large expansion of
import capacities will have to be provided by newly built LNG terminals.
The oil and gas supply situation of the UK should be seen as a warning which demonstrates how soon the days of surplus production of oil and gas, with their corresponding
high export revenues, can be followed by the necessity of steadily rising imports.
ch-01
2/6/07
20
7:36 PM
Page 20
Billion m3/year
history
forecast
500
400
Imports:
+5% per yr
LNG Imports
300
Imports from
Russia, North Africa
Germany
(900 billion m3)
200
100
Italy
(1000 billion m3)
Norway
(4500 billion m3)
0
1960
70
80
90
Year
2000
10
20
FIGURE 1.9 Natural gas supply of Europe probable development until 2020 according to scenario
calculations by LBST26
analysis of the UK, Norway and The Netherlands. Since 2004 the still growing supply from
Norway can no longer compensate for the declining supply from the UK and other
countries Europe has passed its gas production peak. Even a 50 per cent increase of
Norwegian gas production cannot stop the overall decline. Europe is in need of rapidly
rising imports from Russia, North Africa or other parts of the world.
According to this analysis, gas imports must rise at an annual rate of 5 per cent until
2020 just to keep the supply base flat. Future growth of demand would require even
higher rates. Even this zero growth scenario requires the construction and operation of
about four or five new pipelines with a net import capacity of about 30 billion m3/year until
2020.
In Europe, only Norway can still substantially expand its production capacity. A few
years ago its reserve to production ratio (R/P-ratio) amounted to more than 60 years.
However, the strongly rising production and diminishing new discoveries have now
reduced the R/P-ratio to about 30 years. Norway therefore will presumably reach its gas
production peak within the next 10 to 20 years. If Europe wants to increase its gas
demand according to the forecast, gas imports must at least double by 2020. More
precisely, new import capacities of about 150200 billion m3 must be built and put into
operation within the next 15 years. To put this required effort into perspective: the just
proposed and planned pipeline through the Baltic Sea connecting Russia with Germany
will have a capacity of about 28 billion m3 in the first phase (one line) and of 55 billion m3
in the second and final phase with two lines. It is expected that the first line will come into
operation at the earliest in 2010, but more likely in 2011 or 2012. However, by 2020 import
capacities about five times greater will be needed in Europe.
ch-01
2/6/07
7:36 PM
Page 21
history
Tcf/year
40
35
30
25
20
15
10
5
21
forecast
Zappolyarnoye (60Tcf)
+3% p.a.
Kharampur (12Tcf)
Yamburg (170Tcf)
Small fields
(29Tcf)
Astrahan (10Tcf)
Urengoy Severnyy(28Tcf)
Konsomolskoye (28Tcf)
Vyngapur (12Tcf)
Bolshoy Gubkin (16Tcf)
Orenburg(48Tcf)
Medvezhye (75Tcf)
Vuktyl(12Tcf)
Yubilneynoye (12Tcf)
Urengoy
Samotlar (9Tcf)
(250Tcf)
Old fields
(30Tcf)
0
1960
1970
1980
1990
2000
+1% p.a.
ca. 10 Fields (10-15Tcf)
Karasovey (26Tcf)
Leningradskoye (50Tcf)
Shtokmanovskoye (55Tcf)
Semokovskoye (15Tcf)
Rusanovskoye (25Tcf)
Bovanenko (70Tcf)
2010
2020
Year
1.2.2.5 Russia
The required additional amounts of gas cannot come from Russia as the three largest
fields (Urengoy, Yamburg and Medvezhye) containing about one third of the
discovered gas are already in decline.27 It is very questionable whether Russian gas
production still can be expanded for a longer time period. Figure 1.10 shows the fieldby-field analysis of Russian gas production since 1960 and the forecast until 2020.16 This
forecast is based on already known but not yet developed fields and supposes a
hypothetical time schedule for their development. The analysis leads to the conclusion
that probably an annual production increase of 1 per cent can still be realized in the next
10 to 15 years. However, this requires a timely development of new fields. These fields
are situated further north (in the Barent Sea and Kara Sea) or further east; therefore
development will be much more time consuming and much more expensive than the
development of the already producing fields. Also the question remains how much of
this gas will be available for export to Europe as Russian domestic consumption is
expected to rise and East Asian countries (China, Korea and Japan) will compete for
imports.
ch-01
2/6/07
22
7:36 PM
Page 22
Billion m 3
5000
4500
Latin America
4000
3500
3000
2500
2000
Africa
South Asia
East Asia
Middle East
China
OECD Pacific
OECD Europe
1500
Transition Economies
1000
500
0
1960 65 70 75 80 85 90 95 2000 05 10 15 20 25 30 Year
Source: IHS-Energy, BP Statistical Review of world Energy Projection: LBST 2005
For the formulation of global energy scenarios later in this paper two alternative global
gas production scenarios are used:
1 future gas production according to the Association for the Study of Peak Oil12 it
exhibits a production plateau for 20152040 at 3000Mtoe (=3400 billion m3 or 35 per
cent above the 2004 level); and
2 a second low fossil scenario assuming that this peak production plateau can only be sustained for ten years until 2025 and then will be followed by a decline of 3 per cent per year.
ch-01
2/6/07
7:36 PM
Page 23
23
FIGURE 1.12 Installed capacity of nuclear power plants and various forecasts
connected to the grid. Figure 1.12 shows the development of the cumulative nuclear
power capacity worldwide. Of the 550 reactors built more than 100 have already been
decommissioned; on average, old reactors have been decommissioned after less than 25
years of operation.
Assuming that the still operating reactors will be decommissioned after 40 years on
average, world nuclear power capacity will decline rapidly in the next 25 years, from 370GW
now to less than 100GW in 2030.25 In the upper right part of the figure a recent forecast of
the International Atomic Energy Agency is sketched.28 The next dotted line below is the
projection of the IEA in its most recent WEO.1 The shaded area below is the LBST scenario
describing what might be achieved if all efforts are undertaken to bring about a renaissance
of nuclear power: at best one probably could uphold the present level. The new capacity
line in the figure indicates how many new reactors must be built to fit this scenario. As the
number and timing of the required new reactors in this scenario seems to be very ambitious,
it is much more realistic to expect a decline of the installed nuclear capacity.
Figure 1.13 shows the worlds uranium resources based on public statistics.29, 30 The
upper part sketches the uranium used for nuclear weapons by assuming that this amount
will remain constant for the next 25 years. The area below shows the amount which is
already consumed by nuclear power plants in operation. If the present capacity remains
constant for the next 25 years a simple extrapolation shows that the proved uranium
reserves will be consumed by 2030, including the amounts stored from nuclear
disarmament and from reprocessing plants. Even if reasonable additional assured
uranium resources (believed to be producible at costs below US$130/kg) become
available, about 80 per cent of these reserves and resources would be exhausted by 2030.
ch-01
2/6/07
24
7:36 PM
Page 24
1.2.4 COAL
At the present production level, so-called proved coal reserves would last for another 180
years.17, 29 A bell-shaped production pattern (which is fitted to this reserve number and to
the historical growth pattern) allows an estimate that coal production will peak around
20502060. At peak, the production would be about 6070 per cent higher than the
present production rate. This profile is sketched in Figure 1.14.
This figure also shows another much more optimistic production pattern based on an
unrealistic doubling of present reserves. In this case the production peak could be delayed
until 2080 while reaching higher production levels. This hypothetical case is intended to
ch-01
2/6/07
7:36 PM
Page 25
25
Data Source: Historical Data: BP Statistical Review of World Energy, BGR Scenario: LBST 2005
FIGURE 1.14 Possible world coal production profiles based on 180 years and 370 years
of remaining reserves
demonstrate that even the assumption of such an optimistic production profile cannot
avoid the peaking of the supply of all fossil fuels within the next 50 years, as will be shown
below. Much more realistic seems to be an R/P-ratio of 180 years, which is maybe still too
optimistic in view of the bad experiences with the poor quality of oil and gas reserve data.
Furthermore, these production profiles do not take into account that coal is now
mainly used for electricity production. Additionally, in the future coal will also have to
substitute declining oil supplies. But this implies further transformation losses in the order
of 4050 per cent as each attempt to transform coal into a transport fuel (either into
synthetic crude oil via the Fischer-Tropsch process or into hydrogen) results in such losses
of the original energy content. Coal is a very dirty energy source emitting large amounts
of carbon dioxide and other pollutants. With huge efforts these emissions eventually
might be avoided the necessary technologies are available in principle, if not yet in
reality. But doing so provided that the reservoirs for the storage of the carbon are
available would reduce the usable net energy by another 2030 per cent.
The conclusion is that the coal production profiles shown in the figure provide an
upper limit of the future availability of coal and one which is not identical to the marketable
net energy.
ch-01
2/6/07
26
7:36 PM
Page 26
Data Source: Oil, Gas: Colin Campbell/ASPO 2005; Coal, Nuclear Scenario: LBST 2005
FIGURE1.15 High fossil scenarios of future production of fossil and nuclear fuels, based on peak oil
before 2010 and a decline after peak of about 23 per cent per year and a gas peak around 2040
the supply situation could develop in the next few decades which is rather biased to the
high side. Oil production is assumed according to ASPO projections; gas production is
taken from ASPO with a production plateau for 20152040 at a 35 per cent higher
production level than today; coal projections are taken from the bell-shaped profiles
sketched above, based on the more probable R/P-ratio of 180 years and on an upper limit
of 370 years.
Even within this upper limit scenario a further growth of world energy supply comes
to a halt as soon as gas production cannot rise any more. A further decline is unavoidable
in the following decades.
The production of nuclear energy cannot reverse this trend. Projections for nuclear
energy show the energy production from existing reactors and their phasing out after 40
years on average. For this figure, the conversion from nuclear electricity to primary energy
is based on 33 per cent efficiency. However, more appropriately, nuclear should be treated
as primary electricity, thus directly comparable to renewably produced electricity later on.
The low share of only a few per cent and the above sketched limited resource base makes
nuclear energy irrelevant for the global energy supply situation whether one tries to keep
its share or not. Yet from a financial view it is very relevant whether budgets are directed
to the nuclear industry or to other energy sources.
Due to the mature status of oil exploration and production, the energy needed to
supply high quality crude oil to the markets is increasing (through energy losses).
Therefore the difference between produced energy and supplied net energy increases
ch-01
2/6/07
7:36 PM
Page 27
27
Data Source: Oil, Gas: LBST 2005; Coal, Nuclear Scenario: LBST 2005
FIGURE 1.16 Low fossil scenario of future production of fossil and nuclear fuels, assuming oil
peaking in 2010 and decline rates of 5 per cent per year
over the years, thus reducing the energy that can be supplied to the markets below the
values shown in these figures. The accurate assessment of the future development of the
energy losses, which today average about 10 per cent for oil, is very problematic.
In order to get a better understanding of the upper and lower boundaries of possible
future development, a low fossil scenario is also formulated. Such a scenario is
presented in Figure 1.16.
Oil starts to decline at 5 per cent annually between 2010 and 2020, followed by a
softer decline of 3 per cent until 2040, then declines of 2 per cent until 2050 and 1 per
cent thereafter. The plateau of world gas production is expected to end in 2025 (and not
2040 as in the high fossil scenario) with an annual decline rate of 5 per cent in the
following ten years and 3 per cent thereafter. The growth of coal production is also
seen to be smaller than in the high fossil scenario above with a plateau at
3600Mtoe (not 4600Mtoe) lasting from 2020 to 2050 and followed by an annual decline of
1 per cent.
In this low fossil scenario the paramount importance of peak oil is demonstrated. Once
oil peaks and then declines at 5 per cent per year, no other energy source will be able to
stop the aggregate decline of supplies, even for a few years. This scenario is not unlikely.
The high fossil scenario (Figure 1.15) is regarded as being positive from an
economists point of view but is certainly regarded as negative from an ecologists point
of view. The carbon dioxide emissions related to this scenario are shown in Figure 1.17.
Even with fossil fuel consumption peaking around 2010, these high emission levels
might remain constant for 30 years before they begin to decline. And the worst scenario
with respect to greenhouse gas emissions (with an assumed R/P-ratio for coal of 370
years) maintains this emission level until the end of this century.
ch-01
2/6/07
28
7:36 PM
Page 28
FIGURE 1.17 Carbon dioxide emissions according to the production profiles as outlined in the
high fossil scenarios in Figure 1.15
But both scenarios are not as bad as the business as usual scenarios calculated by the
Intergovernmental Panel on Climate Change (IPCC). These scenarios are not based on a
corresponding available resource base. In contrast, the above described low fossil
scenario (Figure 1.16) would even be a help for a climate protection policy as total carbon
dioxide emission could drop by about 40 per cent by 2050 and by about 5560 per cent
by 2100.
ch-01
2/6/07
7:36 PM
Page 29
29
The main purpose of these scenarios is not so much to show how global energy
demand might be supplied in 2100, but to sketch the major characteristics of the
transition period between todays fossil-fuel-based energy economy and a possible future
based on renewable energies.
The following scenarios are modelled with so-called logistic functions which grow
exponentially in the beginning and then approach the ultimate potential attributed to
these sources with continually decreasing growth rates. The general pattern of the
function is fitted to the historical data by adapting two relevant parameters, namely,
T0, the time at which the growth rate is at maximum and the parameter b, which
provides a measure for the growth rate. This procedure is sketched in principle in
Figure 1.18.
For each of the ten world regions as defined in Figure 1.11 and each renewable
energy technology a growth scenario was modelled and adapted to the historical
development in that region. The parameters used for this modelling are listed in Annex II.
In Annex III the historical development of renewable energy generation over the last ten
years is provided. Wind energy has shown by far the largest growth rates, of about
3040 per cent, for the last two decades. The average growth rate of solar electricity
generation is between 20 and 30 per cent, increasing over the last few years to almost 40
per cent.
In the short term the growth is influenced by the experienced growth rate of
recent years. However, long-term installations are highly sensitive to the assumed potential.
Therefore the potential was estimated for each region and each technology individually. Due
to the broad variation of the estimates, an upper and lower figure was assumed for most
FIGURE 1.18 Principle of the logistic growth function and the meaning of its parameters, P, T0 and b
ch-01
2/6/07
30
7:36 PM
Page 30
FIGURE 1.19 Technical potentials of electricity production from renewable energies; keep in
mind that the thermal biomass potential is three times larger than its shown potential for
electricity generation.3143 (The potential for solar thermal power in China and Asia is also very
large; however, it was not considered in this study.)
production was calculated by multiplying the number of active volcanoes within the
specified region with a scaling factor.32, 33
Hydropower: the results are based on a literature survey.32, 34
Photovoltaics: the potential was calculated by multiplying the forecast number of
inhabitants and the gross domestic product with a specific factor which takes care
of available roof areas and solar radiation. The link between population and gross
domestic product was used to estimate the available roof area in 2100. This
methodology is described in detail in Notes 13 and 35. Therefore the basic
assumption is that PV modules are only roof and faade mounted. Obviously, this
underestimates the full PV potential.
ch-01
2/6/07
7:36 PM
Page 31
31
Biomass: biomass is the most discussed renewable resource as the potential varies
by a factor of ten, depending on the assumptions regarding arable land area and
fertilizer use. The results are based on a literature survey and on our own
calculations.36, 37
Wind energy: the results are based on literature survey and our own calculations.3840
Solar thermal power plants: for North Africa and the Middle East the results of the
Mediterranean study (DLR) were used. These results were adapted to other
regions.4143
In Annex I the detailed data are provided, including the potential for thermal energy
conversion.
Nobody today can accurately forecast how much of this potential will be used in the
future. This depends on many aspects which are not of a technical nature. Most important
will be the acceptance by the public in the face of increasing environmental and supply
problems. Society has to and will decide the value of environmental aspects on the one
hand and the value of producing and using energy on the other hand. To partly account
for such decisions, the scenario calculations were based on the minimum values of the
estimates of the technical potential, which by definition include environmental restrictions
as they are obvious today but do not include additional restrictions from lacking public
acceptance.
The aggregate renewable energy scenario is shown in Figure 1.20. It represents the
aggregation of the possible market share potentials for each technology in each world
region. In general it is found that in the long run plenty of renewable energy will become
available. The capacity increase of renewable energies might eventually reach its
maximum around 2060. But most important: primary renewable energies will supply most
energy in the form of electricity and not as fuel or heat.
The focus of the analysis is on the next 20 to 30 years, which are identified as
the critical period. The accurate size of the renewable potential doesnt influence the
growth rates during the next 20 to 30 years, so is of minor importance in the present
context.
Another factor influencing the net energy balance is the energy needed for the
construction of the new generation capacity. For a rough estimate relevant data of today
are used. These are summarized in Table 1.1 and are derived from detailed life cycle
analyses based on todays energy mix in Europe. Since the dominant factor in this balance
is the consumption for extraction and production of the raw materials, most of the energy
needed is in the form of electricity. For these calculations, the electricity consumption is
traced back to the primary fuel consumption based on the present European electricity
mix. Since these data change in the same way as the energy mix and technological
progress, they might be regarded to be a worst case estimate of the energy consumption
during the period of construction.
The detailed calculation including re-powering of the decommissioned power
plants reveals that until 2020 less than 200Mtoe are consumed annually for the
construction of new generation capacity. This figure increases to about 500Mtoe around
2040 and reaches its maximum of about 1000Mtoe around 2070. About two-thirds of the
ch-01
2/6/07
32
7:36 PM
Page 32
TABLE 1.1 Energy consumed for the construction of new power plants31
Hydropower
Wind energy
Geothermal electricity
Geothermal heat
Biomass (electricity)
Biomass (heat)
Photovoltaic (EFG-Si)
Solar thermal electricity
COAL
Mtoe/GW
0.92
0.12
0.13
0.022
0.13
0.06
0.2
0.18
OIL
Mtoe/GW
0.2
0.046
0.026
0.004
0.027
0.013
0.074
0.045
GAS
Mtoe/GW
0.026
0.019
0
0
0
0
0.042
0.073
NUCLEAR
Mtoe/GW
0.088
0.009
0
0
0
0
0.26
0.002
TOTAL
Mtoe/GW
1.24
0.19
0.14
0.026
0.157
0.073
0.58
0.37
energy is consumed as coal and one tenth as oil. As already mentioned, in the long run
the energy efficiency of the production processes as well as the energy mix may improve.
Therefore the total energy consumption for capacity building does not seem to be a
bottleneck restricting the growth rates of renewables during phases of declining oil or gas
supplies.
The full potential of the renewable energy sources was not utilized for the calculations
shown in Figure 1.20. If this were done it would mainly influence the shares of primary
energy supply towards the end of this century (or even beyond) but does not much
influence the development in the coming decades.
ch-01
2/6/07
7:36 PM
Page 33
33
These are the main conclusions concerning the renewable energy scenario:
For technological reasons it is obvious that most of the renewable energy potential
times larger potential than anticipated in this analysis. Other observers see biomass
as a very problematic energy source as it competes with food production and as the
potential might diminish with progressing climate change. How much of the biomass
will enter the market in the form of heat, fuel or electricity will strongly depend on
the technological progress, regional implementation of technologies, cost
competitiveness and demand. This question is left open in these scenario
calculations. Biomass is seen as heat production (different from Figure 1.19). The
conversion of biomass into fuels or electricity would reduce these figures by
between 10 and 70 per cent.
ch-01
2/6/07
34
7:36 PM
Page 34
FIGURE 1.21 High fossil alternative world energy outlook until 2100 the transition from
fossil fuels to renewable energy sources, based on the high fossil scenario (Figure 1.15)
restricted to 55 per cent (4200Mtoe/yr) and not to 75 per cent or 4700Mtoe/yr as before.
The probability of such a scenario is relatively high as world oil supply is now declining at
far more than 5 per cent per year in almost all offshore regions which have passed peak
production. Modern technology has helped to extract oil at a faster rate but this implies
that the eventual decline is even steeper.
In this scenario the next 2030 years will probably see a declining primary world energy
supply. The potential growth rates of renewable energies are large, but the absolute share
is still too small to substitute an oil decline in the order of 5 per cent annually.
Rising energy costs are to be expected because of rising fossil fuel prices and the
more expensive renewable energy technologies. However, renewable energy will become
cheaper as its market share grows. This effect can be estimated by a scaling function
which models declining prices. Typical industrial learning curves have led to a cost
reduction of some 1020 per cent for each doubling of the total cumulative production
volume since the beginning of industrial production.
Typical installation costs of renewable energy technologies and their possible cost
degression potential are given in Table 1.2. The fundamental difference in the utilization of
renewable and fossil energy sources is that the major part of the costs for renewable
energies occur during the installation as capital costs (this is to some extent also true for
nuclear energy), while the costs of fossil sources are dominated by the fuel costs.
Therefore it is legitimate to compare fossil fuel costs with installation costs for a rough
assessment of the financial impacts. The only exception is biomass, which is also
ch-01
2/6/07
7:36 PM
Page 35
35
FIGURE 1.22 Low fossil alternative world energy outlook until 2100 the transition from
fossil fuels to renewable energy sources, based on the low fossil scenario (Figure 1.16)
GEO
4000
5
30
HEAT
GEO
SOLAR
800
1200 (/m2)
9
100 million m2
10
20
30
30
dominated by fuel cost. Figure 1.23 provides a cost comparison of total energy supply
costs on this basis.
Past prices are adjusted for inflation to 2005. The projection is made on the basis that
oil prices will rise to US$100/bbl, gas prices to 0.28/m3, coal prices to 140/t and
biomass prices to 5 Euro-cents/kWh. The conclusion is that energy costs will rise, but this
rise will not be all that dramatic. The most pronounced increase is due to rising fossil fuel
costs. This phase will be followed by declining renewable installation costs, which in turn
will lead to growing market shares.
The detailed cost figures shown in Figure 1.23 should not be regarded as pretending
to be precise quantitative forecasts. The purpose of this rough scenario is only to
demonstrate qualitatively what could be expected. Generally, in the case of possible
ch-01
2/6/07
36
7:36 PM
Page 36
We are here!
FIGURE 1.23 Comparison between fossil fuel costs and renewable installation costs; the
increase of fossil fuel prices sets the order of magnitude; with growing market penetration
renewable energies get cheaper, spurring further growth of market share
energy shortages caused by insufficient fossil fuel supplies, the prices of fossil fuels will
probably rise more than those of renewable energies. In such a situation higher overall
energy price levels must be expected.
ch-01
2/6/07
7:36 PM
Page 37
37
One important aspect is the efficient use of energy. In our present system, with low
energy prices, efficiency is not really attractive and the economic rewards are weak.
This will certainly change with higher prices and also the social valuation of efficient
energy use will increase. This is the case of private virtue versus public responsibility.
Declining oil and natural gas supplies for stationary applications (like electric power
generation and the provision of heat) can relatively easily be substituted by energy
efficiency measures and by strongly rising contributions from renewable energy sources.
Probably there will be plenty of renewable electricity available but not enough fuels for
transport. A serious deficit of fuels will develop over time which can only partly be
compensated by improved efficiency along the whole transport chain.
There exist various options for producing alternative fuels: hydrogen produced with
electricity, hydrogen or synthetic liquid fuels produced with biomass, and hydrogen or
synthetic liquid fuels produced with the remaining coal instead of using it for power
generation. But whatever option is chosen eventually, fundamental changes in lifestyles
and in consequence in the whole economy will be caused by limited supplies and higher
energy prices.
The transport intensity will have to be reduced without endangering mobility, as will
the energy intensity of the production and use of goods. Less available energy and higher
fuel costs will increase transport resistance. There will be a stronger preference for
activities in the neighbourhood and there will be a penalty on longer distances. This will
eventually change land use patterns suburbia in many cases will not work anymore
and perhaps there will be a revival of the cities in the USA. The dominance of the private
car will recede and the perception of what properties an attractive car should have will
also change. Other modes of transport will gain importance and social acceptance.
The industrialized economies rely on growth. Once growth rates are shrinking this is
seen as a symptom of crisis no growth is a disaster. So nobody is prepared for a future
with continually declining fossil energy supplies. This scenario is not on the radar of public
awareness. Rising energy prices will change the perception and only then will society look
for sustainable pathways into the future. It will be a major task to adapt our way of life,
our economies and our political and social systems to this situation. There are no ready
solutions. A comprehensive public debate has not started yet. The outcome is open.
Another question is: What level of energy consumption is sustainable? What level of
human energy use can nature bear without destroying the basis of (human and other) life
on Earth? One line of thought is the idea of a 1.5kW society every human being may only
use this amount of energy all the time. If this level of energy use were to be distributed
justly among all human beings it would mean a substantial reduction of energy use in
industrialized countries but also a substantial rise of energy use in poor countries.
To achieve such a goal, Germany would have to cut average energy use by a factor 4
(down from 5.5kW) and the USA by a factor 8! Half of this reduction probably is achievable
through more efficient technologies, but the other half will have to be brought about by a
change in lifestyles. Both approaches will be needed.
The transition certainly will not be easy and there will be many conflicts and many losers.
One reason is that the build-up of new energy supply structures needs very long lead times:
it is very doubtful whether there is still enough time for a transition without turmoil.
ch-01
2/6/07
38
7:36 PM
Page 38
The challenge will be to master this transition period in a compatible and controllable
manner. The viable path between economic disruptions on the one hand and ecological
disruptions on the other is very narrow. The later we start to realize the real challenges and
the necessity for drastic changes, the more difficult it will be to cope with the challenges
lying ahead.
ch-01
2/6/07
7:36 PM
Page 39
39
ch-01
2/6/07
40
7:36 PM
Page 40
41 H. Klaiss and F. Staiss (eds) (1992) Solarthermische Kraftwerke im Mittelmeerraum, Deutsche Forschungsanstalt fr Luftund Raumfahrt/Zentrum fr Sonnenenergie und Wasserstoffforschung, Springer, Berlin, London and New York.
42 F. Trieb (ed) (2005) Concentrating solar power for the Mediterranean region, final report by the German Aerospace
Center (DLR), Institute of Technical Thermodynamics, study commissioned by the Federal Ministry for the Environment,
Nature Conservation and Nuclear Safety, Germany, 16 April.
43 A. Leitner (2002) Fuel from the sky: Solar powers potential for Western energy supply, BDI consulting / NREL/SR-55032160, National Renewable Energy Laboratory, Golden, Colorado, July.
44 W. Weiss, I.Bergmann and G. Faninger (2006) Solar Heat Worldwide, Solar Heating & Cooling Program, International
Energy Agency, Paris, France.
45 European Solar Thermal Industry (ESTIF), see www.estif.org.
46 European Photovoltaics Industry Association (EPIA), see www.epia.org.
47 Global Wind Energy Council, see www.gwec.net.
48 IEA (various years) Energy statistics and balances of OECD countries, Paris.
49 IEA (2005) Energy statistics and balances of non-OECD countries, Paris.
ch-01
2/6/07
7:36 PM
Page 41
41
ELECTRICITY [1000Twhheat/yr]
Wind
PV
SOT
Geo
HEAT [Twhheat/yr]
Biomass
Geo
Solar
14
5.5
14
3.2
4.9
5.1
3.4
1.2
3.4
0.1
0.4
0.8
5.3
5.6
23.9
>1
0.9
>8
~0.3
0.3
>0.3
3.6
2.7
3.6
1.9
2.1
3
6.8
2.5
6.8
0.1
0.24
1
0.6
1.1
15.3
>1
1.2
>10
~0.1
0.1
>0.1
3.8
4.0
4.2
1.2
1.3
1.9
2.6
1.5
2.8
0.1
0.4
0.3
2.3
2.2
5.2
>1
1
>3
~0.2
0.2
>0.2
10.6
6.0
10.6
3.5
3.5
5.6
<0.1
0
<0.1
0.2
0.7
1.7
2.7
3.5
38.9
>2
2
>17
~0.1
0.3
>0.1
4.6
4.2
4.6
2.2
2.4
3.6
<0.1
0
<0.1
<0.1
0.03
0.1
2
3.2
30
4.6
2.5
4.6
1.1
1.1
1.8
0
0
0
0.4
0.79
1.7
0.9
1.1
4.9
4.6
2.8
4.6
1.5
1.8
2.5
0.3
0.25
0.3
<0.1
0.01
<0.1
0.8
1.7
8.6
5.4
5.2
5.4
3.6
4
5.8
0.2
0.1
0.2
0.4
0.7
2.7
6.1
7
31.4
<0.1
0.1
0.9
1.2
1.7
1.9
8.3
9
224
<0.1
0.05
0.3
0.1
0.1
3.8
1
0.1
>3
~0.1
0.1
>0.1
10.6
8.4
10.6
6.3
6.5
10
0.2
0.35
1.3
5.7
5.8
39
>2
0.4
>13
~0.5
1.3
>0.5
40
44
414
1
0.6
1
>4
4
>17
1
0.4
1
>4
6
>27
~0.5
1
>0.5
~0.8
0.5
>0.8
~0.6
0.3
>0.6
~0.3
1
>0.3
ch-01
2/6/07
42
7:36 PM
Page 42
SOT
Geo
Biomass1
HEAT
Geo2
Solar3
2500
2040
6.5
2900
2055
5.7
500
2050
6.2
50
2030
11
5000
2030
20
750
2050
10
300
2045
20
900
2040
5.5
1500
2036
4.5
1000
2060
10
30
2035
10
100
2045
12
600
2046
6.5
110
2026
12
1500
2023
5
1100
2035
4.4
600
2050
5.5
50
10
2040
200
2025
22
500
2040
10
250
2027
10
2000
2060
5.7
3500
2070
5
100
2050
6.5
300
2037
11
1000
2060
8.4
300
2050
3.5
1500
2040
4.8
1600
2045
4
3
2039
7
200
1950
40
300
2040
8
1000
2020
5.5
1000
2060
5.4
770
2050
4
100
2047
12
100
1970
11
2000
2060
5
500
2040
6
1000
2050
8
900
2040
3.8
100
2035
3.5
1
2030
8
150
1990
7,5
2
2035
15
400
2030
6
1500
2045
4.5
2000
2060
4.5
50
2040
4
100
2070
12
600
2020
10
3000
2065
7
1000
2030
6
30
2050
6
680
2050
3.8
3000
2060
5.5
6
2040
10
10
2020
13
30
2050
10
100
2035
7
3000
2060
6
4000
15,000
2090
2070
6.8
5.5
60
2060
9
500
2005
16
200
2060
8
1000
2045
7
1 Potential in Mtoe
2 Potential in GW thermal energy
3 Potential in million m2
ch-01
2/6/07
7:36 PM
Page 43
43
Year
For reasons of comparison the installed collector area is converted into power by the factor 0.7kW per m2. The data have been
collected over the years by the LBST from various different statistics, including data from Notes 4349. Though the absolute share
of photovoltaics is very small, its growth rate is accelerating, exceeding 40 per cent over recent years. Its absolute contribution is
lagging about ten years behind wind energy.
FIGURE 1.A.1 Cumulative installed capacity of wind energy converters, geothermal power plants,
photovoltaic modules and solar collectors
~910% p.yr.
Year
Photovoltaics, geothermal, wind and biomass are restricted to electricity generation. For reasons of comparison the heat
generation from solar collectors is also included. Not included are contributions from hydropower or heat from geothermal
sources and biomass. Fuels production from biomass is also not included. The data have been collected over the years by the
LBST from various statistics including data from Notes 4349.
ch-01
2/6/07
44
7:36 PM
Page 44
Wind power
Biomass (incl. traditonal use)
Year
Year
Biomass includes so-called traditional use in developing countries. Electricity from hydropower and other sources is directly
converted into Mtoe. The share of other renewables is relatively small, but exhibits by far the largest growth rates. The data
have been collected over the years by the LBST from various statistics including data from Notes 4348. The conversion from
capacity to energy production was performed by the LBST.
ch-02
2/6/07
7:36 PM
Page 45
Abstract
The quantum well solar cell represents the highest efficiency nano-structured solar cell
demonstrated to date. The concept and practical constraints for this cell are discussed
together with its potential application in high efficiency multi-junction solar cells and
thermophotovoltaic devices. Finally the ultimate efficiency limit of the quantum well
solar cell is discussed.
2.1 HISTORY
Since its inception, the quantum well (QW) has been applied widely in semiconductor
devices. Its roots stem from the resonant tunnel diode and it has found applications in a
large number of electronic devices, such as photodetectors, light-emitting diodes, lasers,
optical modulators and high mobility transistors. The almost universal utility of QWs is due
to the ability to engineer the optical and electronic properties of a QW structure to suit a
particular application. It is then reasonable to suppose that some advantages may exist
when using QWs in photovoltaic structures.
Early work (Chaffin et al, 1984) suggested that a superlattice could provide a means of
achieving highly efficient multi-junction solar cells, allowing the constituent band gaps to
be matched to the solar spectrum, without recourse to lattice mismatched techniques or
complex quaternary materials. However, the most decisive step in bringing QWs to
photovoltaics was made in the early 1990s with the proposal of the multiple quantum well
(MQW) solar cell (Barnham and Duggan, 1990). Here it was suggested that the QWs
should be located in the intrinsic region of a pin diode and the proposal was quickly
followed by a successful experimental demonstration in the AlGaAs/GaAs material system
(Barnham et al, 1991) showing a considerable increase in short-circuit current (Jsc).
The ability to adjust the absorption profile of a bulk semiconductor through the
inclusion of QWs gives rise to a number of practical, near-term applications for QW solar
cells, such as a component of a multi-junction solar cell, or in thermophotovoltaics where
low band-gap materials are required.
ch-02
2/6/07
46
7:36 PM
Page 46
N. J. EKINS-DAUKES
QWs and other low dimensional structures also offer some possibilities for testing
unconventional concepts in photovoltaics. For example, the proposal by Barnham and
Duggan also predicted that some advantage in the open-circuit voltage (Voc) could be
attained over bulk semiconductor material. This aspect has stimulated an interesting
debate over what electronic properties are necessary to achieve a photovoltaic voltage
enhancement in low dimensional semiconductors.
Many excellent review papers have already been written on lattice matched QW solar
cells (Barnham et al, 2002; Nelson, 2001 and 1995) and are accompanied by a large volume
of journal papers and conference proceedings. To distinguish this review paper from earlier
ones, the role of strain-balance is emphasized and is presented as a means of broadening
the range of materials that can be usefully applied in QW photovoltaic devices.
ch-02
2/6/07
7:36 PM
Page 47
47
k||
FIGURE 2.1 a) Layer structure showing three QWs surrounded by barrier material; b) Band diagram for
three QWs, showing two confined states for electrons and holes; c) Energy vs in-plane momentum dispersion
relation for a QW, showing two confined electron levels (e1 and e2) and the first heavy hole (hh1) and the first
light hole (lh1) levels
ch-02
2/6/07
48
7:36 PM
Page 48
N. J. EKINS-DAUKES
for two bound electron levels (e1 and e2) and the first heavy and light hole levels. The
degeneracy in the hole levels at k=0 is lifted when the QW is strained (Barnham and
Vvedensky, 2001).
The quantum confinement of carriers also serves to restrict the density of states.
Figure 2.2 compares the step-like behaviour of the QW 2D density of states to the
continuous 3D density of states; the expressions for the 2D and 3D densities of states are
expressed below:
m
h2
(1)
1 2m 3
(
)2 E
2 2 h2
(2)
g2D (E) =
g3D (E) =
where m* is the carrier effective mass. It is clear that in the 2D case, the density of states
is independent of energy for each bound level and depends only on the effective mass. It
is this simple dependence of the QW density of states that gives rise to the many
applications for QWs in electronic devices. For example, the strength of optical transitions
is governed by the density of states, so materials can be chosen to optimize the oscillator
strength of a particular transition, and the well width can be adjusted to achieve the
desired transition energies.
Finally, the number of bound states in a QW is of particular importance for the
QW solar cell, as each bound level adds further states from which absorption can
FIGURE 2.2 Density of states for a 3D bulk semiconductor g3D and for a 2D quantum well structure g2D
ch-02
2/6/07
7:36 PM
Page 49
49
take place. The number of bound states in a QW is given by the following expression
(Bastard, 1988):
n(L) = 1 + Int[
2m Vb L2
]
2 h2
(3)
where Vb is the potential of the confining barrier, L is the well width and Int returns the
lowest integer value of the argument. The role of the effective mass is again important
and can lead to situations, such as in an Al0.3GaAs/GaAs QW, where the band offset gives
a deeper well for the electrons than for holes, yet the larger hole mass more than
compensates for the band offset, giving more heavy hole bound states than bound
electron states.
When a series of QWs is grown, if the barriers are thick, such that only a negligible
fraction of each QW wave function penetrates into the neighbours, then each QW can
be considered isolated and the structure is termed a multiple QW (MQW). The properties
of the MQW can be considered simply as a linear multiple of single QWs. However, when
the barriers separating each well are sufficiently thin to allow sufficient overlap of the
wave functions from adjacent wells, then the QWs become coupled and a superlattice
is formed (Dingle et al, 1975). Under these circumstances, the energy levels
characteristic of an isolated QW broaden to form mini-bands that extend throughout the
superlattice. There is no clearly defined point at which an MQW becomes a superlattice,
as narrowing of the barriers leads to an ever increasing broadening of the confined
levels. Strictly it is only by defining an arbitrary bandwidth that the transition from MQW
to superlattice can be determined. However, for most structures discussed in this review,
the thermal broadening of the confined levels at room temperature is greater than any
superlattice effects, so the structures can be safely called MQW devices. Some
consideration is given to superlattices in section 2.3.4.
Before concluding this section, it is important to note that the QW band alignment
considered above is one where both the electron and hole are confined and is called a
type-I alignment. On occasions, the band alignment of the QW and barrier material are
such that only the electron or hole is confined, but not both, and the QW is then said to
be type-II. To date only type-I QW arrangements have been pursued for photovoltaic
applications as the optical transitions are generally stronger. Nevertheless, type-II band
alignments have some desirable properties, such as the suppression of Auger
recombination (Tang et al, 1995), and have been exploited for optoelectronic applications
(Gevaux et al, 2001).
ch-02
2/6/07
50
7:36 PM
Page 50
N. J. EKINS-DAUKES
FIGURE 2.3 a) QW p-i-n layer structure; b) Band diagram for the QW p-i-n solar cell showing the
photogeneration and recombination processes, together with the carrier capture and escape routes
ch-02
2/6/07
7:36 PM
Page 51
51
FIGURE 2.4 Typical quantum efficiency for a GaAsP0.06/In0.1GaAs 75, 35 MQW structure at various biases
InGaAs. The inclusion of just 4 per cent Al into the GaAsP barriers introduces sufficient
background impurities1 to give an almost negligible spectral response in forward bias.
Similar results are observed if the QWs are placed in the doped regions of the device, at
which point the diffusion length is reduced to the order of the barrier width (Ekins-Daukes,
2000; Barnham et al, 2002).
There are also instances where there is insufficient thermal energy to allow efficient
carrier escape from the QWs, for example in InP/InGaAs QWs (Zachariou et al, 1998a) and
in particularly deep strain-balanced InGaAs/InGaAs QWs (Rohr et al, 2002a). However, if the
electric field is maintained across the QWs and the temperature is sufficiently high, then
efficient carrier escape is maintained beyond the operating point for most MQW solar cell
devices. This is illustrated in the case of the GaAsP/InGaAs device by the light IV curve
shown in Figure 2.6. The temperature coefficient for MQW solar cells has been found to
be marginally better than for equivalent p-i-n solar cells made from the well and barrier
ch-02
2/6/07
52
7:36 PM
Page 52
N. J. EKINS-DAUKES
FIGURE 2.5 Quantum efficiency for an Al0.04GaAsP0.06/In0.1GaAs 75, 35 MQW showing how background
impurities lead to a collapse of the QE at forward bias
material (Ballard et al, 1998 and 2001). Historically the AlGaAs/GaAs material system was
the first material system in which the p-i-n QW solar cell was demonstrated (Barnham et
al, 1991) and was followed by other lattice matched material systems, such as InGaP/GaAs
(Osbourne, 1994; Zachariou et al, 1998b), InP/InGaAs (Freundlich et al, 1994; Zachariou et
al, 1996; Zachariou, 1996) and InGaAsP/InP (Rohr et al, 1998 and 1999; Raisky et al, 1998;
Rohr et al, 2000a; Raisky et al, 2001). A number of strained material systems have also
been investigated, for example GaAs/InGaAs (Barnes et al, 1994; Ragay et al, 1994a;
Barnes et al, 1996; Barnes, 1994), InP/InAsP (Freundlich et al, 1994), InP/InGaAs
(Serdiukova et al, 1997; Freundlich, 2000) and also some strain-balance structures based
on GaAsP/InGaAs (Ekins-Daukes et al, 1999), GaAs/InGaAs (Ekins-Daukes, 2000; EkinsDaukes et al, 1998), InGaAs/InGaAs (Rohr, 2000; Rohr et al, 2000b and 2002b). The strained
and strain-balance materials are discussed further in section 2.3.3.
2/6/07
7:36 PM
Page 53
Current / mAcm2
ch-02
SC=32.1
53
mAcm2
Voc = 09.55V
FF = 82.9%
Eff=18.7%
Bias/V
Source: Bushnell (2002)
FIGURE 2.6 Light IV curve for a GaAsP0.06/In0.1GaAs 75, 35 MQW structure under approximately 1 sun
AM0 equivalent 3000K tungsten halogen illumination
Despite the complexity of the absorption profiles of the MQW solar cell, it has been
possible to model the spectral response accurately using the envelope function
approximation outlined earlier (Paxman et al, 1993; Renaud et al, 1994; Nelson, 1995).
Computer modelling has enabled advanced cell structures to be designed, featuring
compositionally graded layers and back surface mirroring, to improve minority carrier
transport and optical absorption (Connolly, 1997; Connolly et al, 1998).
ch-02
2/6/07
54
7:36 PM
Page 54
N. J. EKINS-DAUKES
To give a broad indication of the recombination behaviour that can be expected from
a QW solar cell, Figure 2.7 shows values of J0 plotted against the cell absorption edge for
a large number of solar cells, with band gaps spanning the solar spectrum. A theoretical
estimate of the radiative limit is also plotted, showing the lower limit for J0, assuming unity
absorptivity (Shockley and Queisser, 1961).
Ln(J0)
FIGURE 2.7 Plot of J0 vs absorption edge for a variety of lattice matched, mismatched and MQW devices
The values of J0 have been extracted by a simple process, taking experimental values
for Jsc and Voc and assuming that the solar cell behaves as an ideal diode. Under these
conditions J0 is estimated as follows:
Jsc = J0 e
qVoc
kT
(4)
ch-02
2/6/07
7:36 PM
Page 55
55
InGaAsP/InGaAsP MQW (Connolly and Rohr, 2003), bulk GaInP (Yang et al, 1997), bulk
GaAs (Tobin et al, 1990), lattice mismatched InGaAs (Takamoto et al, 2000; Hoffman et al,
1998), bulk GaInNAs (Friedman et al, 1998; Li et al, 2000), bulk In0.53GaAs (Takamoto et al,
1997), bulk InGaAsP (Takamoto et al, 1997) and bulk Ge (Nagashima et al, 2003).
It is clear from Figure 2.7 that similar performance is observed from most of the MQW
structures as bulk cells. Of particular interest are the low band-gap MQWs (Connolly and
Rohr, 2003) that show exceptionally low recombination, in comparison with the
neighbouring bulk p/n cells, making the cells attractive for thermophotovoltaics, as
discussed in more detail in section 2.4.2. It is also noteworthy that the strain-balanced
GaAsP/InGaAs cells (Ekins-Daukes et al, 1999) have a similar J0 to the lattice mismatched
InGaAs cells, leading to possible applications in multi-junction solar cells, discussed in
section 2.4.1.
When attempting to model the Voc or dark IV of an MQW cell, it is usually assumed
that the MQW system is sufficiently close to quasi-thermal equilibrium to allow electron
and hole quasi-Fermi levels to be used to describe the carrier populations in the QW. The
extent to which quasi-thermal equilibrium is established in an MQW is discussed in
section 2.5, but nevertheless, the dark currents in a number of QW systems have been
modelled accurately (Nelson et al, 1995; Nelson, 1995) including the effects of
asymmetric QW location in the i-region (Nelson et al, 1999).
The success of computer modelling of MQW structures has led to more general
MQW device optimization studies (Connolly et al, 2000 and 2001). While actual device
performance clearly depends on a large number of material parameters, these models
now have a certain degree of predictive power that make them useful for assessing
suitable MQW device designs.
ch-02
2/6/07
56
7:36 PM
Page 56
N. J. EKINS-DAUKES
FIGURE 2.8 Strained GaAs/InGaAs 3period MQW device: a) Schematic diagram showing the layer
configuration; the relaxed lattice parameter is indicated laterally; b) Electronic band-structure for the device
effects of strain relaxation are so severe that strained GaAs/InGaAs structures can never
offer a substantial increase in current over a GaAs cell without suffering a catastrophic
loss in voltage (Ekins-Daukes et al, 2001).
In the strained GaAs/InGaAs structure, strain builds up with each InGaAs layer.
However, it is possible to minimize the build-up of strain by incorporating alternate tensile
and compressively strained layers. Through an appropriate choice of alloy composition
and layer thickness, and accounting for differences in elastic constant, it is possible to
achieve structures which are locally strained but exert no net force on the substrate or
neighbouring repeat units. The layer thickness required to achieve a strain-balanced
structure can be estimated using the following expression (Ekins-Daukes et al, 2002):
a0 =
A1 t1 a1 a22 + A2 t2 a2 a21
A1 t1 a22 + A2 t2 a21
(5)
where a0 is the substrate lattice parameter, a1 and a2 are the respective QW and barrier
lattice parameters, t1 and t2 the respective QW and barrier layer thicknesses, and A1 and
A2 constants for the QW and barrier and defined through the elastic stiffness coefficients
Cxx as follows:
A = C11 + C12
2
2C12
C11
(6)
ch-02
2/6/07
7:36 PM
Page 57
57
FIGURE 2.9 a) External quantum efficiency for a strained GaAs/InGaAs MQW sample and GaAs control; b)
dark IV curves for GaAs/InGaAs strain-balance MQW cell, a relaxed GaAs/InGaAs MQW cell, a strained
GaAs/InGaAs MQW cell and a GaAs control
ch-02
2/6/07
58
7:36 PM
Page 58
N. J. EKINS-DAUKES
strain-balance stack. The layer thicknesses determined from Equation 5 are based on
linear elastic theory and Vegards law (Vegard, 1921) is usually used to relate the material
composition to lattice parameter. These are good approximations, but will become
increasingly inaccurate at high strain. Finally, the material parameters used to determine
the strain-balance condition are usually those measured at room temperature, yet the
device is grown at an elevated temperature (up to 650C), and differences in thermal
expansion between the component layers ensure that no structure can be exactly strainbalanced at both the growth temperature and room temperature.
A typical strain-balanced structure is shown in Figure 2.10. Here GaAsP is used as the
tensile material to strain compensate the compressive InGaAs layer. Such strain-balanced
structures have demonstrated unprecedented photovoltaic performance for QW solar
cells (Ekins-Daukes et al, 1999). The quantum efficiency and dark IV are shown in
Figure 2.11. The quantum efficiency plot indicates the additional absorption due to the
QWs over a GaAs p-i-n control sample. The dark current profiles show excessively high
rates of recombination for the strained and relaxed GaAs/InGaAs cells, but the
strain-balanced cell has a dark current profile comparable to the GaAs p-i-n control cell.
Therefore, by using the strain-balanced QW approach, strained, lower band-gap
semiconductors can be incorporated into a solar cell with no additional misfit dislocations
and can therefore maintain a high Voc. A remaining challenge is to increase the absorption
of the MQW stack, as the 20 MQW sample only has a QE of approximately 20 per cent
due to the overall low optical thickness of QW absorbing material. This could be solved
by simply growing more QWs into the structure, and the current state-of-the-art device is
a 50 MQW GaAsP/InGaAs device (Bushnell et al, 2003b). However, one cannot increase
FIGURE 2.10 Strain-balance GaAsP/InGaAs 3period MQW device: a) Schematic diagram showing the layer
configuration; b) Electronic band structure for the device
ch-02
2/6/07
7:36 PM
Page 59
59
FIGURE 2.11 a) External quantum efficiency for the strain-balanced GaAsP/InGaAs MQW sample and
GaAs control; b) Dark IV curves for GaAsP/InGaAs strain-balance MQW cell, a relaxed GaAs/InGaAs MQW cell,
a strained GaAs/InGaAs MQW cell and a GaAs control
ch-02
2/6/07
60
7:36 PM
Page 60
N. J. EKINS-DAUKES
the number of QWs indefinitely. First, increasing the number of QWs places an ever
tighter tolerance on the growth process (Ekins-Daukes et al, 2002), so a growth technique
capable of producing a 20 QW structure may not be suitable for a 100 QW sample.
Second, the necessity to maintain a field across the QW for efficient carrier escape places
a stringent demand on the background doping level in the i-region. The devices discussed
above typically have i-regions approximately 1m wide, which through a simple
electrostatic calculation (Ekins-Daukes, 2000) requires a background doping level below
1 1015cm-3. If the background doping level rises to a point that the MQW is no longer
fully depleted, then the catastrophic loss in photocurrent observed in Figure 2.5 is
experienced. Third, the strained QW and barriers must not exceed their individual critical
thickness (Matthews and Blakeslee, 1974), thereby placing some limits on the well width
and number of bound levels in the QW. Indeed, even when the well and barrier remain
below their critical thickness, strain-induced composition modulation can degrade the
MQW under certain growth conditions (Bushnell et al, 2003b; Bushnell, 2002).
Such problems lead one to conclude that MQW devices could benefit from
light-trapping schemes (Yablonovitch and Cody, 1982; Steinke, 1996; Hepp, 1998). One
effective technique has been to incorporate a distributed Bragg reflector (DBR) into the
structure (Bushnell et al, 2003a), thereby doubling the optical path through the QW layers.
To approach unity absorption, however, the optical path length needs to be extended
further. Textured coatings and grating structures have been proposed to increase the
optical path through the QWs and some practical means for their formation investigated
(Bushnell et al, 2002; Bushnell, 2002).
ch-02
2/6/07
7:36 PM
Page 61
61
showed that the proton irradiation introduced a mid-gap trap that is likely to be associated
with the QW or QW Barrier interfaces (Walters et al, 2000b and 2001). Further work on the
radiation resistance of MQW solar cells is required, especially on GaAs-based QW
devices, which could usefully enhance the GaInP/GaAs/Ge multi-junction space solar cell,
discussed in section 2.4.1.
2/6/07
62
ch-02
7:36 PM
Page 62
N. J. EKINS-DAUKES
(Dimroth et al, 2000), the advantage being that strain relaxation and the associated
degradation in material quality due to dislocations are avoided altogether.
To give an indication of the performance of these strain-balanced structures in a
tandem solar cell, Table 2.1 shows the projected Voc for both the lower cell and the
tandem structure, together with the Jsc and the tandem efficiency values. The calculation
is based on published data taken from a GaInP top cell (Yang et al, 1997) and a variety of
MQW candidates for the lower junction (Bushnell et al, 2003a; Ekins-Daukes, 2000).
Values for J0 are estimated from Jsc and Voc as described earlier, using equation 4. The
GaInP junction is modelled with J0 =5 1023mAcm2 and can deliver 20mAcm-2, making
the 2J cell current limited by the lower junction in all cases.
The GaInP/GaAs efficiency provides the baseline for comparison at 26 per cent and is
in reasonable agreement with experimental values (Takamoto et al, 2001). An improved
performance is achieved with both the GaInP/SB MQW and the GaInP/DBR SB-MQW, the
ch-02
2/6/07
7:36 PM
Page 63
63
TABLE 2.1 AM0 Junction parameters for a GaInP top cell together with parameters for lower junction
cells composed of GaAs p/n, strain-balanced MQW, DBR enhanced strain-balanced MQW, strained MQW and
relaxed MQW; the thick GaInP cell is modelled with J0 = 5 1023 mAcm2 and can deliver 20mAcm2,
making the 2J cell current limited by the lower junction in all cases.
GaAs p/n
SB-MQW
DBR SB-MQW
Strn. MQW
Rlxd. MQW
V0C/V
1.006
0.920
0.923
0.831
0.611
EFF.
27.1%
26.1%
28.8%
25.9%
23.2%
latter giving an efficiency in excess of 28 per cent AM0. These figures illustrate the relative
importance of an increase in Jsc over a loss in lower junction Voc. The GaInP/strained and
GaInP/relaxed MQW cell experiences an efficiency loss, due to the excessive dark current
evident from Figure 2.11b.
The figures projected in Table 2.1 look encouraging for the MQW cell as a component
of a multi-junction solar cell. However, the efficiency estimates are somewhat generous
to the MQW in terms of the fill factor. While J0 has been chosen to fit the Jsc and Voc, it
has been assumed that the diode ideality is unity, which is reasonable for p/n devices and
yields a fill factor around 90 per cent for the structures described in Table 2.1. However,
in practice the diode ideality is closer to two for MQW and p-i-n devices under 1 sun
illumination. This serves to reduce the fill factor to around 82 per cent and reduces the
efficiency advantage for the DBR cell to a negligible margin.
Before the introduction of the strain-balance technique, the MQW Voc was the limiting
factor in a multi-junction device. Now the strain-balance MQW devices give good Voc, so
work should proceed to raise the fill factor and identify the processes responsible for the
high MQW diode ideality. In addition, the absorption in the strain-balanced structure is still
quite weak, even with the DBR structure, and the QW current could be raised to match
the GaInP cell by employing the more sophisticated light-trapping techniques discussed
in section 2.3.3.
2.4.2 THERMOPHOTOVOLTAICS
In a thermophotovoltaic system the radiative emission from a heat source is collected
using a photovoltaic device (Coutts and Fitzgerald, 1998). The heat source is usually
followed by a selective emitter, which radiates strongly only over a narrow energy band.
It is therefore desirable to match the solar cell to the selective emitter, and for this
purpose the band-gap engineering offered by QW structures can be useful (Connolly and
Rohr, 2003). The first practical demonstration of a QW thermophotovoltaic cell was based
on lattice matched InGaAs QWs on InP (Griffin et al, 1997 and 1998) and showed an
absorption edge at 1.55m. The InGaAsP material system is extremely versatile as it offers
a wide range of band gaps that are lattice matched to InP. A particular virtue of this
material is that both the QWs and barriers can be grown from InGaAsP. This facilitates the
ch-02
2/6/07
64
7:36 PM
Page 64
N. J. EKINS-DAUKES
MOVPE growth process as neither the group III nor group V sources are shut off
completely, as is necessary when growing InP/InGaAs structures. InGaAsP MQW devices
can be designed to possess an absorption edge between 0.9 and 1.6m (Rohr et al, 1998,
1999 and 2000a). Such devices are particularly suitable for erbia selective emitters
operating at temperatures around 2000K.
There are some environmental advantages to lowering the selective emitter
temperature to 1500K, and for this purpose an emitter operating at longer wavelengths is
required (Abbott et al, 2002). Suitable emitters for this temperature include thulia and
holmia, requiring photovoltaic cells that can absorb emission at 1.75m and at 1.95m
respectively. Strained InGaAs QWs on InP are suitable for these long wavelengths
(Serdiukova et al, 1997; Freundlich, 2000) and the same strain-balance technique can be
used to ensure that the device is free of dislocations (Rohr, 2000; Rohr et al, 2000b and
2002b). A strain-balanced InGaAs/InGaAs MQW device grown on an InP substrate has
recently been demonstrated, with an absorption edge at 2m (Rohr et al, 2002a).
ch-02
2/6/07
7:36 PM
Page 65
65
ch-02
2/6/07
66
7:36 PM
Page 66
N. J. EKINS-DAUKES
2.6 CONCLUSION
Over the last 15 years a considerable quantity of work has been performed on the QW
solar cell and this marks one of the first attempts to apply low dimensional structures to
photovoltaics. The ability to engineer the device band gap has led to practical applications
in multi-junction solar cells and thermophotovoltaics, while the strain-balance technique
broadens the range of materials whose properties can be exploited in MQW or
superlattice structures. To date only the band gap has been deliberately engineered in QW
solar cell structures, but it is likely that other properties, such as the optical transitions and
phonon modes, will also be controlled in order to produce highly efficient devices in the
future.
ACKNOWLEDGEMENTS
Extensive discussions with Keith Barnham, Massimo Mazzer and Jenny Nelson are
gratefully acknowledged. David Bushnell kindly provided recent experimental data and
Martin Green and Peter Wrfel are thanked for useful discussions concerning efficiency
limits. The Japan Society for the Promotion of Science (JSPS) and Masafumi Yamaguchi
are thanked for their respective financial and scientific support.
ch-02
2/6/07
7:36 PM
Page 67
67
NOTE
1 The contamination is most likely to be due to C which is inevitably incorporated into AlGaAs during MOVPE growth on
account of the strong AlC bond.
REFERENCES
Abbott, P., Rohr, C., Connolly, J. P., Ballard, I., Barnham, K., Ginige, R., Corbett, B., Clarke, G., Bland, S. W. and Mazzer, M. (2002)
A comparative study of bulk InGaAs and InGaAs/InGaAs straincompensated quantum well cells for thermophotovoltaic
applications, proceedings of 29th IEEE PV Specialists Conference, IEEE, New Orleans, US, May, pp10581061
Anderson, N. G. (1995) Ideal theory of quantumwell solar cells, J. Appl. Phys., vol 78, no 3, pp18501861
Anderson, N. (2002) On quantum well solar cell efficiencies, Physica, vol E14, no 12, pp126131
Arajo, G. and Marti, A. (1994) Absolute limiting efficiencies for photovoltaic energy conversion, Solar Energy Materials and
Solar Cells, no 33, pp213240
Arajo, G. L. and Marti, A. (1995) Electroluminescence coupling in 24 multiple quantum well diodes and solar cells, Appl.
Phys. Lett, vol 66, no 7, pp894895
Arajo, G., Marti, A., Ragay, F. and Wolter, J. (1994) Efficiency of multiple quantum well solar cells, proceedings of 12th
European Photovoltaic Solar Energy Conference, H. S. Stephens & Associates, Bedford, UK, pp14811484
Ballard, I., Barnham, K. W. J., Connolly, J. P., Nelson, J., Rohr, C., Roberts, C., Roberts, J. and Button, C. (2001) The effect of
temperature on the efficiency of multi quantum well cells for solar and thermophotovoltaic applications, proceedings of
17th European Conference and Exhibition on PV Solar Energy, WIP and ETA, Munich, Germany and Florence, Italy, pp4144
Ballard, I., Barnham, K. W. J., Nelson, J., Connolly, J. P., Roberts, C., Roberts, J. S. and Pate, M. A. (1998) The effect of
temperature on the efficiency of multiquantum well solar cells, proceedings of 2nd World PV Energy Conversion
Conference European Commission, Ispra, Italy and Vienna, Austria, July, pp36243626
Barnes, J. (1994) An experimental and theoretical study of GaAs/InGaAs quantum well solar cells and carrier escape from
QWs, Ph.D. thesis, Imperial College, University of London, available from www.sc.ic.ac.uk/~q_pv/
Barnes, J., Ali, T., Barnham, K. W. J., Nelson, J. and Tsui, E. S. M. (1994) Gallium arsenide/indium gallium arsenide
multiquantum well solar cells, proceedings of 12th European PV Solar Energy Conference, H. S. Stephens & Associates,
Bedford, UK, pp13741377
Barnes, J., Nelson, J., Barnham, K., Roberts, J., Pate, M., Grey, R., Dosanjh, S., Mazzer, M. and Ghiraldo, F. (1996)
Characterization of GaAs/InGaAs QWs using photocurrent spectroscopy, J. Appl. Phys., no 79, pp77757779
Barnham, K. (1992) Solar Cell, UK Patent Application PCT/GB92/01913, EC Patent Application 92921571.3, US Patent
Application 211850, Australian Patent Application 27905/92
Barnham, K. W. J. and Duggan, G. (1990) A new approach to highefficiency multiband-gap solar cells, J. Appl. Phys., no 67,
pp34903493
Barnham, K., and Vvedensky, D. (2001) Low Dimensional Semiconductor Structures: Fundamentals and Device Applications,
Cambridge University Press, Cambridge, UK
Barnham, K. W. J., Braun, B., Nelson, J., Paxman, M., Button, C., Roberts, J. S. and Foxon, C. T. (1991), Shortcircuit current
and energy efficiency enhancement in a low dimensional structure photovoltaic device, Appl. Phys. Lett., no 59, pp135137
Barnham, K. W. J., Connolly, J., Griffin, P., Haarpaintner, G., Nelson, J., Tsui, E., Zachariou, A., Osborne, J., Button, C., Hill, G.,
Hopkinson, M., Pate, M., Roberts, J. S. and Foxon, T. (1996) Voltage enhancement in quantum well solar cells, J. Appl.
Phys., vol 80, no 2, pp12011206
ch-02
2/6/07
7:36 PM
68
Page 68
N. J. EKINS-DAUKES
Barnham, K., Marques, J. L., Hassard, J. and OBrien, P. (2000) Quantum dot concentrator and thermodynamic model for the
global redshift, Appl. Phys. Lett., vol 76 no 9, pp11971199
Barnham, K., Ballard, I., Connolly, J., Ekins-Daukes, N., Kluftinger, B., Nelson, J. and Rohr, C. (2002) Quantum well solar cells,
Physica, vol E14, no 12, pp2736
Bastard, G. (1988) Wave Mechanics Applied to Semiconductor Heterostructures, Les Editions de Physique, les Ulis
Bedair, S., Roberts, J., Jung, D., Moody, B., El-Masry, N. and Katsuyama, T. (2000) GaAsP/InGaAsN strained layer
superlattices for solar cell applications, proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp12691272
Bremner, S., Corkish, R. and Honsberg, C. (1999) Detailed balance efficiency limits with quasi-Fermi level variations, IEEE
Trans. Elect. Dev., vol 46, no 10, pp19321939
Brown, A. and Green, M. (2003) Intermediate band solar cell with many bands: Ideal performance, J. Appl. Phys., vol 94,
no 9, pp61506158
Bushnell, D. (2002) Optimisation of strain compensated multi-quantum-well solar cells, Ph.D. thesis, Imperial College,
University of London, available from www.sc.ic.ac.uk/~q_pv/
Bushnell, D. B., Barnham, K. W. J., Connolly, J. P., Ekins-Daukes, N. J., Airey, R., Hill, G. and Roberts, J. S. (2002)
Lighttrapping structures for multiquantum well solar cells, proceedings of 29th IEEE PV Specialists Conference, IEEE, New
Orleans, US, May, p1035
Bushnell, D., Ekins-Daukes, N., Barnham, K., Connolly, J., Roberts, J., Hill, G., Airey, R. and Mazzer, M. (2003a) Shortcircuit
current enhancement in Bragg stack multi-quantum-well solar cells for multi-junction space applications, Solar Energy
Materials and Solar Cells, vol 75, no 12, pp299305
Bushnell, D. B., Barnham, K. W. J., Connolly, J. P., Mazzer, M., Ekins-Daukes, N. J., Roberts, J. S., Hill, G., Airey, R. and Nasi, L.
(2003b) Effect of barrier composition and well number on the dark current of quantum well solar cells, proceedings of 3rd
World Conference on Photovoltaic Energy Conversion, Osaka, Japan, May
Chaffin, R. (1987) Quantum Well Multi-Junction Photovoltaic Cell, US Patent 4,688,068
Chaffin, R., Osbourn, G., Dawson and L., Biefeld, R. (1984) Strained superlattice, quantum well, multi-junction photovoltaic
cells, proceedings of 17th IEEE PV Specialists Conference, IEEE, US, pp743746
Connolly, J. (1997) Modelling and optimising GaAs/AlGaAs mqw solar cells, Ph.D. thesis, Imperial College, University of
London, available from www.sc.ic.ac.uk/~q_pv/
Connolly, J. P. and Rohr, C. (2003) Quantum well cells for thermophotovoltaics, Semiconductor Science and Technology, vol
18, no 5, p216
Connolly, J. P., Barnham, K. W. J., Nelson, J., Roberts, C., Pate, M. and Roberts, J. S. (1998) Short circuit current
enhancement in GaAs/AlGaAs MQW solar cells, proceedings of 2nd World PV Energy Conversion Conf. European
Commission, Ispra, Italy, pp36313634
Connolly, J. P., Nelson, J., Barnham, K. and Ballard, I. (2000) Multiple quantum well solar simulator, proceedings of 28th IEEE
PV Specialists Conference, Anchorage, IEEE, US, pp13041307
Connolly, J. P., Nelson, J., Ballard, I., Barnham, K. W. J., Rohr, C., Button, C., Roberts, J. and Foxon, T. (2001) Modelling multi
quantum well solar cell efficiency, proceedings of 17th European Conference and Exhibition on PV Solar Energy, WIP, Munich,
Germany
Corkish, R. and Green, M. (1993) Recombination of carriers in quantum well solar cells, proceedings of 23rd PV Specialists
Conference, IEEE, New York, pp675689
Corkish, R. and Honsberg, C. (1997) Dark currents in double-heterostructure and quantumwell solar cells, proceedings of
26th IEEE PV Specialists Conference, pp955960
Coutts, T. J. and Fitzgerald, M. C. (1998) Thermophotovoltaics, Scientific American, vol 279, no 3, pp9095
Dimroth, F., Schubert, U. and Bett, A. (2000) 25.5% efficient GaInP/GaInAs tandem solar cells grown on GaAs substrates,
IEEE Elect. Dev. Lett., vol 21, no 5, pp209211
ch-02
2/6/07
7:36 PM
Page 69
69
Dingle, R., Gossard, A. and Wiegmann, W. (1975) Direct observation of superlattice formation in a semiconductor
heterostructure, Phys. Rev. Lett., vol 34, no 21, pp13271330
Ekins-Daukes, N. (2000) An investigation into the efficiency of strained and strain-balanced quantum well solar cells, Ph.D.
thesis, Imperial College, University of London, available from www.sc.ic.ac.uk/~q_pv/
Ekins-Daukes, N., Mazzer, M., Barnes, J., Ballard, I., Barnham, K., Holmes, S., Grey, R., Roberts, J., Pate M. and Hill, G. (1998)
New approaches in GaAs/InGaAs quantum well solar cells, proceedings of 2nd World Conf. and Exhibition on Photovoltaic
Solar Energy Conversion, European Commission, Ispra, Italy, p3627
Ekins-Daukes, N., Barnham, K., Connolly, J., Roberts, J., Clark, J., Hill, G. and Mazzer, M. (1999) Strain-balanced
GaAsP/InGaAs quantum well solar cells, Appl. Phys. Lett., vol 75, no 26, pp41954197
Ekins-Daukes, N., Bushnell, D., Zhang, J., Barnham, K. and Mazzer, M. (2000) Strain-balanced materials for highefficiency
solar cells, proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp12731276
Ekins-Daukes, N., Barnes, J., Barnham, K., Connolly, J., Mazzer, M., Clark, J., Grey, R., Hill, G., Pate and M., Roberts, J. (2001)
Strained and strain-balanced quantum well devices for high efficiency tandem solar cells, Solar Energy Materials and
Solar Cells, no 68, pp7187
Ekins-Daukes, N., Kawaguchi, K. and Zhang, J. (2002) Strain-balanced criteria for multiple quantum well structures and its
signature in X-ray rocking curves, Crystal Growth & Design, vol 2, no 4, pp287292
Ekins-Daukes, N., Ballard, I., Calder, C., Barnham, K., Hill, G. and Roberts, J., (2003a) Photovoltaic efficiency enhancement
through thermal up-conversion, Appl. Phys. Lett., vol 82, no 12, pp19741976
Ekins-Daukes, N., Calder, C., Ballard, I., Barnham, K., Nelson, J., Hill, G. and Roberts, J. (2003b) Quantum well and bulk quasiFermi
level behaviour under illuminated conditions, proceedings of 3rd World Conference on Photovoltaic Energy Conversion, Osaka,
Japan, May
Fatemi, N. S., Pollard, H. E., Hou, H. Q. and Sharps, P. R. (2000) Solar array trades between very high-efficency multi-junction
and Si space solar cells, proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp1083804
Freundlich, A. (1998) Strained Quantum Well Photovoltaic Energy Converter, US Patent 5,851,310
Freundlich, A. (2000) Quantum Well Thermophotovoltaic Convertor, US Patent 6,150,604
Freundlich, A. (2002) Multi-Quantum Well Tandem Solar Cell, cont. US Patent 6,372,980
Freundlich, A. and Serdiukova, I. (1998) Multi-quantum well tandem solar cells with efficiencies exceeding 30% AM0,
proceedings of 2nd World Conf. and Exhibition on Photovoltaic Solar Energy Conversion, European Commission, Ispra, Italy,
Vienna, Austria, July, p3707
Freundlich, A., Rossignol, V., Vilela, M., Renaud, P., Bensaoula, A. and Medelci, N. (1994) InPbased quantum well solar cells
grown by chemical beam epitaxy, proceedings of 1st World Conference and Exhibition on Photovoltaic Solar Energy
Conversion, IEEE, US, pp18861889
Friedman, D., Geisz, J., Kurtz S. and Olson, J. (1998) 1ev solar cells with GaInNAs active layer, J. Crystal Growth, no 195,
pp409415
Gevaux, D. G., Green, A. M., Phillips, C. C., Vurgaftman, I., Bewley, W. W., Felix, C. L., Meyer, J. R., Lee, H. and Martinelli,
R. U. (2001) GaInSb/InAs/AlGaAsSb W quantum-well light emitting diodes, Appl. Phys. Lett., vol 79, no 25,
pp40734075
Green, M. (2000) Prospects for photovoltaic efficiency enhancement using low-dimensional structures, Nanotechnology, vol
11, no 4, pp401405
Griffin, P. R., Barnes, J., Barnham, K. W. J., Haarpainter, G., Mazzer, M. Zanotti-Fregonara, C., Grunbaum, E., Olson, C., Rohr,
C., David, J. P. R., Roberts, J. S., Grey, R. and Pate, M. A. (1996) Effect of strain relaxation on forward bias dark currents
in GaAs/InGaAs multiquantum well pin diodes, J. Appl. Phys. no 80, pp58155820
Griffin, P., Ballard, I., Barnham, K., Nelson, J., Zachariou, A., Epler, J., Hill, G., Button, C. and Pate, M. (1997) Advantages of
quantum well solar cells for TPV, in T. J. Coutts, C. S. Allman and J. P. Benner (eds) Thermophotovoltaic Generation of
ch-02
2/6/07
7:36 PM
70
Page 70
N. J. EKINS-DAUKES
Electricity: Third NREL Conference, vol 401 of AIP Conference Proceedings, American Institute of Physics, Woodbury,
New York, Colorado Springs, Colorado, US, May, pp411422
Griffin, P., Ballard, I., Barnham, K., Nelson, J., Zachariou, A., Epler, J., Hill, G., Button, C. and Pate, M. (1998) The application
quantum well solar cells to thermophotovoltaics, Solar Energy Materials and Solar Cells, no 50, pp213219
Hepp, M. (1998) Light trapping in quantum well solar cells grown on patterned substrate, Masters thesis, Imperial College
of Science, Technology and Medicine, University of London., available from www.sc.ic.ac.uk/~q_pv/
Hoffman, R., Fatemi, N., Stan, M., Jenkins P. and Weizer, V. (1998) High efficiency InGaAs on GaAs devices for monolithic
multi-junction solar cell applications, proceedings of 2nd World Conf. and Exhibition on Photovoltaic Solar Energy
Conversion, European Commission, Ispra, Italy, pp36043608
Honsberg, C., Bremner, S. and Corkish, R. (2002) Design tradeoffs and rules for multiple energy level solar cells, Physica, vol
E14, no 12, pp136141
Kettemann, S. and Guillemoles, J. (1995) Limiting efficiency of LDS solar cells, proceedings of 13th European Photovoltaic
Solar Energy Conference, Stephens, Bedford, UK, pp119121
Kettemann, S. and Guillemoles, J. (2002) Thermoelectric field effects in lowdimensional structure solar cells, Physica, vol
E14, no 12, pp101106
Kluftinger, B. (2000) An experimental study of the quasiFermi level separations in quantum well solar cells, Ph.D. thesis,
Imperial College, University of London, available from www.sc.ic.ac.uk/~q_pv/
Kluftinger, B., Barnham, K., Nelson, J., Foxon, T. and Cheng, T. (2000) Temperature dependent study of the quasi-Fermi level
separation in double quantum well pin structures, Microelectronic Engineering, no 5152, pp265274
Kluftinger, B., Barnham, K., Nelson, J., Foxon, T. and Cheng, T. (2001) Temperature dependent study of radiative losses in
double quantum well solar cells, Sol. Energy Mater. Sol. Cells, no 66, pp510509
Koch, S., Meier, T., Hoyer, W. and Kira, M. (2002) Theory of the optical properties of semiconductor nanostructures, Physica,
vol E14, no 12, pp4552
Kojima, N., Goetzberger, O., Ohshita, Y. and Yamaguchi, M. (2000) Approach to the carbonbased superlattice structure solar
cell, proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp873875
Kondepudi, D. and Prigogine, I. (1998) Modern Thermodynamics, John Wiley & Sons
Kurtz, S., Faine, P. and Olson, J. (1990) Modeling of 2junction, series-connected tandem solar cells using top cell thickness as
an adjustable parameter, J. Appl. Phys., vol 68, no 4, pp18901895
Levine, B. (1993) Quantum well infrared photodetectors, J. Appl. Phys., vol 74, no 8, R1R81
Li, N. Y., Sharps, P. R., Stan, M., Newman, F., Hills, J. S., Hou, H. Q., Gee, J. M. and Aiken, D. J. (2000) Development of 1.25
eV InGaAsN for triple junction solar cells, proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp986989
Luque, A. and Marti, A. (1997a) Entropy production in photovoltaic conversion, Phys. Rev. B., vol 55, no 11, pp69946999
Luque, A. and Marti, A. (1997b) Increasing the efficiency of ideal solar cells by photon induced transitions at intermediate
levels, Phys. Rev. Lett., vol 78, no 26, pp50145017
Luque, A., Marti, A. and Cuadra, L. (2001), Thermodynamic consistency of subband-gap absorbing solar cell proposals, IEEE
Trans. Electron Devices, vol 48, no 9, pp21182124
Marti, A., Cuadra, L. and Luque, A. (2000) Quantum dot intermediate band solar cell, proceedings of 28th IEEE PV
Specialists Conference, IEEE, US, pp940943
Matthews, J. W. and Blakeslee, A. E. (1974) Defects in epitaxial multilayers (I), J. Crystal Growth, no 27, pp118125
Mazzer, M., Grnbaum, E., Barnham, K., Barnes, J., Griffin, P., Holt, D., Hutchinson, J., Norman, A., David, J., Roberts, J. and
Grey, R. (1996) Study of misfit dislocations by EBIC, CL and HRTEM in GaAs/InGaAs lattice strained multi-quantum well
pin solar cells, Materials Science and Engineering, no B42, pp4351
McFarlane, S. C., Barnes, J., Barnham, K. W. J., Tsui, E. S. M., Button, C. and Roberts, J. S. (1999) Space charge effects in
carrier escape from single quantum well structures, vol 86, no 9, pp51095115
ch-02
2/6/07
7:36 PM
Page 71
71
Monier, C., Freundlich, A. and Vilela, M. (1999) Oscillator strength of excitons in InGaAs/GaAs QWs in the presence of a large
electric field, vol 85, no 5, pp27132718
Nagashima, T., Okumura, K., Murata, K. and Yamaguchi, M. (2003) A germanium backcontact type cell for
thermophotovoltaic applications, proceedings of 3rd World Conference on Photovoltaic Energy Conversion, Osaka; Japan,
May
Nelson, J. (1995) Quantum well structures for photovoltaic energy conversion, in M. Francome and J. Vossen (eds) Physics
of Thin Films, Academic Press, New York, vol 21, pp311368
Nelson, J. (2001) Quantum well solar cells, in M. D. Archer and R. Hill (eds) Clean Electricity from Photovoltaics, Imperial
College Press, London, pp7187
Nelson, J. (2003) The Physics of Solar Cells, Imperial College Press, London
Nelson, J., Paxman, M., Barnham, K. W. J., Roberts, J. S. and Button, C. (1993) Steady state carrier escape from single
QWs, IEEE Journal of Quantum Electronics, no 29, pp14601467
Nelson, J., Kluftinger, B., Tsui, E., Barnham, K., Roberts J. and Button, C. (1995) QuasiFermi level separation in quantum well
solar cells, proceedings of 13th European Photovoltaic Solar Energy Conference, H. S. Stephens & Associates, Bedford,
UK, pp150153
Nelson, J., Barnes, J., Ekins-Daukes, N., Kluftinger, B., Tsui, E., Barnham, K. W. J., Foxon, T., Cheng, T. and Roberts, J. S.
(1997) Observation of suppressed radiative recombination in single quantum well pin photodiodes, no 82, pp62406246
Nelson, J., Ballard, J., Barnham, K. W. J., Connolly, J., Roberts, J. S. and Pate, M. (1999) Effect of QW location on quantum
well solar cell dark currents, vol 86, no 10, pp58985905
Okada, Y., Seki, S., Takeda, T. and Kawabe, M. (2002) Control of dark currents in multi-quantum well solar cells fabricated by
atomic H assisted molecular beam epitaxy, J. Cryst. Growth, no 237, pp15151518
Osbourne, J. (1994) InGaP/GaAs quantum well solar cells, Masters thesis, Imperial College, University of London
Paxman, M. (1992) An experimental study of GaAs/AlGaAs multiple quantum well solar cells, Ph.D. thesis, Imperial College,
University of London
Paxman, M., Nelson, J., Braun, B., Connolly, J., Barnham, K. W. J., Foxon, C. T. and Roberts, J. S. (1993) Modeling the
spectral response of the quantum well solar cell, no 74, pp614621
Peng, R., Mazzer, M. and Barnham, K. (2003) Efficiency enhancement of ideal photovoltaic solar cells by photonic excitations
in multiintermediate band structures, Appl. Phys. Lett., vol 83, no 4, pp770772
Ragay, F., Wolter, J., Marti, A. and Arajo, G. (1994a) Experimental analysis of GaAsInGaAs MQW solar cells, proceedings of
1st World Conference and Exhibition on Photovoltaic Solar Energy Conversion, IEEE, US, p1754
Ragay, F., Wolter, J., Marti, A. and Araujo, G. (1994b) Experimental analysis of the efficiency of MQW solar cells,
proceedings of 12th European Photovoltaic Solar Energy Conference, H. S. Stephens & Associates, Bedford, UK,
pp14291433
Raisky, O. Y., Wang, W. B., Alfano, R. R., Reynolds, Jr., C.L. Stampone, D. V. and Focht, M. W. (1998) In1-xGaxAs1-yPY /InP
multiple quantum well solar cell structures, J. Appl. Phys., vol 84, no 10, pp57905794
Raisky, O. Y., Wang, W. B., Alfano, R. R. and Reynolds, Jr., C.L. (2001) Carrier screening effects in photoluminescence
spectra of InGaAsP/InP multiple quantum well photovoltaic structures, Appl. Phys. Lett., vol 79, no 3, pp430432
Renaud, P., Vilela, M., Freundlich, A., Bensaoula, A. and Medelci, N. (1994) Modeling p-i-n multi quantum well n solar cells: A
contribution for a near optimum design, proceedings of 1st World Conference and Exhibition on Photovoltaic Solar Energy
Conversion, IEEE, US, pp17871790
Rimada, J.C. and Hernandez, L. (2001a) Modelling of ideal AlGaAs quantum well solar cells, Microelectronics, vol 32, no 9,
pp719723
Rimada, J.C. and Hernandez, L. (2001b) A new approach to ideal AlGaAs MQW solar cells, Mod. Phys. Lett., vol B15, no
1719, pp778781
ch-02
2/6/07
7:36 PM
72
Page 72
N. J. EKINS-DAUKES
Rohr, C. (2000) InGaAsP quantum well cells for thermophotovoltaic applications, Ph.D. thesis, Imperial College, University of
London
Rohr, C., Connolly, J. P., Barnham, K. W. J., Griffin, P. R., Nelson, J., Ballard, I., Zachariou, A., Button, C. and Clark, J. (1998)
InGaAsP quantum well cells for thermophotovoltaics, proceedings of 2nd World Conf. and Exhibition on Photovoltaic Solar
Energy Conversion, Ispra, Italy and Vienna, Austria, July 1998, European Commission, pp230233
Rohr, C., Connolly, J. P., Barnham, K. W. J., Griffin, P. R., Nelson, J., Ballard, I., Button, C. and Clark, J. (1999) Optimisation of
InGaAsP quantum well cells for hybrid solarthermophotovoltaic applications, in T. J. Coutts, J. P. Benner and C. S. Allman
(eds) Thermophotovoltaic Generation of Electricity: Fourth NREL Conf., vol 460 of AIP Conference Proceedings, Denver,
Colorado, US, October 1998, American Institute of Physics, Woodbury, New York, pp8392
Rohr, C., Barnham, K. W. J., Connolly, J. P., Nelson, J., Button, C. and Clark, J. (2000a) Potential of InGaAsP quantum well
cells for thermophotovoltaics, proceedings of 26th International Symposium on Compound Semiconductors, no 166 in
Institute of Physics Conference Series, Berlin, Germany, August 1999, Institute of Physics Publishing, Bristoland,
Philadelphia, US, pp423426
Rohr, C., Connolly, J., Barnham, K. W., Mazzer, M., Button, C. and Clark, J. (2000b) Strain-balanced In0.62Ga0.38As/In0.47Ga0.53As
(InP) quantum well cell for thermophotovoltaics, proceedings of 28th IEEE Photovoltaic Specialists Conference, IEEE, US,
pp12341237
Rohr, C., Abbott, P., Ballard, I., Connolly, J. P., Barnham, K. W., Nasi, L., Ferrari, C., Lazzarini, L., Mazzer, M. and Roberts, J.
(2002a) Straincompensated InGaAs/InGaAs quantum well cell with 2m band-edge, in T. J. Coutts, G. Guazzoni and
J. Luther (eds) The Fifth Conference on Thermophotovoltaic Generation of Electricity. Vol. 653 of AIP Conference
Proceedings, American Institute of Physics, Woodbury, New York, pp344353
Rohr, C., Connolly, J. P., Ekins-Daukes, N., Abbott, P., Ballard, I., Barnham, K. W., Mazzer, M. and Button, C. (2002b)
InGaAs/InGaAs strain-compensated quantum well cells for thermophotovoltaic applications, Physica E: Low-dimensional
Systems and Nanostructures, vol 14, no 12, pp158161
Serdiukova, I., Newman, F., Aguillar, L., Vilela, M. F., Monier, C. and Ferundlich, A. (1997) Strained In0.53Ga0.47As/InxGa1-xAs
(x>0.6) multiquantum well thermophotovoltaic converters, proceedings of 26th IEEE PV Specialists Conference, IEEE, US,
pp963966
Serdiukova, I., Monier, C., Vilela, M. and Freundlich, A. (1999) Critical built-in electric field for an optimum carrier collection in
multi-quantum well pin diodes, Appl. Phys. Lett., vol 74, no 19, pp28122814
Settler, M. and Lundstrom, M. (1994) A detailed investigation of hetero junction transport using a rigorous solution to the
Boltzmann equation, IEEE Trans. Elec. Dev., vol 41, no 4, pp592600
Shockley, W. and Queisser, H. (1961) Detailed balance limit of efficiency of pn junction solar cells, J. Appl. Phys., vol 32,
no 3, pp510519
Steinke, L. (1996) Light trapping in MQW solar cells, Masters thesis, Imperial college, University of London, available at
www.sc.ic.ac.uk/~q_pv/
Stradling, R. and Klipstein, P. (1989) Growth and Characterization of Semiconductors, Adam Hilger, UK
Suzuki, Y., Kikuchi, T., Kawabe, M. and Okada, Y. (1999) Atomic hydrogen assisted molecular beam epitaxy for the fabrication
of multi-quantum well solar cells, J. Appl. Phys., vol 86, no 10, pp58585861
Takamoto, T., Ikeda, E., Kurita, H., Tanabe, T., Tanaka, S., Matsubara, H., Mine, Y., Takagishi, S. and Yamaguchi, M. (1997)
InGaP/GaAs and InGaAs mechanically stacked triplejunction solar cells, proceedings of 26th PV Specialists Conference,
IEEE, pp10311034
Takamoto, T., Agui, T., Ikeda, E. and Kurita, H. (2000) High efficiency InGaP/InGaAs tandem solar cells on Ge substrates,
proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp976981
Takamoto, T., Agui, T., Ikeda, E. and Kurita, H. (2001) High efficiency InGaP/In0.01Ga0.99As tandem solar cells lattice matched to
Ge substrates, Solar Energy Materials and Solar Cells, no 66, pp511516
ch-02
2/6/07
7:36 PM
Page 73
73
Tang, P., Pullin, M., Phillips, C. and Stradling, R. (1995) Photoluminescence studies of NIPI superlattices in InSb and InAs
suppression of Auger recombination due to type-II potentials, Semicond. Sci. Technol., vol 10, no 4, pp476482
Tibbits, T., Ballard, I., Barnham, K., Bushnell, D., Ekins-Daukes, N., Airy, R., Hill, G. and Roberts, J. (2003) The potential for
strain-balanced quantum well solar cells in terrestrial concentrator applications, proceedings of 3rd World Conference on
Photovoltaic Energy Conversion, Osaka, Japan, May
Tobin, S., Vernon, S., Bajgar, C., Woktczuk, S., Melloch, M., Keshavarzi, A., Stellwag, T., Venkatensan, S., Lundstrom, M. and
Emery, K. (1990) Assessment of MOCVD and MBE grown GaAs for high efficiency solar cell applications, IEEE Trans.
Electron Devices, vol 37, no 2, pp469477
Varonides, A. and Berger, A. (1997) Proceedings of 15th European Photovoltaic Solar Energy Conference, H. S. Stephens &
Associates, Bedford, UK, pp17121715
Vegard, L. (1921) Die Konstitution der Mischkristalle und die Raumfllung der Atome, Z. Phys., no 5, pp1726
Walters, R., Summers, G., Messenger, S., Freundlich, A., Monier, C. and Newman, F. (2000a) Radiation hard multiquantum
well InP/InAsP solar cells for space applications, Prog. Photovolt. Res., vol 8, no 3, pp349354
Walters, R., Summers, G., Messenger, S., Romero, M., Arajo, D., Garcia, R., Freundlich, A., Newman, F. and Vilela, M.
(2000b) Electron beam induced current and cathodoluminescence study of proton irradiated InAsP/InP quantum well solar
cells, proceedings of 28th IEEE PV Specialists Conference, IEEE, US, pp13121316
Walters, R., Summers, G., Messenger, S., Romero, M., AlJassim, M., Garcia, R., Arajo, D., Freundlich, A., Newman, F. and
Vilela, M. (2001) Electron beam induced current and cathodoluminescence study of proton irradiated InAsP/InP
quantumwell solar cells, J. Appl. Phys., vol 90, no 6, pp28402846
Weisbuch, C. and Vinter, B. (1991) Quantum Semiconductor Structures: Fundamentals and Applications, Academic Press
Wrfel, P. (1995) Is an illuminated semiconductor far from thermodynamicequilibrium, Sol. Energy Mater. Sol. Cells, vol 38,
no 14, pp2328
Wrfel, P. (2003) Improvement of solar cell efficiencies by impurity transitions, Sol. Energy Mater. Sol. Cells, vol 79, no 1,
pp153161
Yablonovitch, E. and Cody, G. (1982) Intensity enhancement in textured optical sheets for solar cells, IEEE Trans. Electron
Devices, ED29, pp300305
Yamaguchi, M. and Luque, A. (1999) High efficiency and high concentration in photovoltaics, IEEE Trans. Elec. Dev., vol 46,
no 10, pp21392144
Yang, M. J., Yamaguchi, M., Takamoto, T., Ikeda, E., Kurita, H. and Ohmori, M. (1997) Photoluminescence analysis of InGaP
top cells for high-efficiency multi-junction solar cells, Solar Energy Materials and Solar Cells, no 45, pp331339
Zachariou, A. (1996) An experimental study of InP/InGaAs quantum well solar cells, Ph.D. thesis, Imperial College of Science,
Technology and Medicine, University of London, available from www.sc.ic.ac.uk/~q_pv/
Zachariou, A., Barnham, K. W. J., Griffin, P., Nelson, J., Button, C. C., Hopkinson, M., Pate, M. and Epler, J. (1996) A new
approach to pdoping and the observation of efficiency enhancement in InP/InGaAs quantum well solar cells, proceedings
of 25th IEEE PV Specialists Conference, IEEE, US, pp113116
Zachariou, A., Barnes, J., Barnham, K. W. J., Nelson, J., Tsui, E. S. M., Epler, J. and Pate, M. (1998a) A carrier escape study
from InP/InGaAs single quantum well solar cells, J. Appl. Phys., vol 83, no 2, pp877881
Zachariou, A., Barnham, K. W. J., Griffin, P., Nelson, J., Osborne, J., Hopkinson, M. and Pate, M. (1998b) GaInP/GaAs
quantum well solar cells, proceedings of 2nd World Conf. and Exhibition on Photovoltaic Solar Energy Conversion,
European Commission, Ispra, Italy, pp223226
ch-03
2/6/07
8:21 PM
Page 74
Abstract
This article briefly reviews and summarizes recent key developments in organic and
polymeric photovoltaic materials and devices. Specifically, the article briefly reviews the
four major stages and/or types of the organic and polymeric photovoltaic developments
from single layer Schottky junction cell, to donor/acceptor bilayer Tang cell, to
donor/acceptor blend heterojunction cell to donor/acceptor bicontinuous phase
separated nano-structured cell. The current relatively low photoelectric power
conversion efficiencies (less than 6 per cent) of purely organic and polymeric
photovoltaic materials and devices are attributed mainly to the heavy photon loss, the
exciton loss and carrier loss. However, high efficiency can be achieved as these
losses can be addressed via optimizations in both space and energy domains as
discussed. Organic versus inorganic photovoltaics is also briefly compared.
Keywords organic and polymeric photovoltaics; plastic solar cells; organic and inorganic
semiconductors; band theory; excitons; light harvesting; portable power
3.1 INTRODUCTION
While inorganic crystalline-based solar cell technology has become relatively mature
over 30 per cent photoelectric power conversion efficiencies have been demonstrated,1-4
and much higher theoretical efficiencies have been predicted3 this technology
suffers from a relatively high manufacturing cost and a current severe shortage of the
feedstock materials. As the need for renewable and solar energy technologies is rapidly
growing, alternative materials or technologies that could reduce the solar cell
manufacturing costs and have sufficient feedstock supplies become very attractive and
critical. Organic or polymeric photovoltaic materials and technology fall right into this
category.57
In comparison to the traditional inorganic solar cell, recently developed organic and
polymeric conjugated semiconducting materials appear very promising for photovoltaic
applications for several reasons:
ch-03
2/6/07
8:21 PM
Page 75
75
their lightweight, flexible shape, versatile materials synthesis and device fabrications,
The Sun
Future Camping/Hiking
FIGURE 3.1 Lightweight and flexible thin film plastic solar cells are very attractive for camping tents or
any mobile units.
2/6/07
76
8:21 PM
Page 76
S-S. SUN
the
room
temperature
thermal
energy
(the
top
diamond
curve
in
Figure 3.2), in other words a secondary force is needed to dissociate the photogenerated
exciton. This is called the secondary photo carrier generation process (or excitonic
mechanism) and will be discussed in the following section. In fact, the initial sunlight
harvesting steps of photosynthesis in natural plants usually follow the secondary process
more closely.10 However, as shown in Figure 3.2, VB to CB band-to-band-like photo carrier
generation (also called the primary photo carrier generation process or SSH model8) is also
possible if the orbital overlap is really good, such that the exciton radius is very large, or the
dielectric constant of the material is very large so the exciton Coulombic binding energy is
below the thermal energy level. A Wannier type of exciton curve affiliated with large dielectric
constant materials such as silicon crystal is also shown (as triangles) in Figure 3.2. Due to the
typical small band size (BS), the excitation energy gap Eg in organic or polymeric materials only
represents the energy difference between the relatively localized LUMO and HOMO, not the
traditional delocalized VB and CB. When a photon with energy equal to or over Eg excite the
organic or polymeric molecule, an electron first transfers from the HOMO to the LUMO and
then quickly relaxes with the hole to form a Frenkel type exciton, also called polaron-exciton
in organics.8 Depending on the dielectric constant of the materials, the Frenkel type exciton
typically has a size of less than 2nm with binding energies larger than 0.1eV.
kT
100
10
Coulombic
Energy E (eV)
ch-03
0.1
0.01
0.001
0
10
20
30
40
FIGURE 3.2 Scheme of exciton Coulombic binding energy (in log scale) versus exciton radius.
50
ch-03
2/6/07
8:21 PM
Page 77
77
ch-03
2/6/07
8:21 PM
78
Page 78
S-S. SUN
excitons can only travel 550nm (also called average exciton diffusion length, AEDL),
much less than the thickness of most PV materials films, which are typically over 100nm;
most excitons are therefore wasted. This exciton loss is a major loss mechanism. Those
early cells are also called Schottky cells or defect cells due to the carriers mainly being
generated in the Schottky junction or defect sites. These single layer solar cells may be
called first generation cells.
Schottky
Junction
Vacuum Level
SWFE
LWFE
Schottky Junction
hv
LUMO
hv
ex
em
ex
E
SWFE
LWFE
HOMO
(a)
(b)
FIGURE 3.3 First generation single layer type organic photovoltaic cell or Schottky cell with (a) device
structure and (b) energy diagram
ch-03
2/6/07
8:21 PM
Page 79
79
recombination is also relatively small due to electrons and holes now moving in two
separate layers or domains. However, a major limiting factor for the double layer cell is still
the relatively thick donor or acceptor layer (typically over 100nm) versus the relatively short
average exciton diffusion length (550nm), in other words many photogenerated excitons
decay before they reach the D-A interface (Figures 3.4a and b). On the other hand, if the
D/A Interface
Vacuum Level
SWFE
LWFE
D/A Interface
hv
D-LUMO
hv
ex
A-LUMO
em
ex
Donor
Acceptor
em
ex
D-HOMO
A-HOMO
(a)
(b)
FIGURE 3.4 Second generation donor/acceptor bilayer type organic photovoltaic cell or Tang cell with (a)
device structure and (b) energy diagram
N N
N M N
O
n
PPV
n
Phthalocyanines
O
MDMO-PPV
MEH-PPV
O
S
O
RO-PPV-10
n
R
PTh
PBZT (D)
ch-03
2/6/07
8:21 PM
80
Page 80
S-S. SUN
R N
O
R
C60
Perylenes
R
SO2
OR
NC
RO
PCBM
CN-PPV
BBL
SF-PPV
vacuum
2
vacuum
Donor (hole acceptor)
MDMO MEH
PPV
-PPV -PPV
2.7 eV
-
C10O
-PPV
-2
2.6 eV
LUMOs
Ca
-2.83 eV -2.80 eV
-2.91 eV
PCBM C
60
SF-PPV
CN-PPV
3.83 eV
3.43 eV
3.60 eV
4
Al
4.3 eV
ITO
4.7 eV
PSS-PEDOT
5.2 eV
Au
5.1 eV
5.17 eV 5.20 eV
5.20 eV
5.31 eV
-5
4.90 eV
5.22eV
5.29 eV
HOMOs
5.80 eV 5.89 eV
-6
6.10 eV 6.10 eV
FIGURE 3.7 Frontier orbital levels of representative polymeric electron donors and acceptors; the work
functions of several representative electrodes are also shown on the sides
film thickness is too thin, this results in severe photon losses. Figure 3.5 shows mostly
used or studied organic donors. Figure 3.6 shows mostly used organic acceptors, and
Figure 3.7 shows the frontier HOMO/LUMO levels of some commonly used conjugated
ch-03
2/6/07
8:21 PM
Page 81
81
semiconducting polymers. The donor/acceptor (or p/n) double layer solar cells may be
called second generation organic cells.
ch-03
2/6/07
82
8:21 PM
Page 82
S-S. SUN
LWFE
SWFE
Vacuum
D-LUMO
A-LUMO
SWFE
LWFE
(a)
D-HOMO
A-HOMO
(b)
FIGURE 3.8 Third generation donor/acceptor blend type organic photovoltaic cells or bulk heterojunction
cells with (a) device structure and (b) energy diagram
ch-03
2/6/07
8:21 PM
Page 83
83
A
SWFE
LWFE
Vacuum
D/A Interface
D-LUMO
A-LUMO
D-HOMO
A-HOMO
LWFE-Femi
SWFE-Femi
(b)
(a)
FIGURE 3.9 Fourth generation donor/acceptor bicontinuous phase separated and nano-structured organic
photovoltaic cell with (a) device structure and (b) energy diagram
ch-03
2/6/07
84
8:21 PM
Page 84
S-S. SUN
can be added up. In this way, large solar cell voltages together with a high power
conversion efficiency can be achieved.8 For instance, over 30 per cent power conversion
efficiencies have been reported in inorganic triple-junction tandem-type solar cell.14 For
organics, a 5.7 per cent power conversion efficiency (under 1 sun AM 1.5 condition) and
an open-circuit voltage of 1.2 volts was reported in a tandem-type PV device containing
two serially connected and stacked sub-cell units.56 In this device, though both sub-cells
are composed of copper phthalocyanine (CuPc) as the donor and C60 as the acceptor, the
front sub-cell was tuned (via optical engineering) to absorb mainly the 650nm wavelength
low energy light, while the back sub-cell mainly captures the 450nm wavelength high
energy photons due to the optical match issue.56 In an ideal tandem cell, however, it is
desirable that the front sub-cells absorb high energy photons while the rear sub-cells
capture low energy light.2, 3, 51, 52 In this way, the largest-gapped cell at the front would
capture highest energy photons first, but allow lower energy photons to pass through to
the lower-gapped cells behind, where the lower energy photos can be captured by lower
energy gap sub-cells, and so on. Even if some excitons in the front large-gapped cells do
not dissociate and relax to emit a smaller energy photon, that photon can be captured by
the next lower-gapped cells. However, current density needs to be balanced between subcells.13, 51, 52 As a whole, as the cell could capture most sunlight, it should appear as a dark
color.7
ch-03
2/6/07
8:21 PM
Page 85
85
Light
FIGURE 3.10 General scheme of an organic/polymer solar cell with buffering layers
O
S
+
S
O
+
S
S
O
S
O
O
OH
O
O S OO S OO S O
ch-03
2/6/07
86
8:21 PM
Page 86
S-S. SUN
also possible that PSS-PEDOT simply prevents the acceptor from being in direct contact
with the ITO, so charge recombination at the ITO interface is minimized. However,
PSS-PEDOT layer thickness should be optimized since too thin a layer may yield many pin
holes, while thicker layers cause larger increase in the series resistance.61
Insertion of a thin layer of lithium fluoride (LiF) between the active layer and the SWFE
metal electrode has been shown to improve electron injection in some organic LEDs.62
The LiF layer is typically very thin (< 1nm), since thicker layers are found to be detrimental
to electron injection. For instance, 3 thickness of LiF was found to be optimal for both
the fill factor and the current density in one system.62 While the exact causes were not
very clear, the improved device optoelectronic performance of incorporating LiF in some
cases may be attributed to, for instance, the LUMO of the LiF lying between the acceptor
LUMO and work function of the metal, or the LiF-modified work function of the metal
electrode due to the dipolar nature of LiF layer. Both would facilitate the electron transfer.
Another cause might be that the LiF layer prevents the donor from being in direct contact
with the metal, or prevents the chemical reactions between the organic active layer and
the metal, or reduces the serial resistance of the interface between the metal and the
active layer.61
The equivalent circuit of an organic/polymeric solar cell J-V curves can be
approximated by:63
(1)
where J0 is the saturation dark current density, q is the elementary charge, n is the diode
ideality factor, V is the applied voltage of the cell, RS and RP are the serial and parallel
resistivity, Jsc is the short-circuit photocurrent density (A/cm2), k is the Boltzman constant,
and T is the absolute temperature. For a pin-hole free organic active layer, the parallel
(shunt) resistivity was approximately the same for many organic photodiodes (typically
over 1k), but the total serial resistivity (RS) may vary depending on materials and device
fabrications. In one device, for instance, Rs decreased from 10 with no LiF layer to 4
for LiF thickness between 3 and 9 thick and increased to 5 for LiF 1215 thick for one
cell.63 In comparison to inorganic solar cells, organics typically have very large RS due to
the typically very poor orbital overlaps.
When characterizing and comparing organic solar cells, like in inorganic solar cells, the
open-circuit voltage (Voc), short-circuit current density (Jsc), the fill factor (FF), and the
overall photoelectric power conversion efficiencies () are critical parameters. Like in
inorganic solar cells, the open-circuit voltage (Voc) can be experimentally obtained from the
currentvoltage curves of an illuminated device when the current is zero. Because of the
special charge generation and separation mechanisms in organic devices, the efficiencylimiting factors are therefore distinct from those in conventional inorganic solar cells (for
example silicon p-n junction solar cells). For instance, while the maximum photovoltages
(Voc) achievable in silicon cells are generally limited to the magnitude of the built-in
potential, it is common to observe experimentally Voc greater than the built-in potential in
organic-based photovoltaic devices.55 It is believed the Voc of an organic donor/acceptor
binary cell is closely correlated to the HOMO of the donor and the LUMO of the
ch-03
2/6/07
8:21 PM
Page 87
87
acceptor.55, 64 The short-circuit current density (Jsc) is the photocurrent density value under
zero applied bias. It is a result of the photo-induced charge separation (or exciton
dissociation) and the charge transport driven by the internal field and the chemical
potentials of the material.
The photoelectric power conversion efficiency (PCE or PCE) is the ratio of cell
maximum output electrical power (Pout) over the total radiated input optical power (Pin):
(2)
where a fill factor (FF) is related to the percentage of maximum electrical power that can
be extracted from the solar cell. The fill factor for devices is described as
(3)
where Jm and Vm are the values for the current density and voltage at maximum power
Pout/max = (JV)max.
To achieve a high FF factor in organic solar cells, it is desirable that the parallel
resistivity is large to prevent short or leakage current, while the serial resistivity is small to
enable a large forward current. The series resistivity simply adds up from all series
resistive contributions in the device, including both carrier transport in materials and
interface charge transfer. One key problem in organics is the very high series resistance
due to poor molecular frontier orbital overlaps, so charges have to hop between
molecules instead of diffuse in bands. However, molecular self-assembly may solve this
problem in the future.
For all currently reported organic/polymeric photovoltaic materials and devices, none of
the above mentioned five steps have been optimized. For instance, in the first step, the
photon loss is heavy due to most organic semiconductors having too high energy gaps
(typically larger than 2eV), meaning that only a very small fraction of the sunlight could be
captured. In the second step, the exciton loss is severe unless a phase separated and
ordered nano-structure can be materialized. In the third step, both the exciton loss and
carrier loss are severe unless the frontier orbital energy offsets are at their optimal value
ch-03
2/6/07
88
8:21 PM
Page 88
S-S. SUN
so that the exciton dissociation is maximized and at the same time the charge
recombination is minimized. In the fourth step, the carrier loss is notoriously severe and
well known in most organic/polymeric semiconductors due to the typically poor orbital
overlaps, so that sizable bands are difficult to form, and carrier transport is mainly through
inter- and intra-molecular hopping mechanisms. In the last step, the carrier loss may be
severe if the energy levels or chemical states between the organic semiconductors and
the electrodes are not optimized. It is thus not surprising that the power conversion
efficiencies of most currently reported organic or polymeric solar cells are relatively small
in comparison to their inorganic counterparts. However, there is plenty of room to improve,
and the optimizations should be done in both energy and spatial domains.
ch-03
2/6/07
8:21 PM
Page 89
89
less susceptible to distortion of the main chain conjugation (Figures 3.12 and 3.13). Since
both donor and acceptor blocks are electron conjugated chains, if they are selfassembled in planes perpendicular to the molecular plane like a stacking morphology
well known in a typical conjugated system (Figure 3.13),7173 good carrier transport in
both donor and acceptor phases would become feasible.
Source: S. Sun, Photovoltaic Devices Based on a Novel Block Copolymer, US patent publication # 20040099307
D
A
Source: S. Sun, Photovoltaic Devices Based on a Novel Block Copolymer, US patent publication # 20040099307
FIGURE 3.13 Scheme of a potential DBAB- type block copolymer secondary structure
While the DBAB- block copolymer backbone structure may be called primary structure
(Figure 3.12), the conjugated chain orbital closely stacked and ordered morphology may
be called secondary structure (Figure 3.13). This secondary structure has been known to
possess dramatically enhanced charge carrier mobility as demonstrated in, for instance,
ch-03
2/6/07
90
8:21 PM
Page 90
S-S. SUN
D
A
D
D
A
D
A
A
D
A
FIGURE 3.14 Scheme of a potential DBAB- type block copolymer tertiary structure
ch-03
2/6/07
8:21 PM
Page 91
91
(4)
Here X=D (donor) or A (acceptor). The parameter Yeq reflects to a certain degree the
efficiency of exciton
>charge conversion. It was experimentally observed that the charge
separation could be orders of magnitude faster then the exciton decay in MEHPPV/fullerene binary systems.33, 34 Secondly, the charge separations (steps 3 and 7) are
ch-03
2/6/07
92
8:21 PM
Page 92
S-S. SUN
also competing with charge recombination (step 4). The ratio of charge separation rate
constant (ks) over charge recombination rate constant (kr) can therefore be defined as
recombination quenching parameter (RQP, mathematically represented as Yrq):
YrqX = ksX / kr
(5)
For any light harvesting applications, such as solar cell applications, it is desirable that
both Yeq and Yrq parameters are as large as possible.
From semi-classical electron transfer theory, when LUMO orbital energy offset E is
set as variable (Figure 3.15),8389 and based on measured, calculated and estimated
parameters of RO-PPV/SF-PPV-I donor/acceptor pairs developed recently,42, 45 a plot of
normalized YeqD, kr, and YrqD versus E is shown in Figure 3.16.51, 52 As Figure 3.16 shows,
when E varies, kr, YeqD and YrqD all exhibit their own maximum values. For the RO-PPV/SFPPV-I pair, the fastest photo-induced charge separation occurs when the RO-PPV/SF-PPVI LUMO offset (E, driving force) equals the sum of charge separation reorganization
energy and the donor exciton Coulombic potential energy.51, 52 Also, the fastest charge
recombination occurs at a LUMO offset far away from the optimum charge separation
offset as well as the actual RO-PPV/SF-PPV-I offset. Therefore, the charge recombination
in the RO-PPV/SF-PPV-I pair does not seem to be of a major concern as long as the LUMO
offset is nearby the EeqD. Figure 3.16 also shows the recombination quenching parameter
YrqD (ksD/kr) does not reach its maximum until a positive energy offset. At this positive
energy offset, the photo-induced charge separation might be too slow to be attractive for
efficient photovoltaic function; therefore, the positive ErqD value appears not critical in
this particular case. It is desirable that through careful molecular design, the ErqD is
Vacuum
3
D-LUMO
EA
A-LUMO
IP
1
4
6
D-HOMO
A-HOMO
D/A Interface
FIGURE 3.15 Scheme of molecular frontier orbitals and photo-induced electron transfer as well as Dexter
energy transfer processes in a donor/acceptor binary light harvesting system
2/6/07
8:21 PM
Page 93
Kr
ch-03
3.5
2.5
1.5
Yeq(D)
0.5
93
Yrq(D)
0.5
1.5
FIGURE 3.16 Donor RO-PPV exciton quenching parameter (Yeq(D)=ks(D)/kd(D), middle solid curve), charge
recombination rate constant (Kr, left long dashed curve), and charge recombination quenching parameter
(Yrq(D)=ks(D)/kr(D), right short dashed curve) versus LUMO offset of RO-PPV/SF-PPV-I pair
coincident with or close to EeqD, and that ErD is far away from EeqD. Similar results can
also be deducted for the acceptor and for the donor/acceptor pair.44, 52
This donor/acceptor charge separation model can readily be used in the impurity or
defect photo doping cases (including the single layer Schottky junction cell), as the
impurities or structural defects can be either donor or acceptor type in the energy domain.
The model can also be used in ground state chemical doping cases.7 Chemical doping
typically refers to a mobile or free charge carrier generation phenomenon when one
material is mixed or in direct contact with a second material at room temperature without
the intentional use of strong external excitation forces such as light or high heat. This is
sometimes also called ground state charge separation. However, there are two different
situations involved. In the first situation, the acceptor LUMO is lower then the donor
HOMO. The energy offset E between the D-HOMO and A-LUMO therefore acts as a key
driving force (in addition to other potential driving forces such as thermal force kT) for the
electron transfer from the donor HOMO directly to the acceptor LUMO. In this case, the
most efficient or optimal charge separation would occur when the energy offset E plus
the thermal force kT equals the charge separation reorganization energy and the attractive
Coulomb force. Even at absolute zero temperature when kT=0, the electron transfer still
occurs due to the presence of the driving E. However, in the second scenario, the
acceptor LUMO might be the same or a little higher then the donor HOMO and the
thermal driving force kT becomes critical. In this case, the most efficient electron transfer
would occur when the driving kT is equal to the charge separation reorganization energy
plus the Coulomb attractive force plus the D-HOMO/A-LUMO energy offset increase.
Because of the kT driving force involved, this process actually is not really ground state
electron transfer, though it may occur at room temperature. Thermally induced chemical
doping mechanism may also be used to explain static electrical charge generation when
two different materials are rubbed against each other, as the friction at the interface might
ch-03
2/6/07
94
8:21 PM
Page 94
S-S. SUN
generate enough thermal energy to incur electron transfer. In chemical doping cases,
since the dopant is typically a minority component and it traps one charge carrier, the
opposite charged carrier at non-dopant majority phase then becomes the mobile or free
charge carrier and can transport at the majority component phase.
ch-03
2/6/07
8:21 PM
Page 95
95
the light radiation direction appears ideal. This would enable a broad capture of most solar
photons and the summation of induced photovoltages.
With optimizations in both space and energy/time domains, it is expected that high
efficiency photoelectric power conversion efficiency organic light harvesting systems,
including organic photovoltaic cells, photo detectors or any artificial photo-charge
synthesizers/converters, can be realized. The dream of renewable, clean, inexpensive,
portable and low cost energy supply can be a reality.
ACKNOWLEDGEMENTS
The author would like to thank all those involved in the research cited in the literature and
for the research/educational grant supports from a number of funding agencies including
NASA, the Air Force Office of Scientific Research, the National Science Foundation, the
Department of Education (Title III award) and the Dozoretz foundation.
REFERENCES
1 M. D. Archer and R. Hill (eds) (2001) Clean Electricity From Photovoltaics, Imperial College Press, London
2 A. Luque and S. Hegedus (eds) (2003) Handbook of Photovoltaic Science and Engineering, John Wiley, Chichester, UK
3 M. A. Green (2003) Third Generation Photovoltaics: Advanced Solar Energy Conversion, Springer, Berlin
4 A. Hepp, S. Bailey and R. Raffaelle (2005) Inorganic photovoltaic materials and devices: Past, present, and future in S.
Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and Devices, CRC Press, Boca Raton, Florida,
US, p19
5 C. Brabec, V. Dyakonov, J. Parisi, and N. Sariciftci (2003) Organic Photovoltaics: Concepts and Realization, Springer, Berlin
6 S. Sun and N. S. Sariciftci (eds) (2005) Organic Photovoltaics: Mechanisms, Materials and Devices, CRC Press, Boca
Raton, Florida, US
7 S. Sun (2006) Organic and polymeric solar cells, in H. S. Nalwa (ed) Handbook of Organic Electronics and Photonics,
American Scientific Publishers, Los Angeles, California, US, Vol 2, Chapter 23
8 A. J. Heeger (2000) Semiconducting and metallic polymers: the fourth generation of polymeric materials, Nobel Lecture
in Chemistry.
9 Y. I. Frenkel (1931) Phys. Rev., no 37, p1276
10 R. Blankenship (2005) Natural organic photovoltaic solar energy transduction, in S. Sun and N. S. Sariciftci (eds) Organic
Photovoltaics: Mechanisms, Materials and Devices, CRC Press, Boca Raton, Florida, US, p37
11 V. Podzorov, E. Menard, A. Borissov, V. Kiryukhin, J. A. Rogers and M. E. Gershenson (2004), Phys. Rev. Lett., 93,
086602.
12 Pochettino (1906) Acad. Lincei Rendiconti, no 15, pp355363
13 R. Bube (1960) Photoconductivity of Solids, Wiley, New York
14 S. Anthoe (2002) Rom. Rep. Phys., no 53, p427
15 C. W. Tang and A. C. Albrecht (1975) J. Chem. Phys., no 62, p2139
ch-03
2/6/07
96
8:21 PM
Page 96
S-S. SUN
H. Hoppe and N. Sariciftci (2005) Bulk heterojunction solar cells, in S. Sun and N. S. Sariciftci (eds) Organic
Photovoltaics: Mechanisms, Materials and Devices, CRC Press, Boca Raton, Florida, US, p217
40 G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery and Y. Yang (2005) Nature Materials, no 4, pp864868
41 M. Drees, R. Davis and R. Heflin (2005) Polymer-fullerene concentration gradient photovoltaic devices by thermally
controlled interdiffusion, in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and Devices,
CRC Press, Boca Raton, Florida, US, (2005) p559
42 S. Sun, Photovoltaic Devices Based on a Novel Block Copolymer, US patent publication no 20040099307 (application
dated 14 November 2002)
43 S. Sun (2003) Sol. Energy Mat. Sol. Cells, no 79, pp257264
44 S. Sun, Z. Fan, Y. Wang and J. Haliburton (2005) J. Mater. Sci., no 40, pp14291443
45 C. Zhang, S. Choi, J. Haliburton, R. Li, T. Cleveland, S. Sun, A. Ledbetter, C. Bonner (2006) Macromolecules, no 39,
pp4317-4326.
46 K. Coakley and M. McGehee (2004) Chem. Mater., no 16, pp45334542
47 M. Jin and L. Dai (2005) Vertically aligned carbon nanotubes for organic photovoltaic devices, in S. Sun and
N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and Devices, CRC Press, Boca Raton, Florida, US,
p579
48 Y. Kang, N. Park and D. Kim (2005) Appl. Phys. Lett., no 86, 113101.
49 L. Schmidt-Mende A. Fechtenktter, K. Mllen, E. Moons, R. H. Friend and J. D. MacKenzie (2001) Science, no 293,
p1119
ch-03
2/6/07
8:21 PM
Page 97
97
50 B. Kippelen, S. Yoo, J. Haddock, B. Domercq, S. Barlow, B. Minch, W. Xia, S. Marder and N. Armstrong (2005) Liquid
crystal approaches to organic photovoltaics, in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms,
Materials and Devices, CRC Press, Boca Raton, Florida, US, p271
51 G. V. Tsarenkov (1975) Sov. Phys. Semicond., vol 9, no 2, pp166171
52 G. Sassi (1989) J. Appl. Phys., no 65, pp322328
53 S. Sun (2005) Sol. Energy Mat. Sol. Cells, no 85, pp261267
54 S. Sun (2005) Mater. Sci. Eng. B., vol 116, no 3, pp251256
55 C. J. Brabec, A. Cravino, D. Meissner, N. S. Sariciftci, T. Fromherz, M. Minse, L. Sanchez and J. C. Hummelen (2001)
Adv. Funct. Mater., no 11, pp374380
56 J. Xue, S. Uchida, B. Rand and S. R. Forrest (2005) Appl. Phys. Lett., no 86, p5757
57 X. Crispin, S. Marciniak, W. Osikowicz, G. Zotti, A. W. D. van der Gon, F. Louwet, M. Fahlman, L. Groenendaal, F. De
Schryver and W. R. Salaneck (2003) Journal of Polymer Science Part B Polymer Physics, no 41, pp25612583
58 A. J. Mkinen, I. G. Hill, R. Shashidhar, N. Nikolov and Z. H. Kafafi (2001) Applied Physics Letters, no 79, pp557559
59 Y. Cao, G. Yu, C. Zhang, R. Menon and A. J. Heeger (1997) Synthetic Metals, no 87, pp171174
60 S. A. Carter, M. Angelopoulos, S. Karg, P.J. Brock and J.C. Scott (1997) Applied Physics Letters, no 70, pp20672069
61 A. Djuristic and C-Y. Kwong (2005) in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and
Devices, CRC Press, Boca Raton, Florida, US, p453
62 W. J. H. van Gennip, J. K. J. van Duren, P. C. Thne, R. A. J. Janssen and J. W. Niemantsverdriet (2002) J. Chem. Phys.,
no 117, pp50315035
63 P. Lane and Z. Kafafi (2005) in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and
Devices, CRC Press, Boca Raton, Florida, US, p49
64 Y. Gao (2005) in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and Devices, CRC Press,
Boca Raton, Florida, US, p421
65 N. Hadjichristidis, S. Pispas and G. Floudas (eds) (2003) Block Copolymers: Synthetic Strategies, Physical Properties, and
Applications, John Wiley & Sons, Inc., New York
66 M Lazzari, and M. Lopez-Quintela (2003) Adv. Mater., no 15, pp15841594
67 F. S. Bates and G. H. Fredrickson (1990) Ann. Rev. Phys. Chem., no 41, pp525557
68 B. de Boer, U. Stalmach, P. F. van Hutten, C. Melzer, V. V. Krasnikov and G. Hadziioannou (2001) Polymer, vol 42, no 21,
pp90979109
69 R. Segalman, C. Brochon and G. Hadziioannou (2005) Solar cells based on diblock copolymers: a PPV donor block and a
fullerene derivatized acceptor block, in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials
and Devices, CRC Press, Boca Raton, Florida, US, p403
70 X. L. Chen and S. A. Jenekhe (1996) Macromolecules, no 29, p6189
71 T. A. Skotheim, R. L. Elsenbaumer and J. R. Reynolds (eds) (1998) Handbook of Conducting Polymers, 2nd ed, Marcel
Dekker, New York
72 Z. Bao, A. Dodabalapur and A. J. Lovinger (1996) Appl. Phys. Lett., no 69, p4108
73 T. Nguyen, J. Wu, V. Doan, B. Schwartz and S. H. Tolbert (2000) Science, no 288, pp652656
74 S. Sun, Z. Fan, Y. Wang, C. Taft, J. Haliburton and S. Maaref (2002) in D. Fichou and K. Kafafi (eds) Organic Photovoltaics
II, SPIE, Bellingham, WA, US, vol 4465, pp121-128
75 S. Sun, Z. Fan, Y. Wang, J. Haliburton, C. Taft, K. Seo and C. Bonner (2003) Syn. Met., no 137, pp883884
76 S. Sun (2004) in P. Lane and Z. Kafafi (eds) Organic Photovoltaics IV, SPIE, Bellingham, WA, US, vol 5215, pp195205
77 S. Sun (2004) in P. Lane and Z. Kafafi (eds) Organic Photovoltaics V, SPIE, Bellingham, WA, US, vol 5520, pp126135
78 S. Thomas, C. Zhang and S. Sun (2005) J. Poly. Sci. (A) Polym. Chem., vol 43, pp42804287
79 S. Sun, Z. Fan, Y. Wang, K. Winston and C. E. Bonner (2005) Mater. Sci. Eng. B., vol 116, no 3, pp279282
ch-03
2/6/07
98
8:21 PM
Page 98
S-S. SUN
80 S. Sun, C. Zhang, J. Haliburton, A. Ledbetter, C. Bonner, M. Drees and N. Sariciftci (2005) in P. Lane and Z. Kafafi (eds)
Organic Photovoltaics VI, SPIE, Bellingham, WA, US, vol 5938, pp59380D1-8
81 M. Knupfer (2003) Appl. Phys. A, no 77, pp623626
82 S. Sensfuss and M. Al-Ibrahim (2005) in S. Sun and N. S. Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials
and Devices, CRC Press, Boca Raton, Florida, US, p529
83 R. A. Marcus (1993) Rev. Mod. Phys., no 65, pp599610
84 V. Balzani (ed) (2000) Electron Transfer in Chemistry, Wiley-VCH, New York
85 E. Peeters, P. Hal, J. Knol, C. Brabec, N. Sariciftci, J. Hummelen and R. Janssen (2000) J. Phy. Chem. B, no 104,
pp1017410190
86 E. Neuteboom, S. J. Meskers, P. van Hal, J. van Duren, E. Meijer, R. Janssen, H. Dupin, G. Pourtois, J. Cornil, R.
Lazzaroni, J. Brdas and D. Beljonne (2003) J. Am. Chem. Soc., no 125, pp86258638
87 J. Cornil, V. Lemaur, M.C. Steel H. Dupin, A. Burquel, D. Beljonne and J. L. Brdas (2005) Electronic structure of organic
photovoltaic materials: modeling of exciton-dissociation and charge-recombination processes, in S. Sun and N. S.
Sariciftci (eds) Organic Photovoltaics: Mechanisms, Materials and Devices, CRC Press, Boca Raton, Florida, US,
pp161182
88 J. Miller, L. Calcaterra and G. Closs (1984) J. Am. Chem. Soc., no 106, pp30473049
89 A. Weller (1982) Photoinduced electron transfer in solution, Phys. Chem., no 133, pp9398
ch-04
2/7/07
11:51 AM
Page 99
Abstract
This chapter provides an overview of methods to characterize the thermal performance
of heat exchangers immersed in thermal storage vessels intended for solar domestic
water heating systems and to assess the mechanical durability and scaling potential of
polymers for this application. Recent research at the University of Minnesota is
summarized and recommendations are made for future research to further the
development of polymer heat exchangers and piping.
Keywords heat exchangers; domestic hot water; integral collector storage; polymers; natural
convection; nylon; heat transfer
4.1 INTRODUCTION
Immersed heat exchangers are common in European combistore water storage tanks for
combined domestic hot water and space heating, but in the US, low temperature solar
thermal systems are used almost exclusively for domestic hot water and immersed heat
exchangers are atypical. This situation is expected to change as lower cost and simplified
solar domestic hot water (SDHW) systems are developed for the US market. The recent
focus of US efforts to reduce system cost is an indirect integral collector storage (ICS)
system (Figure 4.1). An immersed heat exchanger is required to discharge the
unpressurized solar storage. During discharge, the domestic pressurized water flows
through the heat exchanger and is delivered to the conventional water heater. Because
the load-side flow is driven by the domestic water pressure, no mechanical pump is
required. Recent research and development for this system, as well as for systems
intended for cold climates, addresses the barriers and opportunities to shift from copper
and glass components to integrated systems manufactured using mass production
techniques, such as those associated with polymeric materials. Significant cost savings
ch-04
2/7/07
11:51 AM
100
Page 100
ch-04
2/7/07
11:51 AM
Page 101
Yo
101
FIGURE 4.1 Conceptual drawing of an indirect integral solar collector storage (ICS) system with a load-side
immersed heat exchanger for solar domestic water heating
ch-04
2/7/07
102
11:51 AM
Page 102
(a)
(b)
FIGURE 4.2 Conceptual sketch of an immersed heat exchanger in a vertical thermal storage tank; the heat
exchanger is shown here in two positions depending on the intended mode of operation: a) top mounted heat
exchanger intended to discharge the tank, and b) bottom mounted heat exchanger intended to charge the tank
The thermal performance of the heat exchanger is described by the overall heat
transfer coefficient-area product (UA), which is the inverse of the sum of the convective
and conductive thermal resistances:
(1)
In equation (1), Ri is the thermal resistance of the forced convection flow through the heat
exchanger, Rw is the conductive resistance across the heat exchanger wall and Ro is the
natural convection thermal resistance on the outer storage-side of the heat exchanger.
The effectiveness of an immersed heat exchanger is related to the UA by:
(2)
ch-04
2/7/07
11:51 AM
Page 103
103
The subscript s denotes the storage fluid and the subscript f denotes the heat
exchange fluid. The bracketed term on the right hand side of equation (2) is referred to as
the number of transfer units (NTU) of the heat exchanger. For metal heat exchangers, the
conductive resistance across the tube wall (Rw) may be neglected (unless there is
substantial scaling due to use in hard water or if double wall tubes are used). If the tube
is plastic, thin-walled tubes with a relatively high ratio of outer diameter-to-wall thickness,
termed the standard diameter ratio (SDR), are desirable to minimize Rw (Liu et al, 2000;
Davidson et al, 2002). The relative magnitude of the thermal resistances depends on many
factors including the flow rate and flow regime (laminar or turbulent) through the heat
exchanger, the geometry of the combined heat exchanger and storage tank, and the
temperature difference between the outer heat exchange surface and the storage fluid
(the state of charge). The natural convection heat transfer may limit the rate of heat
transfer, especially if the internal flow is turbulent. In this case, it is very important to
accurately characterize Ro:
(3)
h o Ao
(6)
ch-04
2/7/07
104
11:51 AM
Page 104
The characteristic length scale in expressions (5) and (6) is the outer diameter D of an
individual tube in a tube bundle or coil. The empirical constant C and exponent m should
depend only on the geometry of the heat exchanger/tank and the range of Rayleigh
number. The real challenge in developing correlations for this application is measurement
and identification of an appropriate characteristic temperature difference between the
heat exchange surface and the storage fluid, T = Tw Ts. Early studies of smooth and
finned coiled and bayonet heat exchangers (Feiereison et al, 1982; Farrington, 1986;
Farrington and Bingham, 1986 and 1987) did not measure the heat exchanger surface
temperature but rather used various definitions of the log-mean temperature difference to
correlate data. The storage fluid was assumed to be fully mixed. The correlations were
generally unsuccessful in the sense that the constant C was temperature dependent.
Natural convection to a single tube (Liu et al, 2003) and several tube bundle heat
exchangers (Liu et al, 2003 and 2004) immersed in a thin rectangular storage tank with a
height to width aspect ratio of 9.3:1 and inclined at 30 degrees to the horizontal was
measured at the University of Minnesota. The tube wall temperature was measured
directly. When the storage fluid was initially isothermal, it was possible to correlate the
data using the average storage temperature. When the storage fluid was initially stratified,
it was necessary to use the local storage fluid temperature surrounding the heat
exchanger tubes to define Ts and to obtain statistically significant correlations of the
measured data. For a single tube, the data were best correlated by:
NuD = (0.675 0.001)RaD0.25, for 106 RaD 108.
(7)
(8)
(9)
Three important findings emerged from these studies. First, comparison of equations
(7)(9) to the correlations developed for a horizontal tube in an unbounded fluid (Morgan,
1975) proved that the presence of the storage tank enhanced heat transfer. The heat
transfer enhancement is attributed to the buoyancy driven flow in the tank. For a single
tube, the correlation predicts average heat transfer coefficients ~20 per cent larger than
those for a heated tube in an unbounded fluid. Second, fluid motion in the storage fluid
was sufficient to maintain a fully mixed tank for an initially isothermal storage tank. If the
storage fluid was initially stratified, operation of the heat exchanger destroyed
stratification relatively quickly. Third, for tube bundles, the pitch to diameter ratio (P/D) had
no statistically significant impact on the overall heat transfer for 1.5 P/D 3.3. This result
implies that it is possible to pack the heat exchanger tubes in close proximity to each
other without substantially decreasing overall heat transfer. For the same heat exchange
area, multiple tubes reduce the pressure drop across the heat exchanger.
ch-04
2/7/07
11:51 AM
Page 105
105
Use of equations (7)(9) for design of heat exchangers is difficult because it requires
knowledge of the temporal behavior of the storage fluid temperature during the discharge
process. An alternate approach that relies solely on the initial storage temperature was
first suggested by Reindl et al (1992b). From a scale analysis of the problem, a semiempirical expression for the time-dependent Nusselt number based on the initial Rayleigh
number of an initially isothermal tank was proposed. Su and Davidson (2005) extended the
scale analysis to consider discharge of a storage tank with tube bundle heat exchangers
and presented the results in a more generalized dimensionless form.
Su and Davidson (2005) divide the process into four temporal periods referred to as
conduction, quasi-steady, fluctuating and decay. General formulations for the transient
Nusselt number and the volume averaged water temperature in the enclosure were
developed for each period. Three dimensional CFD results illustrate the flow and
temperature fields within the storage tank (Figure 4.3). At the initiation of the discharge, heat
transfer is dominated by conduction. The convective flow is only apparent near the heat
exchanger and the buoyancy-induced velocity is very low. The dimensionless temperature
in the tank during the conduction period may be determined analytically and is given by:
(10)
where n are the positive roots of the Bessel function of the second kind,Y0 (n) = 0, and
r is the dimensionless radial distance measured from the heat exchanger. The volume
averaged storage temperature is defined as:
(11)
where Ts is the average storage temperature at a specified time, and Ts,0 is the fully-mixed
storage temperature at the start of the charge or discharge process (at t = 0). The average
Nusselt number is expressed as:
(12)
The asterisk on the Nusselt number denotes that it is defined in terms of the initial, and
presumed known, difference between the initial surface temperature of the heat exchanger
and the initial temperature of the storage fluid, i.e. T = Tw Ts,0. The conduction period
. The dimensionless time, , is given by:
ends relatively quickly at = Pr
s
(13)
where t0 is the ratio of the convective length scale and a velocity scale for natural
convection. The natural convection velocity scale U is given by:
(14)
Thus the time scale is:
(15)
ch-04
2/7/07
106
11:51 AM
Page 106
FIGURE 4.3 Three dimensional streamlines (top) and isotherms (bottom) for a rectangular storage with a single
immersed tube; T = |Tw To| = 20K (T0 = 353K and Tw = 333K) corresponding to Ra*D = 2.345 107 during
(a) conduction ( = 0.029, t = 0.014s); (b) conduction ( = 1.953, t = 0.955s); (c) quasi-steady ( = 11.778,
t = 5.76s); (d) fluctuating ( = 111.78, t = 54.66s); (e) decay ( = 2441, t = 1194s); and (f) decay ( = 65,710,
t = 32,315s)
Once the conduction period ends, heat transfer becomes dominated increasingly by
natural convection. Initially, the presence of the storage container does not impact the
flow field or heat transfer to the heat exchanger. Consequently, the convective heat
transfer is described by the average Nusselt number for a body immersed in an
unbounded fluid, given in the form of equation (4). This quasi-steady period endures until
ch-04
2/7/07
11:51 AM
Page 107
107
the flow in the enclosure begins to disrupt the boundary layer of the heat exchanger. The
duration depends on the geometry of the tank, the position of the heat exchanger within
the tank and the magnitude of the convective velocity given in equation (14).
At the end of the quasi-steady period, the flow field and heat transfer are influenced
by the presence of the tank. CFD results show there is a short period of intense mixing of
the storage fluid and intermittent disruption and growth of the boundary layer at the heat
exchanger surface. The Nusselt number fluctuates around the quasi-steady value. The
duration of the fluctuating period is not well defined except by observation of numerical
or measured data.
The last period and most important one in terms of overall system performance is the
decay period during which over 90 per cent of the energy is removed (discharge) or added
(charge) to the storage tank. It is during this period that all measured data have been
obtained. The driving temperature difference between the heat exchanger surface and the
storage fluid decreases exponentially. The transient Nusselt number is expressed as:
(16)
where the dimensionless transient temperature difference is given by:
(17)
The geometric parameter is specified by the heat exchanger/storage configuration.
It equals the volume of the storage fluid divided by the product of the heat transfer
surface area of the heat exchanger and the characteristic length scale for natural
convection heat transfer. For N tubes, each of diameter D and outer heat transfer surface
area Ao:
(18)
A generalized expression for the transient Nusselt number is:
(19)
The strength of equation (19) is that it requires only the initial storage temperature and the
heat exchanger geometry and operating conditions, from which Ra*D and can be
determined, to determine heat transfer during the entire decay period. The values of C and
m must be determined empirically and depend only on the geometry. Su and Davidson (2005)
used the data of Liu et al (2003, 2004 and 2005) for single- and multiple-tube heat exchangers
to test the usefulness of equation (19). The data for the single-tube heat exchanger in an
initially isothermal storage were successfully correlated with m = 0.25 and C = 0.50. For the
240-tube heat exchanger with P/D = 3.3, the figures were m = 0.188 and C = 2.6.
ch-04
2/7/07
108
11:51 AM
Page 108
In situations where storage fluid is stratified at the initiation of the charge or discharge
and stratification is maintained by the use of manifolds or baffles, equation (19) is not
applicable. If the fluid is stratified, the average tank temperature does not represent the
thermal environment of the heat exchanger. In this case, use of the initial Rayleigh number
in equation (19) will underestimate the natural convection film coefficient at the beginning
of the charge or discharge process and overestimate it as the storage tank reaches a full
or depleted state of charge, respectively. In this case, correlations of the form of equation
(4) may be used but it will be necessary to characterize the temperature of the storage
fluid surrounding the heat exchanger during use.
Once the UA of the heat exchanger has been estimated, the transient behavior of the
heat exchanger/storage tank can be predicted. The transient storage temperature Ts(t) and
delivered outlet water temperature Tf,out(t) can be estimated from energy balances on the
heat exchanger and storage. Assuming the inlet fluid temperature and mass flow rate
) are fixed and the storage fluid remains fully mixed
through the heat exchanger (m
f
throughout the charge or discharge process, the transient storage temperature and heat
exchanger outlet temperature are given by:
(20)
(21)
The parameter = (cp)f/(cp) is approximately equal to the heat exchanger
effectiveness when the heat exchange and storage fluids are the same fluid. The volume
ratio Vr equals the volume of fluid passed through the heat exchanger divided by the
ft/f /Vs.
storage volume: Vr=m
ch-04
2/7/07
11:51 AM
Page 109
109
into two or more storage compartments is practical and has the advantage of reduced
thermal losses to the ambient.
At the University of Minnesota, we have conducted experiments with a single storage
tank divided into two compartments (Davidson et al, 2005). The dual compartment
storage tank, illustrated in Figure 4.4 for discharge of the store, provides a staged heating
process. The inlet of the heat exchanger is in the first compartment. The heat exchange
fluid entering the second compartment is warmer than that entering the first
compartment and thus the storage fluid in compartment two remains warmer for longer
periods of time. Consequently, for much of the discharge process, the heat exchanger
outlet temperature is higher than it would be using the same heat exchanger in an
undivided storage. The advantage of the approach is that the enhanced heat transfer
attributed to convective motion in the storage (Liu et al, 2003, 2004 and 2005) is
maintained in each compartment. Boies and Homan (2004) point out that the benefit of
this concept can be realized in either of two ways. For a fixed total storage volume, a
higher energy delivery rate is maintained for a longer period, which means more hot water
is produced. Alternately, the same energy output can be provided by a smaller storage
volume.
FIGURE 4.4 Sketch of a divided indirect thermal storage vessel with two storage compartments; the heat
exchanger is positioned for discharge
2/7/07
11:51 AM
110
Page 110
1.18
1.16
NTU = 12
1.14
ch-04
NTU = 6
1.12
1.10
NTU = 4
1.08
Equal outlet
temperatures
1.06
1.04
NTU = 2
1.02
NTU = 1
1.00
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
FIGURE 4.5 Predicted ratio of delivered energy of a divided storage and an undivided storage as a function
of the dimensionless output volume and NTU. The divided storage has two compartments. The dashed line
indicates the volume ratio at which the heat exchanger outlet temperatures of the two storage configurations
are equal
constant during discharge and thermal losses are neglected. Here the ratio of energy
delivered by a divided storage to that delivered by an undivided storage, Er, is plotted
versus Vr for 1 NTU 12. The NTU values shown are those of the undivided storage. For
the divided storage, the NTU is assumed to be divided equally between the two storage
compartments. The energy ratio increases as overall NTU is increased from 1 to 10, with
little change from NTU = 10 to 12. For a given NTU, the energy ratio depends on the
volume ratio, or extent of depletion of the storage. The maximum cumulative delivered
energy is at Vr 0.85. For example, at NTU = 10, Er is 1.15. The dashed line on the plot
indicates the value of Vr for which the outlet temperatures of the undivided and divided
storage are equal. The region to the left of the dashed line, where the divided storage
delivers higher temperature water than the undivided storage, represents the desirable
operating range. The only case for which there is no advantage to the divided storage
system is if the storage is too small to meet the hot water load and the stored energy is
completely discharged during a hot water draw. When the storage unit is completely
discharged, Er = 1.
The measured performance of a divided storage vessel is shown in Figure 4.6. These
data were obtained for a 126 liter rectangular storage vessel with a copper heat exchanger
with a total surface area of 2.38m2 immersed in the storage fluid. Experiments were
conducted with one and two compartments. In the divided storage vessel, the compartments
were separated with a 2.54cm thick polystyrene sheet (R-value = 0.881K-m2/W). Figure 4.6
is a plot of Er versus Vr for nominal NTUs of 2.5 and 7. With NTU = 7, the divided storage
delivers 11 per cent more energy than the undivided storage at Vr = 0.8 (i.e. when 100
liters of hot water or 55 per cent of the stored energy has been delivered). Higher outlet
temperatures are sustained by the divided storage until Vr = 1.2. At this point, the divided
2/7/07
11:51 AM
Page 111
111
1.14
1.12
ch-04
NTU = 7
1.10
1.08
1.06
1.04
NTU = 2.5
1.02
1.00
0.0
0.5
1.0
1.5
2.0
2.5
3.0
storage provides 8 per cent more energy than the undivided storage and the storage is 74
per cent depleted. With NTU = 2.5, the divided storage delivers 5 per cent more energy
at Vr = 0.8. At Vr = 1.3, when the outlet temperatures of the divided and undivided storage
are equal, the divided storage provides 4.6 per cent more energy and the stored energy
is 65 per cent depleted. The measured advantage of the divided storage is slightly lower
than that predicted because the NTU of the heat exchanger decreases as the tank is
discharged. However, even with a simple partition, the divided storage vessel provides the
same benefit as that predicted for multiple storage tanks. The concept is expected to be
easy to implement in solar systems at minimal or no extra cost.
4.2.3 SUMMARY
Substantial progress has been made in recent years to understand and characterize the
thermal/fluid processes in storage vessels with immersed heat exchangers. Natural
convection heat transfer correlations have been developed for tube bundles and the
generalized approach to express measured data in the form of equation (19) holds
promise for fully mixed storage vessels. The empirical constant in this expression will
depend on the combined heat exchanger/tank design and must be determined from
experiment or CFD. The computational problem requires substantial computational
resources. It is anticipated that most developers and designers will choose the
experimental approach. Once the constant is determined, the transient charge or
discharge process can be predicted using only the initial conditions. If baffles or other
stratification devices are shown to be effective, equation (4) may be used. In this case, the
key to predicting the natural convection heat transfer is to determine the thermal
environment to which the heat exchanger is exposed.
ch-04
2/7/07
112
11:51 AM
Page 112
ch-04
2/7/07
11:51 AM
Page 113
113
is likely. The PA66 polymer chain includes an amide group which can by hydrolyzed in
water, leading to chain scission and permanent damage.
The effect of hot water on the molecular structure of polyolefins was analyzed using
infrared spectroscopy, size exclusion chromatography, thermal analysis and visual
observations (Gedde et al, 1994). Cross-linked polyethylene (PEX), polybutene-1,
polypropylene and medium-density polyethylene were studied. These researchers found
that there is an incubation period during which there is a decrease in the antioxidant
concentration in a polyolefin. The incubation period ends when the antioxidants are
completely depleted, leaving the material with an extremely limited lifetime. Oxidation
was found to be the main mechanism responsible for polymer degradation. The study,
although limited to polyolefins, established a method for predicting the lifetime of a
polymer based on antioxidant depletion.
Prior work addressing the effect of a hot potable water environment on mechanical
properties of polymer tubing has focused on polymer strength. The National Sanitary
Foundation (NSF) and the American Society for Testing and Materials (ASTM) have both
developed testing protocols, NSF standard P171 (1999) and ASTM standard F2023 (2000),
to evaluate the burst strength of polymer tubing in hot chlorinated water. In these tests, the
tube rupture must be a result of the oxidative environment as indicated by a failure in which
there are many cracks emanating from the interior tube wall and extending to the tube
surface (stage III failure). The water chlorine level ranges between 3 to 5ppm and pH level
ranges from 6 to 8 such that an ORP of 825mV is maintained throughout the test. (Note that
the ORP of typical tap water ranges from 400 to 550mV (ASTM, 2000).) Tests are performed
over several combinations of pressure and temperature so as to extrapolate the burst
strength of the tubing after 10 to 40 years exposure to hot chlorinated water. Test results
are limited to the particular tube geometries (diameter to thickness ratio) tested. These
standards provide a methodology to predict polymer lifetime based on a material strength
limit, the hydrostatic burst strength. Material deformation (in other words creep compliance)
is not evaluated during testing and is not considered in predicting material lifetime.
Bradley et al (2000) used a similar test to determine the effects of temperature and
chlorine levels on the tensile strength and weight of polymer tubes. Polymers were tested
in conditions of 28C, 60C and 90C with 5ppm chlorine and also at 90C with 0ppm
chlorine. The strength of the materials when immersed in a hot chlorinated environment
for 1500 hours was evaluated under (1) no load and (2) continuous loading conditions. The
harshest condition, 90C and 5ppm chlorine, was very detrimental to PA66, causing it to
lose 70 per cent of its weight and 95 per cent of its tensile strength. Reducing the
temperature to 60C from 90C caused the degradation of PA66 to slow dramatically. PSU,
polyphenylsulfone, polyphenylsulfide and polyvinylidene fluoride exhibited good
resistance to chlorine exposure at high temperature with little weight or tensile strength
loss. The creep compliance was not evaluated. Limited creep data in oxidative
environments for PB and PA66 have been published by researchers at the University of
Minnesota (Walter et al, 2003; Wu et al, 2004). Because the test data were obtained for
only one environmental condition (825mV ORP and 82C), the effects of changes in ORP
and temperature could not be evaluated. No tests of chemical or physical degradation
(such as molecular weight or SEM) were performed.
ch-04
2/7/07
114
11:51 AM
Page 114
2/7/07
11:51 AM
Page 115
115
ch-04
10
0.1
0.1
10
100
1000
Time (hours)
FIGURE 4.7 Creep compliance of PA66 in air and exposed to hot chlorinated water
2/7/07
11:51 AM
116
Page 116
100
No exposure
550 20 mV, 60 1C
830 20 mV, 60 1C
830 20 mV, 80 8C
80
uncertainty, stress
60
40
20
uncertainty, strain
250
Strain at failure (%)
ch-04
200
150
100
50
0
0
50
100
150
200
250
300
Time (hours)
FIGURE 4.8 Tensile strength (top) and strain (bottom) of PA66 before and after exposure to hot
chlorinated water
80C conditions, strain at failure varies depending on the exposure time (Figure 4.8).
Considerable elongation occurs for samples aged from 2 to 5 hours (strain at failure during
this period was as high as 200 per cent). After 12 hours exposure the strain at failure
ranges from 35 to 45 per cent. This variation in elongation over time may be associated
with initial water absorption followed by hydrolysis. Although strain at failure is similar for
unexposed and exposed specimens (exposure time greater than 10 hours), the exposed
material has consistently lower strength and reduced stiffness.
ch-04
2/7/07
11:51 AM
Page 117
117
(a)
(b)
FIGURE 4.9 SEM images of PA66: (a) native unexposed surface and (b) after 292 hours exposure at 830
40mV, 59 2C. The surface of the unexposed PA66 is smooth with no visible defects. Surface cracks, ranging
from 1m to 10m in length, are visible on the surface of all specimens exposed to water
ch-04
2/7/07
118
11:51 AM
Page 118
4.3.3 SUMMARY
These results highlight the importance of designing heat exchanger components based
on data from specimens that have been exposed to an aqueous, oxidative environment.
The challenge is to elicit degradation within the test duration. It is difficult to accelerate
the test sufficiently using a traditional ORP control approach.
The data presented here show that for up to 1100 hours exposure in hot potable
chlorinated water, PSU and PB (with antioxidant additives) maintain their mechanical
properties, while PA66 degrades significantly. These test results are consistent with
the expected degradation mechanisms for each polymer. PSU has a rigid backbone
with strong chemical bonds and was expected to be resistant to oxidation. PSU tubing
is commercially available in limited sizes. Manufacturing process development is
required to cost-effectively produce small diameter, thin-walled PSU tubes required for
heat exchanger applications. The PB included an antioxidant package, such that
degradation of the polymer is delayed until the antioxidant is depleted. The combination
of test environments and durations studied were not sufficient to deplete the
antioxidant. Further investigation of the relationship between depletion of the
antioxidant additives in polyolefins is recommended to determine the expected life of
these materials. Polyamide 6,6 is not recommended for this application as test results
indicate substantial degradation of the material in relatively short times. The polymer
readily absorbs water and the polymer chain includes an amide group which can be
hydrolyzed.
ch-04
2/7/07
11:51 AM
Page 119
119
with water that is supersaturated with respect to calcium carbonate. The supersaturation (S)
is commonly given by:
(22)
where aCa2+ and aCO32 are the activities of calcium and carbonate ions, respectively and Ksp
is the solubility product for crystalline CaCO3 (e.g. calcite) (Snoeyink and Jenkins, 1980;
Stumm and Morgan, 1996). When S is greater than 1, the solution is supersaturated and
the formation of solid calcium carbonate is possible. A higher value of S indicates a
greater thermodynamic driving force for scaling. To calculate S, several chemical reactions
that control species concentrations must be taken into account along with the
temperature, which influences the reaction rates as well as the solubility product. Next,
solid calcium carbonate forms on the polymer surface by nucleation and growth
(Nancollas, 1979; Mullin, 1993). During the initial nucleation stage, the chemical and
physical properties of the polymer surface, as well as the supersaturation and the
temperature, are expected to play a role. During growth, ions from the water are
transported to the surface and the calcium carbonate scale layer grows. Growth can be
controlled by transport to the surface or reaction at the surface. When transport controls,
the flow rate (e.g., Reynolds number) enters in as another important variable in scaling, as
flow impacts the mass transfer to the surface. Lastly, the scale can be removed,
depending on the flow conditions and the adhesion of the scale to the tube. Hence at
least four variables impact the scaling process: the composition of the water, the
temperature, the surface composition and structure, and the flow condition.
ch-04
2/7/07
120
11:51 AM
Page 120
TABLE 4.1 Scale accumulation and scaling rate for tubes in a tube-in-shell heat exchanger1 as
determined by chemical analysis
MATERIAL
EXTERNAL
DIAMETER
(mm)
3.25
WALL
THICKNESS
(mm)
0.41
VOLUME
OF WATER2 (L)
1158
SCALE
ACCUMULATED
CaCO3 (g /m2)
1.275
5.89
2.54
2.18
1.52
4.57
6.6 x 107
2.19
2535
0.082
8.3 x 109
0.76
1.07
1.48
1.78
478
303
327
143
0.172
0.202
0.173
0.395
8.9 x 108
1.0 x 107
8.9 x 108
2.0 x 107
Scale accumulation and rate calculated assuming that all calcium was present as CaCO3.
Volume of water passed through each tube in 540 hours during accelerated scaling phase.
SCALING RATE
(g /m2s)
ch-04
2/7/07
11:51 AM
Page 121
121
FIGURE 4.10 SEM micrographs of scale formed on (a) polyamide 6,6, (b) polypropylene, (c) polybutylene
and (d) copper tubes in a tube-in-shell heat exchanger. SEM micrographs were taken at the end of 540 hours of
accelerated scale testing. A layer of new material of a distinctly different morphology from the native polymer
is apparent
(Aizenberg et al, 1999). The lesser scaling rate for the HTN was attributed to the tubes
larger wall thickness, which prevented the interior surface temperature from reaching as
high a temperature as the other materials. For this water chemistry, the temperature had
little effect on the supersaturation, but the lower temperature in the case of HTN resulted
in slower kinetics of deposition. These conclusions were consistent with a surfacereaction limited model for scaling that took into account differences in temperatures due
to the differing wall thicknesses and thermal conductivities of the materials. From these
experimental and modeling results, several of the candidate polymers for solar thermal
applications (polypropylene (PP), PB, HTN and Teflon) had very similar scaling tendencies.
Scaling studies have also been carried out using flow-through systems under
isothermal conditions. Andritsos and co-workers (Andritsos et al, 1996 and 1997) studied
scale formation at room temperature under turbulent flow conditions. Their experimental
system used a syringe pump to boost the pH, and hence the supersaturation, in-line. This
design allowed them to achieve high supersaturation without the risk of precipitation of
calcium carbonate particles in the bulk liquid. Their results showed a marked increase in
scale deposition rate with temperature and flow velocity, consistent with work by Hasson
ch-04
2/7/07
122
11:51 AM
Page 122
et al (1968). Interestingly, the scaling rate was found to undergo a dramatic increase at a
critical value of supersaturation (S = 8). The authors speculate that below the critical
supersaturation the formation of scale is by a surface-controlled mechanism and above
the critical value a diffusion-controlled mechanism dominates. Likewise, the scale
accumulation with time depended on the supersaturation; at low S, the scale
accumulated slowly in the early stage, with an apparent induction period, but at high S,
the scale quantity increased linearly with time. The researchers compared the
morphologies of the scale deposits on steel, copper and Teflon and found that the tube
material had little effect. The tenacity of the scale was less on Teflon, however.
Using a similar isothermal flow-through system, Sanft et al (2006) carried out
accelerated scaling experiments at room temperature on tubes of cross-linked
polyethylene (PEX), a polypropylene copolymer (PP-r) and copper. Two tests were run
under identical conditions using distilled water with its composition adjusted (S = 7.8,
[Ca2+]t = 3 x 103 M, pH = 9). The accumulation of scale was slow at early times and then
increased, similar to the results of Andritsos et al (1996). There was large variability in the
data, however, which was attributed to the randomness of the nucleation process. Within
this limitation, scaling tendencies of the materials in the study were similar. SEM
micrographs revealed isolated calcite and vaterite particles, as shown in Figure 4.11.
Coverage is incomplete, indicative of the early nucleation-dominated stage of scaling.
Another approach is to expose polymers to supersaturated water without flow. In one
such study, Roques and Girou (1974) noted a connection between the type of material used
for the walls of their reactor, which contained supersaturated water, and the formation of
calcium carbonate on the walls. The induction time for this heterogeneous nucleation on
the walls was decreased in the following order: poly(vinyl chloride), poly(methyl
methacylate), glass and stainless steel. In another study, nucleation times were compared
FIGURE 4.11 SEM micrographs of scale formed on PEX tubes after (a) 5 hours and (b) 7.5 hours of
exposure to flowing hard water at room temperature. After exposure, isolated calcite and vaterite particles are
visible. Coverage is incomplete, indicative of the early nucleation-dominated stage of scaling
ch-04
2/7/07
11:51 AM
Page 123
123
for various substrate materials that were immersed in hard water kept at constant
temperature (Ben Amor et al, 2004). The nucleation time was determined by monitoring the
pH and the calcium ion content of the water. The researchers observed that under low
hardness conditions the nucleation times were much different, decreasing from polyamide
to polyvinyl chloride to chrome to steel, consistent with Roques and Girou. However, when
harder water was used, the nucleation times were all shorter and roughly the same. This
result concurs with evidence from flow-through experiments that the nature of the material
has the greatest impact when scaling conditions are not severe and nucleation dominates.
A series of reports document the formation of calcium carbonate on the surfaces of
fine polymer particles circulating in supersaturated water under isothermal conditions
(Dalas, et al, 1988, 1999 and 2000; Kanakis and Dalas, 2000; Dousi et al, 2003). In these
studies, polymer particles are placed in a constant concentration reactor in which a drop
in pH associated with calcium carbonate formation triggers the addition of reactant
solutions, allowing for quantitative analysis of deposition. The results of studies of
calcium carbonate formation on sulfonated polystyrene and polystyrene divinylbenzene
(Dalas et al, 1988), carboxylated poly(vinyl chloride) copolymer (Dalas et al, 1999) and
cellulose (Dalas et al, 2000) polymers parallel earlier work on calcium carbonate formation
on calcite seeds (Nancollas and Reddy, 1971; Reddy and Nancollas, 1971; Kazmierczak
and Tomson, 1982), supporting the finding that the induction time decreases and the
calcium carbonate formation (nucleation and growth) rate increases as the
supersaturation of the water is increased. These studies generally point to a surfacecontrolled or surface diffusion-controlled mechanism and reveal some of the effects of
polymer chemistry. An important chemical factor is the ability of the polymer to bind
calcium ions. For example, calcium ions can bind to the carboxylate (C=O) functional
group (Dousi et al, 2003). With this binding ability, the local supersaturation at the surface
is high and so the induction time drops. This research on polymer particles gives good
guidance on the selection of candidate polymers for solar water heating systems:
polymers lacking active functional groups, such as PEX, PP-r, and PP, have the greatest
promise in terms of minimizing scaling.
In recent work at the University of Minnesota, Wang (2004) and McGill (2005) studied
the formation of calcium carbonate on polymers in isothermal cells containing hard water.
They immersed thin sheets of the candidate polymers, including those studied by Wang et
al (2005), as well as polysulfone and polyphenylsulfone, in hard water. Calcium carbonate
deposition was quantified by chemical analysis of the deposit. Conditions were chosen to
study nucleation, and hence the randomness of the process led to some variability in the
data. These researchers found the scaling tendencies of the candidate polymers to be the
same within the experimental error. Experiments were conducted at room temperature and
approximately 50C using water that was supersaturated to approximately the same degree
at the two temperatures. A significantly greater amount of scale formed during the elevated
temperature test, showing the importance of kinetics in enhancing the rate of deposition.
4.4.3 SUMMARY
While progress has been made in recent years, more research is needed to understand
and control the scaling of polymers. Results to date indicate all surfaces are prone to
ch-04
2/7/07
124
11:51 AM
Page 124
scale, but the scaling rate may be different if the conditions ensure that nucleation or
surface reaction control the deposition. Scale formation on polymers is linked in part to
the chemical structure of the material. For example, PA66 appears to form scale more
readily due to its susceptibility to water absorption and hydrolysis (Wang et al, 2005). Yet
unknown is the importance of the physical structure, including surface roughness, which
has been shown to influence the scaling of metals (Keysar et al, 1994), and stress
concentration, which is documented as trouble causing in calcification of biomedical
polymers in the body (Banas and Baier, 2000). In general, scaling is enhanced for water
with higher supersaturation. Scaling rate also increases with temperature, an effect due,
to some extent, to an increase in supersaturation and perhaps to a greater extent to the
acceleration of the kinetics of deposition process. The flow conditions become critical
when the scale grows by ion transport to the surface scaling rate increases with flow
rate. This phenomenon is well documented for turbulent flow (Hasson et al, 1968), but
less so for laminar flow, which is characteristic of flow through narrow polymer tubes
such as might be used in heat exchangers. The shear removal of scale from polymer
tubes has not been investigated extensively. Here, the difference in scale adherence,
which will likely be related to the chemical and physical structure of the polymer, may be
important. Lastly, the field of scaling research on polymers as well as metals is in need of
innovation in characterization tools. Some progress has been made on this front with
several reports of new in situ methods: microscopic imaging (Kim et al, 2002), rotating
disc electrode monitoring (for metals) (Chen et al, 2005), attenuated total reflectance
infrared sensing (Smith et al, 2004), quartz crystal microbalance monitoring (Kohler et al,
2001), optical fiber sensing (Lyons et al, 2001), and monitoring of laser reflectance and
scattering (Euvrard et al, 2004; Sanft, 2005).
4.5 CONCLUSION
Thermal characterization of heat exchangers immersed in solar water storage vessels for
domestic hot water systems has been the subject of study for nearly 20 years. Yet
predicting thermal performance remains a challenge because the transient natural
convection heat transfer coefficients for such heat exchangers depend on the geometry
of the heat exchanger/tank combination, the state of charge and the extent of thermal
stratification. Based on recent work summarized in this paper, we recommend equation
(19) to predict the transient natural convection Nusselt number during either charge or
discharge. The advantage of this transient formulation is that only the initial Rayleigh
number and an empirical constant C are required. The empirical constant must be
determined for each heat exchanger/tank combination. The correlation is valid for fully
mixed storage tanks. In most cases, charge or discharge of the tank with an immersed
heat exchanger maintains a fully mixed tank and is likely to destroy existing stratification.
In situations where stratification of the storage fluid is maintained by the use of manifolds
or other devices, equation (19) is not applicable. In this case, correlations of the form of
equation (4) may be used and it will be necessary to characterize the temperature of the
storage fluid surrounding the heat exchanger during use. Either a validated transient zonal
model of the heat exchanger/tank such as that used in TRNSYS or computational fluid
dynamics models will be needed to predict the temperature distribution in the tank.
ch-04
2/7/07
11:51 AM
Page 125
125
As manufacturers and system integrators consider the use of polymer materials for
immersed heat exchangers, durability and reliability of the heat exchanger must be
considered. A critical aspect is the long-term mechanical stability of materials in water at
elevated temperature. In addition, chlorinated water may create an oxidative environment in
which certain polymers are chemically attacked, resulting in permanent loss of mechanical
strength and increase in creep compliance. Water absorption and hydrolysis can also
diminish polymer mechanical properties. The mechanism of polymer degradation will depend
on the polymer molecular structure. Based on testing of polysulfone, (PSU), polybutylene (PB)
and polyamide 6,6 (PA66), we do not recommend the use of polyamide 6,6 for this
application. Data for polybutylene is inconclusive. Although the material did not degrade after
1100 hours of exposure to hot, chlorinated water, PB is expected to fail once the antioxidant
additives have been depleted. Further investigation of the relationship between depletion of
the antioxidant additives in polyolefins is recommended to determine the expected life of
these materials. Polysulfone is a promising material as it is stable in the intended environment
and does not require the use of antioxidants.
All heat exchangers operated in an open loop with water supersaturated with respect
to calcium carbonate or other inverse soluble salts are prone to scale. Measurement of
scaling in a number of candidate polymer tubes indicates the scaling rate for polymers is
of the same order of magnitude as that for copper. Additional research is needed to
assess scale adhesion on polymers and to develop polymers that are resistant to scale.
ACKNOWLEDGEMENTS
The authors appreciate the collaboration of Dr. Jay Burch at the US National Renewable Energy
Laboratory, Professor F. A. Kulacki and graduate students at the University of Minnesota, and
Professor Kelly Homan at the University of Missouri-Rolla. We gratefully acknowledge the
financial support of the National Renewable Energy Laboratory, the US Department of Energy
and the University of Minnesota Initiative for Renewable Energy and the Environment.
REFERENCES
Aharoni, S. M. (1997) n-Nylons: Their Synthesis, Structure and Properties, John Wiley & Sons, New York
Aizenberg, J., Black, A. J. and Whitesides, G. M. (1999) Oriented growth of calcite controlled by self-assembled monolayers
on functionalized alkanethiols supported on gold and silver, J. Am. Chem. Soc., vol 121, pp45004509
Andritsos, N., Kontopoulou, M., Karabelas, A. J. and Koutsoukos, P. G. (1996) Calcium carbonate deposit formation under
isothermal conditions, Can. J. Chem. Eng., vol 74, p911
ch-04
2/7/07
11:51 AM
126
Page 126
Andritsos, N., Karabelas, A. J. and Koutsoukos, P. G. (1997) Morphology and structure of CaCO3 scale layers formed under
isothermal flow conditions, Langmuir, vol 13, pp28732879
ASTM (2000) F 2023-00: Standard test method for evaluating oxidative resistance of crosslinked polyethylene (PEX) tubing
and systems to hot chlorinated water, in Annual Book of ASTM Standards, American Society of Testing and Materials,
West Conshohocken, PA, US
ASTM (2001) D 1238: Standard test method for melt flow rates of thermoplastics by extrusion plastometer, in Annual Book
of ASTM Standards, American Society of Testing and Materials, West Conshohocken, PA, US
ASTM (2004) D789-04: Standard test method for determination of relative viscosity of polyamide (PA), in Annual Book of
ASTM Standards, American Society of Testing and Materials, West Conshohocken, PA, US
Baker, D. K. (2000) An investigation of calcium carbonate scaling rates based on experiments and modeling, Ph.D. Thesis in
Mechanical Engineering, University of Texas, Austin, US
Baker, D. K. and Vliet, G. C. (2001) Designing solar hot water systems for scaling environments, Journal of Solar Energy
Engineering, Transactions of the ASME, vol 123, no 1, pp4347
Baker, D. and Vliet, G. (2003) Identifying and reducing scaling problems in solar hot water systems, Journal of Solar Energy
Engineering, Transactions of the ASME, vol 125, pp6166
Banas, M. D. and Baier, R. E. (2000) Accelerated mineralization of prosthetic heart valves, Mol. Cryst. and Liq. Crystl., vol
354, pp249267
Basell Polyolefins (2004) Product Data Sheet: Polybutylene-1, PB 4137 Grey
Bejan, A. (1982) Entropy Generation Through Heat and Fluid Flow, John Wiley & Sons, New York, p165
Ben Amor, M., Zgolli, D., Tlili, M. M. and Manzola, A. S. (2004) Influence of water hardness, substrate nature and
temperature on heterogeneous calcium carbonate nucleation, Desalination, vol 166, pp7984
Bradley, S. W., El-Hibri, J., Bersted, B. H. and Bradley, W. L. (2000) A study of the effect of chlorinated water on engineering
thermoplastics at elevated temperatures, presented at ANTEC, Orlando, US
Branch, C. A. and Muller-Steinhagen, H. M. (1991) Influence of scaling on the performance of shell-and-tube heat
exchangers, Heat Transfer Engineering, vol 12, pp3745
Boies, A. M. and Homan, K. O. (2004) Improving discharge characteristics of indirect integral collector storage systems with
multi-element storage, ASME International Solar Energy Conference, Portland, OR, US, June 2004, CDROM
Boies, A. M. and Homan, K. O. (2005) System performance of an indirect collector storage preheater in series with a
conventional electric resistance water heater, 2005 ISES Solar World Congress, Orlando, FL, US, August 2005, ASME,
paper ISEC2005-76200, CDROM
Boies, A. M., Homan, K. O., Davidson, J. H. and Liu, W. (2005) A variable effectiveness model for indirect thermal storage
devices, 2005 ASME Summer Heat Transfer Conference, San Francisco, CA, US, July 2005, ASME, paper HT2005-72711,
CDROM
Bott, T. R. (1988) Crystallisation fouling Basic science and models, Fouling Science and Technology, vol 145, pp251260
Budair, M. O., Khan, M. S., Zubair, S. M., Sheikh, A. K. and Quddus, A. (1998) CaCO3 scaling in AISI 316 stainless steel
tubes Effect of thermal and hydraulic parameters on the induction time and growth rate, Heat & Mass Transfer, vol 34,
pp163170
Chen, T., Neville, A., and Yuan, M. (2005) Calcium carbonate scale formation Assessing the initial stages of precipitation
and deposition, J. Petroleum Science and Engineering, vol 46, pp185194
Cowan, J. C. and Weintritt, D. J. (1976) Water-Formed Scale Deposits, Gulf Publishing Co., Houston, TX, US
Dalas, E., Kallitsis, J. and Koutsoukos, P. G. (1988) The crystallization of calcium carbonate on polymeric substrates,
J. Crystal Growth, vol 89, pp287294
Dalas, E., Klepetsanis, P. and Koutsoukos, P. G. (1999) The overgrowth of calcium carbonate on poly(vinyl chloride)-co-vinyl
acetate-co-maleic acid, Langmuir, vol 15, pp8322327
ch-04
2/7/07
11:51 AM
Page 127
127
Dalas, E., Klepetsanis, P. G. and Koutsoukos, P. G. (2000) Calcium carbonate deposition on cellulose, J. Colloid. Inter. Sci.,
vol 224, pp5662
Davidson, J. H., Mantell, S. C. and Jorgensen, G. (2002) Status of the development of polymeric solar water heating
systems, in D.Y. Goswami (ed) Advances in Solar Energy, American Solar Energy Society, vol 15, pp149186
Davidson, J. H., Ragoonanan, V. and Homan, K. (2005) Divided storage in an indirect integral collector storage with immersed
heat exchanger, Proceedings of the 2005 Solar World Congress, Orlando, FL, US
Dousi, E., Kallitsis, J., Chrissanthopoulos, A., Mangood, A. H. and Dalas, E. (2003) Calcite overgrowth on carboxylated
polymers, Journal of Crystal Growth, vol 253, pp496503
Drck, H. (2002) Influence of different combistore concepts on the overall system performance, IEA-SHC Task 26, Industry
Workshop, Oslo, Norway, pp3946
DuPont (2004) Product Data Sheet: Zytel 103 HSL NC010
Epstein, N. (1986), Fouling of heat exchangers, in J. W. Palen (ed.) Heat Exchanger Source Book, Hemisphere Publishing
Corporation, New York, pp677697
Euvrard, M., Membrey, F., Filiatre, C. and Foissy, A. (2004) Crystallization of calcium carbonate at a solid/liquid interface
examined by reflectance of a laser beam, Journal of Crystal Growth, vol 265, pp322330
Farrington, R. B. and Bingham, C. E. (1986) Testing and analysis of immersed heat exchangers, SERI/TR-253-2866, National
Renewable Energy Laboratory, Golden, Colorado, US
Farrington, R. B. and Bingham, C. E. (1987) Testing and analysis of load-side immersed heat exchangers for solar domestic
hot water systems, SERI/TR-254-3094, National Renewable Energy Laboratory, Golden, Colorado, US
Feiereison, T. J., Klein, S. A., Duffie, J. A. and Beckman, W. A. (1982) Heat transfer from immersed coils, Paper No 82
WA/SOL-18, American Society of Mechanical Engineers, New York
Ferguson, R. J. (1984) A kinetic model for calcium carbonate deposition, Materials Performance, vol 23, pp2534
Freeman, A. (2004) Mechanical performance of polysulfone, polybutylene, and nylon 6,6 in hot, chlorinated water, MS
Thesis in Mechanical Engineering, University of Minnesota, Minneapolis, US
Freeman, A., Mantell, S. C. and Davidson, J. H. (2005) Mechanical performance of polysulfone, polybutylene, and polyamide
6,6 in hot chlorinated water, Solar Energy, vol 79, pp624637.
Gedde, U. W., Viebke, J., Leijstrom, H. and Ifwarson, M. (1994) Long-term properties of hot-water polyolefin pipes A
review, Polymer Engineering and Science, vol 34, no 24, pp17731787
Githens, R. E., Minor, W. R. and Tomsic, V. J. (1965) Flexible tube heat exchangers, Chem. Eng. Prog., vol 61, no 7, pp5562
Hasson, D. (1981); Precipitation fouling: A review, in E. F. C. Somerscals and J. G. Knudsen (eds) Fouling of Heat Transfer
Equipment, Hemisphere Publishing Corporation, New York, pp527568
Hasson, D., Avriel, M., Resnick, W., Rozenman T. and Windreich, S. (1968) Mechanism of calcium carbonate scale deposition
on heat-transfer surfaces, I&EC Fundamentals, vol 7, no 1, pp5965
Kanakis, J. and Dalas, E. (2000) The crystallization of vaterite on fiber, Journal of Crystal Growth, vol 219, pp277282
Karabelas, A. J. (2002) Scale formation in tubular heat exchangers Research priorities, International Journal of Thermal
Science, vol 41, pp682692
Kazmierczak, T. F. and Tomson, M. B. (1982) Crystal growth of calcium carbonate: A controlled composition kinetic study,
J. Phys. Chem., vol 86, pp103107
Keysar, S., Semiat, R., Hasson, D. M. and Yahalom, J. (1994) Effect of surface roughness on the morphology of calcite
crystallizing on mild steel, J. Colloid and Interface Sci., vol 162, p311
Khan, M. S., Zubair, S., Budair, M., Sheikh A. and Quddus, A. (1996) Fouling resistance model for prediction of CaCO3 scaling
in AISI 316 tubes, Heat and Mass Transfer, vol 32, pp12
Kim, W. T., Bai, C. and Cho, Y. I. (2002) A study of CaCO3 fouling with a microscopic imaging technique, International Journal
of Heat and Mass Transfer, vol 45, pp597607
ch-04
2/7/07
11:51 AM
128
Page 128
Knudsen, J. G. (1981) Cooling water fouling A brief review, American Society of Mechanical Engineers, Heat Transfer
Division, vol 17, pp2938
Kohler, N., Courbin, G. and Ropital, F. (2001) Static and dynamic evaluation of calcium carbonate scale formation and
inhibition, European Formation Damage Control Conference, Expanding Horizons, Society of Petroleum Engineers, The
Hague, The Netherlands
Liu, W., Davidson, J. H., and Mantell, S. C. (2000) Analysis of polymer heat exchangers for solar water heating: A case
study, Journal of Solar Energy Engineering, Transactions of the ASME, vol 122, no 2, pp8491
Liu, W., Davidson, J. H., Mantell, S. C. and Kulacki, F. A. (2003) Natural convection from a horizontal tube heat exchanger
immersed in a tilted enclosure, Journal of Solar Energy Engineering, Transactions of the ASME, vol 125, no 1, pp6775
Liu, W., Davidson, J. H. and Kulacki, F. A. (2004) Natural convection from a tube bundle in a thin inclined enclosure, Journal
of Solar Energy Engineering, Transactions of the ASME, vol 126, no 2, pp702709
Liu, W., Davidson, J. H., and Kulacki, F. A. (2005) Thermal characterization of prototypical integral collector storage systems
with immersed heat exchangers, Journal of Solar Energy Engineering, Transactions of the ASME, vol 127, no 1, pp2128
Lyons, W. B., Ewald, H., Flanagan, C., Lochmann, S. and Lewis, E. (2001) A neural networks based approach for determining
fouling of multipoint optical fibre sensors in water systems, Meas. Sci. Technol., vol 12, pp958965
Mather, D., Hollands, K. G. T. and Wright, J. L. (2002) Single- and multi-tank energy storage for solar heating systems:
fundamentals, Solar Energy, vol 73, no 1, pp313
Marner, W. J. and Suitor, J. W. (1987) Fouling with Convective Heat Transfer, John Wiley & Sons, New York
McGill, K. N. (2005) Calcium carbonate scaling of polymers in isothermal, stagnant water, Mechanical Engineering,
University of Minnesota, Minneapolis, MN, US
Morgan, V. T. (1975) The overall convective heat transfer from smooth circular cylinders, Advances in Heat Transfer, vol 11,
pp199264
Mullin, J. W. (1993) Crystallization, Butterworth-Heinemann, Oxford, UK
Nancollas, G. H. (1979) The growth of crystals in solution, Advances in Colloid and Interface Science, vol 10, pp215252
Nancollas, G. H. and Reddy, M. M. (1971) The crystallization of calcium carbonate: II Calcite growth mechanism, J. Colloid
and Interface Sci., vol 36, pp824
NSF (1999) NSF Protocol P171: Chlorine resistance of plastic piping materials, National Sanitary Foundation International,
Ann Arbor, MI, US
Reddy, M. M. and Nancollas, G. H. (1971) The crystallization of calcium carbonate: I. Isotopic exchanger and kinetics,
J. Colloid and Interface Science, vol 36, p166
Reindl, D. T. (1992) Source driven transient natural convection in enclosures, Ph.D. Thesis in Mechanical Engineering,
University of Wisconsin, Madison, US
Reindl, D. T., Beckman, W. A. and Mitchell, J. W. (1992a) Transient natural convection from a vertical flat plate in a rectangular
enclosure, Proceedings, ASME 28th National Heat Transfer Conference and Exhibition, San Diego, CA, US, pp9198
Reindl, D. T., Beckman, W. A. and Mitchell, J. W. (1992b) Transient natural convection in enclosures with application to solar
thermal storage tanks, Journal of Solar Energy Engineering, Transactions of the ASME, vol 114, pp175181
Roques, H. and Girou, A. (1974) Kinetics of the formation conditions of carbonate tartars, Water Research, vol 8, p907
Sanft, P. M. (2005) Development of an apparatus for in situ analysis of calcium carbonate formation on polymer surfaces,
MS Thesis in Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, US
Sanft, P., Francis, L. F. and Davidson, J. H. (2006) Calcium carbonate formation on cross-linked polyethylene (PEX) and
polypropylene random copolymer (PP-r), Journal of Solar Energy Engineering, Transactions of the ASME, vol 128, no 2,
pp251254
Scott, G. (1999) Polymers and the Environment, The Royal Society of Chemistry, Cambridge, UK
ch-04
2/7/07
11:51 AM
Page 129
129
Sekulic, D. P. and Krane, R. J. (1992a) Use of multiple storage elements to improve the second law efficiency of a thermal
storage system. Part I: Analysis of the storage process, in A. Valero and G. Tsatsaronis (eds) ECOS 92 International
Symposium on Efficiency, Cost, Optimization, ASME, New York, pp6166
Sekulic, D. P. and Krane, R. J. (1992b) Use of multiple storage elements to improve the second law efficiency of a thermal
storage system. Part II: Completion of the analysis and presentation of results, in A. Valero and G. Tsatsaronis (eds) ECOS
92 International Symposium on Efficiency, Cost, Optimization, ASME, New York, pp6772
Smith, J. K., Yuan, M., Lopez, T. H., Means, M. and Przybylinski, J. L. (2004) Real time and in-situ detection of calcium
carbonate scale in a west Texas oil field, Society of Petroleum Engineers Production and Facilities, vol 19, pp9499
Snoeyink, V. L. and Jenkins, D. (1980) Water Chemistry, John Wiley & Sons, New York
Solvay Advanced Polymers (2002) Product Data Sheet: P-1700 NT11, NT06 & CL2611, Alpharetta, GA, US
Stumm, W. and Morgan, J. J. (1996) Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters, John Wiley & Sons,
New York
Su, Y. and Davidson, J. H. (2005) Natural convective flow and heat transfer in a collector storage with an immersed heat
exchanger: Numerical study, Journal of Solar Energy Engineering, Transactions of ASME, vol 127, no 3, pp324332
Vliet, G. C. and Baker, D. K. (1998) Designing solar hot water systems for scaling environments, Proceedings of the 1998
American Solar Energy Society Annual Conference, Albuquerque, NM, US, pp307317
Walter, D., Mantell, S. C. and Davidson, J. (2003) Performance of polymer tubes in hot chlorinated water, SOLAR 2003:
Conference Proceedings, American Solar Energy Society, pp7782
Wang, Y. (2004) Experimental study of calcium carbonate scaling on polymer tubes and films, MS Thesis in Chemical
Engineering and Materials Science, University of Minnesota, Minneapolis, US
Wang, Y., Davidson, J. H. and Francis, L. F. (2005) Scaling in polymer tubes and interpretation for use in solar water heating
systems, Journal of Solar Energy Engineering, Transactions of the ASME, vol 127, pp314
Weiss, W. (2003) Solar Heating Systems for Houses: A Design Handbook for Solar Combisystems, International Energy
Agency, London and James and James, London
Wu, C., Mantell, S. C. and Davidson, J. H. (2004) Polymers for domestic solar hot water: Long-term performance of PB and
nylon 6,6 in hot water, Journal of Solar Energy Engineering, Transactions of the ASME, vol 126, no 1, pp581586
ch-05
2/7/07
2:25 PM
Page 130
Photocatalytic Detoxification
of Water with Solar Energy
Sixto Malato, Julin Blanco, Diego C. Alarcn, Manuel
I. Maldonado, Pilar Fernndez-Ibez and Wolfgang Gernjak
Abstract
During the past ten years there has been tremendous research and development in the
area of photocatalysis. One of the major applications of this technology is the
degradation of organic pollutants in water, by what are called advanced oxidation
processes (AOPs). This paper reviews the use of sunlight to produce OH radicals for
AOPs. The pilot plant-scale systems necessary for performing solar photocatalytic
experiments, including the basic pilot plant components and the fundamental
parameters related to solar photocatalytic reactions, are described. The paper also
summarizes most of the recent research related to solar photocatalytic degradation of
water contaminants, and how it could significantly contribute to the treatment of very
persistent toxic compounds. It further describes the possibility of using the solarilluminated photo-Fenton reaction to extend the horizon of solar photocatalysis
applications. Various solar reactors for photocatalytic water treatment based mainly on
non-concentrating collectors erected during the last few years are also described in
detail in the last part of this review.
5.1 INTRODUCTION
The main causes of surface and groundwater contamination are industrial discharges (even
in small amounts), excessive use of pesticides, fertilizers (agro-chemicals) and domestic
waste landfills. Wastewater treatment is based on various mechanical, biological, physical
and chemical processes. In fact, it is a combination of many operations, such as filtration,
flocculation, sterilization or chemical oxidation of organic pollutants. After filtration and
elimination of particles in suspension, the ideal process is biological treatment (natural
decontamination). Unfortunately, some organic pollutants, classified as bio-recalcitrant, are
not biodegradable. Advanced oxidation processes (AOPs) may be used for
ch-05
2/7/07
2:25 PM
Page 131
131
decontamination of water containing these compounds (Andreozzi et al, 1999; Gogate and
Pandit, 2004a; Legrini et al, 1993; Pera-Titus et al, 2004). These methods rely on the
formation of highly reactive chemical species which degrade even the most recalcitrant
molecules into biodegradable compounds. Although there are different reacting systems,
all of them are characterized by the same chemical feature: production of OH radicals
(OH), which are able to oxidize and mineralize almost any organic molecule, yielding CO2
and inorganic ions. Rate constants (kOH, r = kOH [OH] C) for most reactions involving
hydroxyl radicals in aqueous solution are usually in the order of 106 to 109M1s1.
They are also characterized by their not very selective attack, which is a useful
attribute for wastewater treatment and solution of pollution problems. The versatility of
AOPs is also enhanced by the fact that there are different ways of producing hydroxyl
radicals, facilitating compliance with the specific treatment requirements. Methods based
on UV, H2O2/UV, O3/UV and H2O2/O3/UV combinations use photolysis of H2O2 and ozone to
produce the hydroxyl radicals. Other methods, like heterogeneous photocatalysis and
homogeneous photo-Fenton, are based on the use of a wide band-gap semiconductor
and the addition of H2O2 to Fe2+ salts, respectively, and irradiation with UV-VIS light. Since
1976, photocatalytic detoxification has been discussed in the literature as an alternative
method for cleaning up polluted water (Carey et al, 1976). Todays renewed interest
(Pignatello, 1992) in the photo-assisted Fenton process, discovered by Fenton in the 19th
century, is underlined by a significant number of studies devoted to wastewater treatment
(Gogate and Pandit, 2004b). Both these processes are of special interest since sunlight
can be used for them. The main disadvantage of AOPs is their high cost (expensive
reactants such as H2O2 and UV generation). Future applications of these processes could
therefore be improved through the use of catalysis and solar energy.
The heterogeneous solar photocatalytic detoxification process consists of making use
of the near-ultraviolet (UV) band of the solar spectrum (wavelength shorter than 380nm)
to photo-excite a semiconductor catalyst in the presence of oxygen (Goswami and Blake,
1996). Under these circumstances, oxidizing species, either bound hydroxyl radicals (OH)
or free holes, which attack oxidizable contaminants, are generated producing a
progressive break-up of molecules yielding CO2, H2O and diluted inorganic acids. The
most commonly used catalyst is the semiconductor TiO2, which is cheap, non-toxic and
abundant. The homogeneous solar photocatalytic detoxification process (photo-Fenton) is
based on the production of OH radicals by Fenton reagent (H2O2 added to Fe2+ salts). The
rate of degradation of organic pollutants with Fenton-like reagents is strongly accelerated
by irradiation with UV-VIS light. This is an extension of the Fenton process which can make
use of UV-VIS light irradiation at wavelengths over 300nm. Under these conditions, the
photolysis of Fe3+ complexes enables Fe2+ regeneration and Fenton reactions due to the
presence of H2O2 (Bauer et al, 1999; Sagawe et al, 2001).
Although these processes have been studied for at least two decades,
industrial/commercial applications, engineering systems and engineering design
methodologies have only been developed recently (Blanco and Malato, 2003; Goswami
et al, 1997 and 2003). This paper summarizes such work done during the last decade and
attempts to continue previous work published in Advances in Solar Energy on this subject
(Goswami, 1995).
ch-05
2/7/07
2:25 PM
Page 132
132 S. MALATO et al
ch-05
2/7/07
2:25 PM
Page 133
133
the parabolic mirrors (R,). As this surface is outdoors it gets dirty and can be
damaged, requiring that the reflectivity should be measured periodically; and
the spectral transmissivity of the absorber tube (T,).
Figure 5.1 shows the path of direct radiation (ID) until it arrives inside the absorber tube.
It must reach the surface and be reflected (part is lost due to R,) in the right direction
(here affected by s) by the real mirror surface (C), before penetrating (T,) in the tube.
Furthermore, the parabolic trough concentration factor must also be considered (ratio of
surface area of the parabola capturing the radiation and surface area of the tube,
Sp/ST). Therefore the effective photon flux corresponding to the direct UV inside the
absorber (ID,E) is:
I D ,E = f I D ,
Sp
Sr
,c ,R ,Tl
(1)
Global Radiation is also collected by the PTCs, but only that small fraction which falls
directy on the transparent absorber tube without intervention of the collector and is only
affected by the transmissivity of the glass, T,:
I G E = f I G ,T
(2)
where IG,E is the effective photon flux corresponding to global UV inside the absorber and
therefore the total radiation reaching a PTC is the sum of both components.
ID
IG
ID
IG
ID
ID,E
IG,E
FIGURE 5.1 The various loss factors () affecting the photon flux (I) inside a PTC photo-reactor
ch-05
2/7/07
2:25 PM
Page 134
134 S. MALATO et al
Both these facilities, which were followed by others, were based on parabolic trough
collectors with hundreds of square meters of collector surface and can be considered the
starting point of solar photocatalytic technology development (Goswami, 1995). Parabolic
trough collectors make efficient use of direct solar radiation and, as an additional
advantage, the thermal energy collected from the concentrated radiation could
simultaneously be used for other applications. The reactor is small, while receiving a large
amount of energy per volume unit. The flow is turbulent and volatile compounds do not
evaporate, making handling and control of the liquid to be treated simple and inexpensive.
The main disadvantages are that the collectors (i) use only direct radiation, (ii) are
expensive and (iii) have low optical and quantum efficiencies (at least for TiO2 applications)
(Alfano et al, 2000).
sun and is open to the atmosphere (Goslich et al, 1997a), with a catalyst supported in
a permanent matrix on the reactor surface (Guillard et al, 2003) or suspended or
dissolved in the fluid (photo-Fenton applications) (Gernjak et al, 2003). Two large pilotscale fixed-bed photocatalytic reactors were tested in Australia (Feitz et al, 2000).
One (6.5 x 0.5m, total volume 200L) was based on a TiO2-coated (5g m2) woven
glass fiber mesh supported on an inclined corrugated support. The other was based
on a packed bed configuration (1 x 2m2) containing TiO2-coated raschig rings (6.5cm
deep). The packed bed configuration processed a similar concentration of phenol
seven times faster under similar solar light conditions.
Pressurized flat plate. This consists of two plates between which the fluid
circulates using a separating wall (Dillert et al, 1999a). A kind of this nonconcentrating reactor is the double skin sheet reactor (DSSR), which consists of a flat
and transparent structured box made of Plexiglas (Well et al, 1997), with channels
and walls for water circulation. The structure of the reactor is schematically drawn in
Figure 5.2. The suspension containing the model pollutant and the photocatalyst is
ch-05
2/7/07
2:25 PM
Page 135
135
pumped through these channels. The Plexiglas used to manufacture the double skin
sheets has high transmittance in the UV-A portion of the solar spectrum below
400nm. A similar system (based on a double skin acrylic panel) was developed by the
Solar Energy and Energy Conversion Laboratory (University of Florida) for the
treatment of BTEX-contaminated groundwater (Goswami, 1995). A detailed
description and discussion of the experiments can be found in Srinivasan et al (1997
and 1998).
Solar ponds. These are small, shallow on-site pond reactors (Bedford et al, 1994;
Gimnez et al, 1999). The flat reactor resides in a cylindrical tank in which several
perforated tubes lie at the bottom, through which air circulates, bubbling the solution
to be treated. The system operates in a discontinuous mode and the catalyst is
maintained in suspension by using the air stream. Air also supplies the oxygen
necessary for the pollutants oxidation.
Based on all of the above, the main advantages and disadvantages of the different
technologies for solar photocatalytic applications can be summarized.
FIGURE 5.2 Structural schematic view of a DSSR reactor showing water flowing in the transparent box.
The channels are shown connected in series. The channels may also be connected in parallel to reduce
pressure drop
ch-05
2/7/07
2:25 PM
Page 136
136 S. MALATO et al
ch-05
2/7/07
2:25 PM
Page 137
137
being static and collecting diffuse radiation. Thus they also constitute a good option for
solar photochemical applications (May et al, 1991). CPCs are static collectors with a
reflective surface designed to be ideal in the sense of non-imaging optics and can be
designed for any given reactor shape. When the absorber has another shape a more
precise way of describing them would be CPC-type collectors. CPCs were invented in the
1960s (Welford and Winston, 1978) to achieve solar concentration with static devices
(Collares-Pereira, 1995), since they were able to concentrate on the receiver all the radiation
that arrives within the collectors angle of acceptance. They do this by illuminating the
complete perimeter of the receiver, rather than just the front of it, as in conventional flat
plates or in the case of tubes laid side by side. These concentrating devices have ideal
optics, thus maintaining both the advantages of the PTCs and static systems. The
concentration factor (CCPC) of a two dimensional CPC collector is given by Equation 3:
CCPC =
1
sin a
(3)
The normal values for the semi-angle of acceptance (a) for photocatalytic applications are
going to be between 60 and 90 degrees. This wide angle of acceptance allows the receiver
to collect both direct and a large part of the diffuse light (1/C of it), with the additional
advantage of decreasing errors of both the reflective surface and receiver tube alignment,
which become important for achieving a low cost photo-reactor. A special case is that in
which a = 90, whereby CCPC = 1 (non-concentrating solar system) and each CPC curve is an
ordinary involute (Figure 5.3a). When this occurs, all the UV radiation that reaches the aperture
area of the CPC (direct and diffuse) can be collected and redirected to the reactor. If the CPC
is designed for an acceptance angle of +90 to 90, all incident solar diffuse radiation can be
collected. The light reflected by the CPC is distributed all around the tubular receiver (Figure
5.3b) so that almost the entire circumference of the receiver tube is illuminated and the light
incident on the photo-reactor is the same as would impinge on a flat plate. The advantages
of CPCs can be summarized as follows. Just as in the parabolic trough collector, in CPC-type
collectors, the photo-reactor is tubular so that the water can be pumped easily. CPC devices
for photocatalytic applications are generally fabricated with aluminum reflectors and the
structure is usually made of a simple frame which, in turn, forms the support for connecting
the glass tubes that make up the photo-reactor (Romero et al, 1999a; Blanco et al, 2000).
CPCs have the advantage of both technologies (PTCs and non-concentrating collectors) and
none of the disadvantages, apparently making them the best option for photocatalytic
processes based on the use of solar radiation (Ajona and Vidal, 2000; Blanco et al, 2000;
Jubran et al, 1999; Malato et al, 2002a and 2003a; Robert et al, 1999). Advantages include:
1 they can make highly efficient use of both direct and diffuse solar radiation without
the need for solar tracking (Blanco et al, 1999);
2 there is no evaporation of possible volatile compounds and water does not heat up;
3 they have high optical efficiency, since they make use of almost all the available
radiation, and high quantum efficiency, as they do not receive a concentrated flow of
photons; and
4 flow is turbulent inside the tube reactor.
ch-05
2/7/07
2:25 PM
Page 138
138 S. MALATO et al
ch-05
2/7/07
2:25 PM
Page 139
139
Sedlak, 1996; Minero, 1999). They serve to describe the rate of degradation, and may be
used for reactor optimization, but they have no physical meaning and may not be used to
identify surface processes. Thus while not a useful tool for describing the active species
involved in oxidation, engineers and solar designers seem to have a common
understanding of the usefulness of the unmodified L-H model.
For L-H standard data treatment it is therefore assumed that the reaction occurs on the
surface, which is also the assumption most widely accepted as possible. Under these
conditions, two extreme situations are defined to illustrate the adsorption on the catalyst
surface: (1) substrate and water compete for the active catalyst sites, and (2) the reactant and
the solvent are adsorbed on the surface without competing for the same active catalyst sites.
According to the L-H model, the reaction rate (r) is proportional to the fraction of surface
covered by the substrate (x). In each case the following expression can be obtained:
r=-
kr KC
dC
= krx =
dt
1 + KC + K sCs
(4)
k KC
dC
= kr x = r
dt
1 + KC
(5)
r=-
where kr is the reaction rate constant, K is the reactant adsorption constant, C is the
concentration at any time, KS is the solvent adsorption constant and CS is its concentration
(in water CS 55.5 M). As CS >> C and CS remains practically constant, the part of the
catalyst covered by water is unalterable over the whole range of C and the previous
equations can be integrated.
Using an L-H model, graphics similar to those depicted in Figure 5.4 may be obtained
from the experimental data and from the linearization of the previous equations. The
effect of the initial concentration on the degradation rate is shown in Figure 5.4; due to
the saturation produced on the semiconductor surface as the concentration of the
reactant increases, it reaches a point at which the rate becomes steady.
It should be emphasized that photo-decomposition gives rise to intermediates, which
could also be adsorbed competitively on the surface of the catalyst. The concentration of
these intermediates varies throughout the reaction up to their mineralization and thus
Equation 6 may also take the following form:
r0
r01
t1/2
FIGURE 5.4 Graphics related to the adjustment of data to an L-H type kinetic model
2/7/07
2:25 PM
Page 140
140 S. MALATO et al
r=
kr KC
(6)
1 + KC + K iCi (i = 1, n )
i =1
where i is the number of intermediates formed during degradation (the solvent is also
included in the summation).
An understanding of the reaction rates and how the reaction rate is influenced by
different parameters is important for the design and optimization of an industrial system.
The L-H reaction rate constants are useful for comparing the reaction rate under different
experimental conditions. Once the reaction constants kr and K have been evaluated, the
disappearance of reactant can be estimated if all other factors are held constant. Due to
this, a series of tests at different initial substrate concentrations must be performed to
demonstrate whether the experimental results could be adjusted with this model. The
concentration range must be wide enough to allow correct fit of the L-H linearization. This
means from the lowest concentration at which the initial rate could be determined until
the limit where the relationship between initial reaction and initial concentration remains
constant (see Figure 5.4).
Since hydroxyl radicals react non-selectively, numerous intermediates are formed en
route to complete mineralization at different concentrations. Because of this, all tests
should be carried out using TOC (total organic carbon) as a crucial parameter (instead of
concentration of parent compound, C), because the photocatalytic treatment must
destroy not only the initial contaminant, but any other organic compound as well. The
results shown in Figure 5.5 are examples of the experiments carried out with mixtures of
different commercial pesticides. It is possible to see that mineralization, once begun,
maintains the same slope until at least 6070 per cent of the initial TOC has been
degraded.
L1
ch-05
Q, kJ L1
FIGURE 5.5 Pesticide decomposition at different initial concentrations. Maximum rate as a function of
TOCmax is shown in the inserted graphic
ch-05
2/7/07
2:25 PM
Page 141
141
As the reaction is not expected to follow simple models like first- or zero-order
kinetics, overall reaction rate constants cannot be calculated. The complexity of the
results, of course, is caused by the fact that the TOC is a sum parameter often including
many products which undergo manifold reactions. One parameter is proposed in order to
obtain a practical point of comparison for various experiments: the maximum gradient of
the degradation curve, which is the slope of the tangent at the inflection point (rQ,0). This
has the unit of a zero-order rate constant (mg/kJ instead of mg/min) and therefore appears
to be easy to handle. Furthermore this gradient can be roughly considered as the initial
rate of the mineralization reaction, because it is preceded by a period of nearly constant
TOC level. The parameter rQ,0 is referred to as maximum rate. In the graphic insert in
Figure 5.5, it may be observed that the initial rate is steady from 2030mg of TOC per liter.
At this concentration, saturation occurs and the reaction rate becomes constant.
Once the optimum initial concentration is known, a model for predicting plant behavior
is necessary. This model must allow calculation of the area of solar collectors required for
treating water contaminated with different amounts of pesticides. As commented before,
the L-H model is not a perfect explanation of the mechanism of the photocatalytic process,
but the behavior of the reaction rate versus reactant concentration can very often be
adjusted to a mathematical expression with it. In the present case, instead of using the LH model (r = kKC/(1+KC)) directly, the use of an alternative model is preferred for fitting
experimental data in large solar photocatalytic plants, using an approximate kinetic solution
of the general photocatalytic kinetic system, which has the analytical form of an L-H
equation. With these considerations, the rate of TOC disappearance is given by Equation 7
(analogous to the L-H model but without its original significance).
rQ ,0 =
1 [TOC ]max
2 + 3 [TOC ]max
(7)
The experimental results shown in Figure 5.5 have been used to calculate the constants
(i). By inversion of Equation 7 these constants can be calculated from the intercept and
the slope of the line of fit (Equation 8), which is shown in the inset in Figure 5.5.
1 3 2
1
=
+
; 3 = 1.67mg -1 kJ; 2 = 5 .07kJ L-1
1
rQ ,0 1 1 [TOC ]max 1
(8)
Using these values, experimental results and the corresponding lines of fit are shown in
Figure 5.6. The lines of fit were drawn with Equation 9, using the constants reported
previously.
[TOC ]max
1
2 ln
+ 3 [TOC ]max - [TOC ]
1
[TOC ]
) = Q
UV
(9)
The experimental results agree reasonably well with the model proposed and the constants
calculated. This equation allows TOC degradation to be predicted as a function of initial TOC
2/7/07
2:25 PM
Page 142
142 S. MALATO et al
L1
ch-05
Q, kJ L1
FIGURE 5.6 Application of the proposed kinetic model for mineralization of a pesticide mixture. The inset
shows the fit of Equation 9
and available radiation, and the reverse, incident energy on the reactor necessary to reach a
specific degree of mineralization. Therefore, useful design equations may be obtained with
an L-H type model, in spite of not fitting the heterogeneous photocatalytic reaction
mechanism. For now, these equations must be obtained at pilot plant size; however, they
will be useful for larger plants if the same type of collector is used.
ch-05
2/7/07
2:25 PM
Page 143
143
FIGURE 5.7 Average UV direct and UV global (direct + diffuse) irradiance for each month of the year at
Plataforma Solar de Almera, Spain (5 years average)
FIGURE 5.8 Overall dichloroacetic acid degradation rate (as TOC disappearance reaction rate) comparison
for a concentrating (PTC) and a non-concentrating (CPC) collector system. The performance ratio (CPC/PTC)
and the mean value of global UV-radiation during the experimental periods are also shown
performing one test near the 15th of each month during one year. The tests were
performed on perfect, sunny days to permit the parabolic trough to operate all day from
sunrise to sunset. The ratio encountered between both systems working in analogous
conditions for degradation is higher than 5 in favor of the CPC, and this ratio would be
even higher taking cloudy periods into account.
ch-05
2/7/07
2:25 PM
Page 144
144 S. MALATO et al
It has been experimentally shown that above a certain flux of UV photons, the reaction
rate changes from one to a half-order dependence on intensity. Most authors believe that
the transition from r = f (I1.0) to r = f (I0.5) occurs due to the excess of photogenerated
species (e, h+ and OH). A very simple explanation could be the following (based on the
first stages of the process). The first stages considered are: (i) formation of electron/hole
pairs (Equation 10), (ii) recombination of the pairs (Equation 11), and (iii) oxidation of a
reactant R (Equation 12).
(10)
(11)
(12)
(13)
where I is the intensity of incident radiation. If it is considered that [e] [h], then in
stationary state:
(14)
When I is very high, a large number of holes and electrons are generated and therefore
kR[h]2 ko[h]R:
(15)
As the reaction rate depends on the amount of hydroxyl radicals present, and these are
generated in the holes, then r0.5 when I is high. Under these conditions, the quantum
yield diminishes because of the high rate of recombination of e/h+ pairs formed. In the
same manner, when I is small, the inverse is true, kR[h]2 ko[h]R:
(16)
At higher radiation intensities, another transition from r = f (I0.5) to r = f (I0) is produced.
At this moment, mass transfer limits the photocatalytic reaction, therefore the rate
is constant although the radiation increases. There may be several factors that limit
the mass transfer, such as the lack of electron scavengers (O2), organic molecules in the
proximity of the TiO2 surface and/or excess of products occupying active centers of the
ch-05
2/7/07
2:25 PM
Page 145
145
catalyst. These phenomena appear more frequently when working with a supported
catalyst (Pozzo et al, 1997) and/or at a low agitation level. This implies a low catalyst
surface in contact with the liquid and smaller turbulence, which diminishes the contact of
reactants with the catalyst and the diffusion of products from the proximity of the catalyst
to the liquid.
These effects may be appreciably attenuated if some reagent that reduces the
importance of the electron/hole recombination is added. When the electrons are trapped,
recombination of e/h+ is impeded. Either way, addition of oxidants can improve the
efficiency of the process at high illumination intensities. Moreover, this type of compound
can increase the quantum yield even at low irradiation levels due to their strong oxidizing
character. The use of inorganic peroxides has been demonstrated to remarkably enhance
the rate of degradation of different organic contaminants because they trap the photogenerated electrons more efficiently than O2 (Chen et al, 1998; Doong and Chang, 1997;
Malato et al, 1998 and 2000a; Poulios et al, 1998). Hydrogen peroxide is the obvious
candidate and it has been tested with a large number of compounds. Also it is a very
commonly used chemical and so is inexpensive. Being an electron acceptor, hydrogen
peroxide can be a beneficial oxidizing agent because it can react with conduction band
electrons to generate hydroxyl radicals which are required for the photo-mineralization of
organic pollutants (Equation 17):
(17)
The effect depends on H2O2 concentration, generally showing in the optimum range, but
that strength usually depends on the H2O2/contaminant molar ratio. An optimal molar ratio
of between 10 and 100 has been found by different authors (Akmehmet and Ynel., 1996;
Bellobono et al, 1994; Harada et al, 1990; Hofstadler et al, 1994; Poulios et al, 1998).
Inhibition could be explained in terms of TiO2 surface modification by H2O2 adsorption,
scavenging of photo-produced holes and reaction with hydroxyl radicals (Equations 18
and 19). This produces additional problems when working with complex mixtures,
because the molar concentration is usually different for each compound in the mixture.
Moreover, for the treatment of real wastewater, the concentration of each product is
unpredictable.
(18)
(19)
Peroxydisulfate can be a beneficial oxidizing agent in photocatalytical detoxification
because SO4 is formed from the oxidant compound by reaction with the semiconductor
ch-05
2/7/07
2:25 PM
Page 146
146 S. MALATO et al
between molar ratio S2O82/contaminant and reaction rate shows peroxydisulfate as a good
choice. It must be mentioned here that in many highly toxic wastewaters where
degradation of organic pollutants is the major concern, the increase of salt content by
addition of an inorganic anion to enhance the organic degradation rate may be justified.
(20)
(21)
(22)
Other oxidants have been used in photocatalysis for reducing solar (or artificial) UV
exposure time: ClO3, BrO3, IO4 and HSO5. Nevertheless, these additives are very
expensive compared to hydrogen peroxide and peroxydisulfate, and their application
would dramatically increase treatment cost. Even more importantly, they do not dissociate
into harmless products (Br and I), and hundreds of milligrams per liter of these anions
are undesirable in water.
ch-05
2/7/07
2:25 PM
Page 147
147
(23)
In Equation 23, r is called relative photonic efficiency (Serpone and Emeline, 2002). When
the reaction rate for the test substances and phenol (secondary actinometer) are obtained
under identical experimental conditions, there is no need to measure the photon flux. The
use of relative photonic efficiency renders comparison of process efficiencies between
studies carried out in different laboratories or pilot plants possible because r is basically
independent of the fundamental photocatalysis parameters (light intensity, reactor
geometry and TiO2 concentration for a given catalyst). However, it depends on the initial
concentration of substrate and on temperature. In any case, based on initial rates of
degradation, r illustrates only one aspect of photo-degradation and is also useful for
comparing different photocatalyst materials for water treatment purposes. The
mineralization rate (measured by TOC analysis) is also included because efficiencies
based on the disappearance of organic carbon (r ,TOC, Equation 24) provide more practical
information:
r, TOC
(24)
The concentration of oxygen also affects the reaction rate, which is faster when the partial
pressure of oxygen (pO2) in the atmosphere in contact with the water increases. In any
case, it seems that the difference between using air (pO2 = 0.21atm) or pure oxygen
ch-05
2/7/07
2:25 PM
Page 148
148 S. MALATO et al
(pO2 = 1atm) is not drastic. In an industrial plant it would be purely a matter of economy
of design. Injection of pure O2 becomes necessary in once-through plants at low flow
rates. At high flow rates or with recirculation, the addition of oxygen is not always
necessary since the illumination time per pass is short. The water again recovers the
oxygen consumed when it reaches the tank (open to the atmosphere and stirred).
ch-05
2/7/07
2:25 PM
Page 149
149
Fe3+ (related species and organic complexes) absorbs solar photons as a function of
its absorptivity (Mazellier et al, 1997). This effect must be considered when determining
the optimum load as a function of light path length in the photo-reactor. These
calculations must be experimentally demonstrated because they are strongly affected by
different complexes formed by Iron (III) during the photo-Fenton process. The optimum
concentrations of 0.20.5mM were obtained after many experiments with different photoreactors under sunlight (Fallmann et al, 1999; Gernjak et al, 2003; Malato et al, 2002b).
ch-05
2/7/07
2:25 PM
Page 150
150 S. MALATO et al
(26)
In the case of organic compounds containing halogens, Equation 27 shows how the
corresponding halide is formed:
(27)
Under photocatalytic oxidative conditions, sulfur is recovered as sulfate in sulfurcontaining compounds according to Equation 28:
(28)
ch-05
2/7/07
2:25 PM
Page 151
151
concentration of hydrogen ions (decrease in pH). For this reason, the analysis of these two
products of the reaction is of interest for the final mass balance. However, the decrease
in pH is not a very reliable parameter of this balance, except in some cases, because it is
influenced by other processes which take place in the medium: the effect of the TiO2
suspension, the formation of CO2 and intermediates.
The oxidation of carbon atoms into CO2 is relatively easy. It is, however, in general,
markedly slower than the de-aromatization of the molecule. Until now, the absence of
total mineralization has been observed only in the case of s-triazine herbicides, for which
the final product obtained was essentially 1,2,5-triazine-2,4,6, trihydroxy (cyanuric acid),
which is, fortunately, not toxic (Minero et al, 1996b). This is due to the strong stability of
the triazine nucleus, which resists most methods of oxidation. For chlorinated molecules,
Cl ions are easily released into the solution. Nitrogen-containing molecules are
mineralized into NH4+ and mostly NO3. Ammonium ions are relatively stable and the
proportion depends mainly on the initial oxidation degree of nitrogen and on the
irradiation time. Organo-phosphorous pesticides produce phosphate ions. However,
phosphate ions in the pH range used remain adsorbed on TiO2. This strong adsorption
partially inhibits the reaction rate, though it remains acceptable. Until now, the analyses of
aliphatic fragments resulting from the degradation of the aromatic ring have only revealed
formate and acetate ions. Other aliphatics (presumably acids, diacids and hydroxylated
compounds) are very difficult to separate from water and to analyze. Formate and acetate
ions are relatively stable, which in part explains why total mineralization takes much longer
than de-aromatization (Franch et al, 2002). Information concerning degradation of
contaminants at pilot plant-scale using solar collectors is available in several recently
published reviews (Alfano et al, 2000; Bahnemann et al, 2004; Dillert et al, 1999a; Gogate
and Pandit, 2004a; Konstantinou and Albanis, 2003; Malato et al, 2002a; Malato and
Agera, 2003; Pera-Titus et al, 2004).
ch-05
2/7/07
2:25 PM
Page 152
152 S. MALATO et al
it is very easy to calculate the average incident radiation on the collector surface (UVG,n)
for each period of t, and therefore calculate the accumulated energy during this period,
Euv,ges. It is reasonable to divide the mass area ratio of the pollutant by the accumulated
energy, yielding a value for an entity called efficiency: Eff = Si (Euv,ges)1. Its dimension is
mass per unit energy. Instead of using Euv,ges and Eff, we have proposed using QUV and rQ
(Malato et al, 2000c). QUV fits better with the term radiant energy (Q) recommended by
IUPAC (Braslavsky and Houk, 1988) and makes it possible to use C (reactant
concentration) in the plots, which is easier to understand than Si. Besides, rQ is better than
Eff to define a parameter which is not dimensionless. At the same time, we think that
symbol rQ (-dC/ dQUV) is more consistent with r (dC/dt), which is commonly used in the
scientific community.
Consequently, the amount of energy collected by the reactor (per unit of volume) from
the beginning of the treatment until each sample is collected may be found by:
Ar
Vt
(29)
Dt n = tn - t n -1
where tn is the experimental time of each sample and QUV,n is the accumulated energy (per
unit of volume, kJ L1) incident on the reactor for each sample taken during the
experiment. This equation comprises all those proposed by Goslich et al (1997a),
substituting the parameter used by these authors (Euv,ges) by QUV,n.
Figure 5.9 shows the improvement obtained using this equation to calculate the
reaction rate in a two-day photocatalytic degradation experiment with a model compound.
Obviously, UV power changes during the day, and clouds during the first day make this
variation still more noticeable, but with Equation 29, the data for both days can still be
combined and compared with other photocatalytic experiments. Consequently, with QUV,
the initial reaction rate (rQ,0 = -dC/dQUV; C1(QUV) = C0 - kapQUV) is expressed in terms of
pesticide degraded per kJ of UV incident on the collector surface. A similar treatment
could be done for first order kinetics. As Goslich et al have remarked, if rQ (in their case
Eff) is known, collector efficiency is already included through the use of incident surface
radiation, since different rQ, with the same substance and different solar collectors, means
collector efficiency is different.
2/7/07
2:25 PM
Page 153
153
C, mg.L1
C, mg.L1
UVG, W.m2
ch-05
QUV, kJ.L1
FIGURE 5.9 Plots of pesticide concentration as a function of experiment time (top) and accumulated
energy (bottom). Solar UV power throughout the experiment is also shown
As commented by Wolfrum and Turchi (1992), this procedure for deducing rQ is exact only
if the concentration of reactives is constant along the entire reactor (in other words if the
system is at steady state). Solar detoxification pilot plants are frequently operated in a
recirculating batch mode as depicted in Figure 5.10. In this scheme, the fluid is
continuously pumped between a reactor zone and a tank in which no reaction occurs until
the desired degradation is achieved (Bahnemann et al, 2004). Since these are dynamic
systems, the photo-reactor (see Figure 5.10) outlet concentration (C1) is not exactly the
same as the mixed tank outlet concentration (C2). In a large field system, the amount of
conversion each time the mixture passes through the reactor is noticeable (Goswami,
1995). As the relatively clean water in the reactor is mixed with the dirty water in the batch
tank, the water sent from it to the reactor has an increasingly lower concentration. Because
the rate usually decreases with concentration, the overall rate in the reactor responds
likewise. Thus unless properly accounted for, the presence of the tank will alter the
perceived performance of the photo-reactor. Two solutions are available to solve this
problem. The first solution is to use a very high flow rate to achieve low conversion each
time through. This high flow rate must allow more than 1 per cent conversion per pass
(C1(t) C2(t)) to be avoided. We have selected this 1 per cent because it is lower than the
error associated with any analytical chemical method applied for Ci analysis. The second
solution is that the concentration in the reactor outlet at time t, C1(t) is determined from the
inlet concentration (which does not change with time) and the reaction kinetics. However,
because the system is a transient process, the normal steady-state plug flow reactor
ch-05
2/7/07
2:25 PM
Page 154
154 S. MALATO et al
equation cannot be used to model the photo-reactor. Because the inlet concentration
changes with time, C2(t) is defined by what went into the reactor one residence time prior
to t, C2(t-), and the kinetics. The batch tank could still be modeled as a well-mixed tank.
Solving these equations is more difficult than the low-conversion-per-pass case and a
numerical routine is required to fit the data from the batch test. This routine also takes into
account the volume of the piping between the reactor outlet and the batch tank. This
methodology can be used to numerically solve for the concentration profile in any batch
process, even though this method is only necessary when the recirculation flow is not high
enough. As also pointed out by Mehos et al (1992), one must account for the mixing in the
dark tank only if substantial conversion per pass is obtained (C2 > C1).
ch-05
2/7/07
2:25 PM
Page 155
155
night. After this sedimentation period the supernatant liquid (500L) is pumped out of the
tank and the treatment cycle can be started again by filling the tank with a new batch of
wastewater. The initial values of the TOC were determined to vary between 26.9 and
7.3mg L1. During the solar-catalytic treatment a decrease in the concentration of the
organic pollutants, determined as COD (chemical oxygen demand) and TOC, was
observed in most cases. More than 50 per cent of the organic pollutants initially present
in the mixed water inside the pilot plant could thus be degraded within 8 to 11 hours of
illumination. Naturally, the total mass of the degraded contaminants was found to depend
on the initial pollutant concentration, the time of illumination and, in particular, on the solar
UV energy flux density.
In 1997 Freudenhammer et al reported their results from a pilot study using thin film
fixed bed reactors (TFFBRs), which was performed in various Mediterranean countries and
showed that biologically pretreated textile wastewater can be cleaned by solar
photocatalysis with a maximum degradation rate of 3g COD m2 per hour (Freudenhammer
et al, 1997). It was concluded that photocatalysis should be a suitable technology as the final
stage of purification of biologically or physically pretreated wastewater in particularly sunrich areas. However, it was also pointed out that there is a strong demand for cheap
photocatalysts of higher activity for industrial applications of solar wastewater treatment on
a larger scale to be competitive with treatment methods already established on the market.
Based on these results, a pilot plant, financed by the European Commission, has been
built at the site of a textile factory in Tunisia (Menzel Temime). The TFFBR was chosen
because previous studies showed sufficient degradation rates with the selected textile
wastewater in combination with its simple, low cost construction and the low energy
consumption. However the possibility of using suspended catalysts has also been
considered, to integrate results obtained with suspended catalysts showing, in some
cases, a higher efficiency than the fixed system,. The pilot plant and the flow chart have
been published recently (Bahnemann et al, 2004).
Two TFFBRs of width 2.5m and length 10m, corresponding to a total illuminated area of
50m2, were built in concrete and are oriented to the south with an inclination angle of 20.
The reactors can be operated in parallel or as a cascade flow and in a continuous or a
recycling mode, depending on the reaction kinetics. The pumps are designed for a
maximum flow of 3m3 per hour. The plant is connected to the sewerage system of the textile
factory for continuous and long-term experiments. Two sequencing batch reactors (SBRs)
and a membrane aeration system are connected to the TFFBR for pre- and post-treatment,
each with a total volume of 15m3. An automatic injection system of chemical compounds
can be used for pH regulation or other necessities. The plant is designed to operate the two
reactors independently. This allows a comparison of different catalysts, fixation techniques,
hydrodynamics and effluents/model compounds under identical conditions. The plant can
be operated with suspended and fixed catalysts. A sedimentation tank is connected to the
reactors (via a storage tank) to separate and recycle the catalyst. Instrumentation and control
instruments are installed and connected to a computer system to allow automatic
operation: pumps and valves control flow rates, tank levels, UV-A intensity, dissolved
oxygen, pH and conductivity. Experiments are currently being conducted to test the
operation conditions of this newly installed facility (Bousselmi et al, 2004).
ch-05
2/7/07
2:25 PM
Page 156
156 S. MALATO et al
Under the Solar Detoxification Technology for the Treatment of Industrial NonBiodegradable Persistent Chlorinated Water Contaminants (SOLARDETOX) project, a
consortium coordinated by Plataforma Solar de Almera, Spain, has been formed in
Europe for the development and marketing of solar detoxification treatments for
recalcitrant water contaminants. The main project objective was the technical and
economic optimization of real solar detoxification applications. The main innovations were
in the engineering, there being no specific developments in the solar detoxification
technology itself, since it was derived from the already existing solar thermal technology
with only minor modifications. This general objective was divided into several parts:
assessment of the performance and best working conditions for newly developed
The main goal (financed by the EC-DGXII (European Commission Directorate General XII)
through the Brite Euram III Program, 19972000) was to develop a commercial nonconcentrating solar detoxification system using compound parabolic collector (CPC)
technology, with a concentration ratio of one. The beauty of the solar CPC system (as
commented in section 5.2.3) is its intrinsic simplicity while remaining cost-effective and
easy to use it requires low capital investment. Field demonstration was intended to
identify any pre- or post-processing requirements, potential operating problems, and
capital and operating costs.
Based on accumulated experience in pilot plant design, construction and testing (Blanco
et al, 2000), a full-size demonstration plant was erected at the facilities of HIDROCEN (Madrid,
Spain). This plant was designed to treat 1m3 of water contaminated with 100m2 of collector
aperture area (see Figure 5.11). The main plant characteristics are: i) 2 rows of 21 collectors
each; ii) total collector (tilted 40, local latitude) aperture area = 98m2; iii) total loop volume =
675L; iv) total plant volume = 975L; v) 200mg/L TiO2 slurry; and vi) 1.5 x 1.5m collectors with
sixteen 29.2mm internal diameter (ID) tubes. The CPC reflector is made of a highly reflective
anodized aluminum sheet held by a galvanized frame supporting 16 parallel 1.5m-long tubes,
each with an appropriate connector for the adjacent tube. A complete module is formed by
a series of collectors connected in a row. The final prototype plant consists of EW oriented
parallel rows of 21 collectors each. The structure was slightly tilted (1 per cent) in the same
direction to drain rain water and avoid its accumulation in the CPC troughs. Final system
design is completely modular, with the collectors attached in series using HDPE (high density
ch-05
2/7/07
2:25 PM
Page 157
157
polyethylene) quick connections between glass reactor tube absorbers. Water flows
simultaneously through all parallel tubes and there is no limit to the number of collector
components modules may have. Water enters and exits rows through two manifolds at
opposite ends. As this plant is to be a demonstration of what a commercial plant would be
like, operation is fully automatic and maintenance requirements are minimal. Among the
electronic control devices the most innovative system is a solar UV-A sensor that integrates
solar UV with time during the treatment. This sensor is connected to a programmable logic
controller (PLC) and once sufficient energy (see Equation 29) for finishing the treatment has
been achieved (based on preliminary testing for plant design according to the specific
contaminated water to be treated), the PLC stops the main pump and advises the operator
that the treatment has been completed. The PLC also receives other data signals (flow rate,
tank level, temperature, etc) for controlling system pumps and valves.
2m3
2m3
0.1m
0.2m3
0.3m
100m2
800L
~5
0m
FIGURE 5.11 Schematic diagram of solar detoxification demonstration plant constructed in SOLARDETOX
project at HIDROCEN, Madrid
Since late 1999, this plant has been fully functional and its operation is easily described.
Wastewater enters the system from a 2m3 wastewater storage tank while detoxified
water from a previous run returns to the catalyst sedimentation tank. When the system is
completely full (0.8m3), a pump recirculates through the collectors and a small tank
(0.1m3). The concentrated TiO2 slurry and the air necessary for the reaction are injected
into the circuit. Once the water is detoxified, the entire volume passes to the
ch-05
2/7/07
2:25 PM
Page 158
158 S. MALATO et al
sedimentation tank and the system is filled with more wastewater for another batch.
Meanwhile, the detoxified water and the TiO2 in the sedimentation tank undergo pH
adjustment to provoke fast sedimentation of the catalyst (FernndezIbez et al, 2003).
The concentrated catalyst slurry is transferred from the bottom of the tank to another
smaller tank from which the catalyst enters the photo-reactor. The supernatant is removed
through an outlet in the side almost at the bottom of the sedimentation tank and enters
another tank (0.3m3) where the small fraction of the initial catalyst contained in the
supernatant (<7mg L1) is removed by micro-filtration with a suitable membrane before
disposal of water. The membrane has two outlets, one for clean water and one for the
concentrated slurry. The concentrated slurry is recirculated through the membrane until
there are several grams of catalyst per liter. Then this is added to the TiO2 injection tank
and also reused.
More recently (2004) a new CPC-based plant has been installed. This project focuses
on problems that originate in intensive greenhouse agriculture, a sector that has been
growing exponentially in recent years in the Mediterranean basin. There are currently over
200,000 hectares of greenhouses, most of them in EU countries. This type of agriculture
requires up to 200 times more pesticides than conventional agriculture. The
environmental problems are one of the greatest disadvantages for the development of
this economic sector. One of these problems is the uncontrolled dumping of plastic
pesticide containers, which usually still contain residues. This poses a serious risk of
pollution of soil and groundwater. In the area of El Ejido, a town in the province of Almera
in southern Spain, intensive agriculture in 400km2 of greenhouses consumes
approximately 2 million plastic bottles of pesticide per year. So far, these empty plastic
bottles have simply been discarded. Although the amount of product remaining in each
bottle after use is minimal, the numbers are so high that they become a danger to the
environment (poisoning of fauna and flora, not only on land, but also at sea by bottles
carried out to sea by flooding; contamination of water supply by ground filtration). The
solution is to selectively collect these containers for recycling. But before the plastic can
be recycled, it must be washed and the water used for this becomes polluted by the
pesticides. This water must be treated before it is discharged. Therefore, the
development of a simple, clean treatment process for this water in the place where it is
produced is necessary. It is in the detoxification of this water that solar photocatalysis
suggests itself as a very promising process, made even more so by the availability of yearlong strong sunlight as a cheap energy source in this region. Thus the availability of
sunlight and the lack of other alternatives justify the application of a new technology that
has not yet been evaluated on an industrial scale. The ALBAIDA Company and CIEMAT
(Spain) has presented a project entitled Environmental Collection and Recycling of Plastic
Pesticide Bottles using Advance Oxidation Process, Driven by Solar Energy to the
European LIFEENVIRONMENT program, which was approved and began in October
2001. The definitive plant has been in operation since 2004.
The plastic container recycling starts with shredding and then industrial washing of the
shredded plastic, which produces water polluted with highly toxic persistent compounds
(pesticides). The shredded plastic is sent to a series of three baths of the same size (see
Figure 5.12). This is where the separation of the plastic and the wash water would begin.
ch-05
2/7/07
2:25 PM
Page 159
159
The wash water would be drained from each of the baths at floor level and flow to a
storage tank. This water is considered a hazardous toxic product as it carries the dissolved
organic matter that was in the washed pesticide containers. As the water is continuously
recycled and reused, the contaminants must be treated. It is important that care be used
in selecting the containers, since it is very common for the grower to throw away any kind
of used or useless container, while this specific plant is exclusively for plastic.
10mg/L
mg/L
1g/L
CPCs FIELD
(150m2)
FIGURE 5.12 Schematic diagram of the container washing and photocatalytic water treatment
plant concept
Before entering the solar field, the contaminated water is treated physically and
chemically. The water still contains solid waste, which is eliminated by a rough screen. It
then flows to another tank where reactants needed for the photocatalytic reaction are
added. Once the sample is prepared, it enters the solar field of several photo-reactors
connected in series, through which the water flows from one to another. When the water
leaves the solar field, it is tested to see if the process has considerably diminished the
TOC concentration in the treatment water. If not, it is sent back through the solar field
again to undergo the solar detoxification process in the collectors again. When the TOC
measurement is sufficient, the water is reused, re-entering the washing unit in the
container storage bay. When the water has recirculated four or five times, it can be filtered
by an active carbon filter, where any substance resistant to the solar process would be
retained, and discharged into a storage pool for later reuse. The plant was constructed to
treat 1.6m3 of contaminated water with 150m2 of collector aperture area. The main
plant characteristics are: i) 4 rows of 14 collectors each; ii) total collector (tilted 37,
local latitude) aperture area = 150m2; iii) total loop volume = 1060L; iv) total plant
volume = 1600L; v) photo-Fenton using Fe (II) 1mM; and vi) collectors with twenty
ch-05
2/7/07
2:25 PM
Page 160
160 S. MALATO et al
29.2mm ID glass tubes. The final prototype plant consists of EW oriented parallel rows
of 14 collectors each. Final system design is completely modular. Collectors are attached
in series using polypropylene quick connections between glass reactor tube absorbers.
Water flows simultaneously through all four rows and in series through all tubes of each
row. Operation is fully automatic and maintenance requirements are minimal.
(150m )
FIGURE 5.13 Schematic diagram of solar detoxification demonstration plant constructed by ALBAIDA at
La Mojonera, Almera, Spain
ch-05
2/7/07
2:25 PM
Page 161
161
needing to be solved. It is clear that naked TiO2 needs extra help to undertake practical
applications and this may cause it to lose some of the charm of its easy operation.
Two basic lines of R&D (increasing quantum yield) have been working on modifying
catalyst structure and composition and by adding electron acceptors (Herrmann, 1999). A
third approach has focused on finding new catalysts able to work with band gaps which
better overlap the solar spectrum. There have been many attempts within the first and
third approach, such as improving specific surface, doping and deposition with metal ions
and oxides (Blake, 2001). Successful innovative catalyst compositions have been
developed, but they have not been used in large-size plants because no cheap solution
has yet been developed. Our experience in testing at large solar facilities and with
different contaminants qualifies the use of electron acceptors as the most versatile way
of improving reaction rates (for the moment), opening the opportunity to extend the use
of heterogeneous photocatalysis to complicated wastewater (Malato et al, 2002a).
Photo-Fenton treatment presents an interesting alternative to titanium dioxide
photocatalysis, because when the reaction rates of these AOPs are compared, photoFenton shows usually higher reaction rates and is less susceptible to inner filter effects
by the compounds of the wastewater itself (Fallmann et al, 1999; Malato et al, 2002b).
Both issues are mainly due to the fact that iron complexes can absorb light to higher
wavelengths (up to 580nm, Bauer et al, 1999) depending on their composition and on the
homogeneous nature of the process. On the other hand, commonly mentioned
disadvantages of the photo-Fenton treatment are the low pH at which the process is
performed and the need to remove the iron after the process. Therefore its applicability
will be positively influenced by possibilities of water reuse where reactives employed
could be re-valorized, as in fertilizer in irrigation.
FIGURE 5.14 Schematic of a solar treatment coupled with an aerobic biological treatment
ch-05
2/7/07
2:25 PM
Page 162
162 S. MALATO et al
ch-05
2/7/07
2:25 PM
Page 163
163
ACKNOWLEDGEMENTS
The authors wish to thank the European Commission (Research DG) for its financial
assistance within the Energy, Environment and Sustainable Development Program
(Contract No EVK1CT200200122, CADOX Project) and EU-DG Research, Confirming
the International Role of Community Research for Development (Contract No
ICA4CT200210001, SOLWATER Project). They also wish to thank Mrs. Deborah
Fuldauer for correcting the English. Mr. Gernjak wishes to thank the Austrian Academy of
Sciences for financial support through a DOC grant.
REFERENCES
Ajona, J. A. and Vidal, A. (2000) The use of CPC collectors for detoxification of contaminated water: Design, construction
and preliminary results, Solar Energy, no 681, pp109120
Akmehmet, Y. and Ynel, Y. (1996) Photocatalytic degradation of organic contaminants in semiconductor suspensions with
added H2O2, J. Environ. Sci. Health A, vol 31, no 1, pp123138
Alfano, O. M., Bahnemann, D., Cassano, A. E., Dillert, D. and Goslich, R. (2000) Photocatalysis in water environments using
artificial and solar light, Catalysis Today, no 58, pp199230
Anderson, J. V., Link, H., Bohn, M. and Gupta, B. (1991) Development of US solar detoxification technology: An introduction,
Solar En. Mat., no 24, pp538549
Andreozzi, R., Caprio, V., Insola, A. and Martota, R. (1999) Advanced oxidation processes (AOP) for water purification and
recovery, Catalysis Today, no 53, pp5159
Angelakis, A. N., Marekos, M. H. F., Bontoux, L. and Asano, T. (1999) The status of wastewater reuse practice in the
Mediterranean basin: Need for guidelines, Wat. Res., vol 33, no 10, pp22012217
Bahnemann, D. (2004) Photocatalytic Water Treatment: Solar Energy Applications, Solar Energy, no 77, pp445459
Bauer, R., Waldner, G., Fallmann, H., Hager, S., Klare, M., Krutzler, T., Malato, S. and Maletzky, P. (1999) The photo-Fenton
reaction and the TiO2/UV process for waste water treatment Novel developments, Catalysis Today, no 53, pp131144
Bedford, J., Klausner, J. F., Goswami, D. Y. and Schanze, K. S. (1994) Performance of nonconcentrating solar photocatalytic
oxidation reactors part II: Shallow pond configuration, Journal of Solar Energy Engineering, vol 116, no 1, pp813
Bekblet, M., Lindner, M., Weichgrebe, D. and Bahnemann, D. W. (1996) Photocatalytic detoxification with the thin-film fixedbed reactor (TFFBR): Clean-up of highly polluted landfill effluents using a novel TiO2-photocatalyst, Solar Energy, no 56,
pp455469
Bellobono, I. R., Carrara, A., Barni, B. and Gazzotti, A. (1994) Laboratory- and pilot-plant-scale photodegradation of
chloroaliphatics in aqueous solution by photocatalytic membranes inmovilizing titanium dioxide, J. Photochem. Photobiol.
A: Chem., no 84, pp8390
Blake, D. M. (1994) Solar Processes for the Destruction of Hazardous Chemicals. Alternative Fuels and the Environment, Lewis
Publishers, Boca Raton, FL, US
Blake, D. M. (2001) Bibliography of Work on the Photocatalytic Removal of Hazardous Compounds from Water and Air,
National Technical Information Service, US Dept. of Commerce, Springfield, VA, US
ch-05
2/7/07
2:25 PM
Page 164
164 S. MALATO et al
Blake, D. M., Magrini, K., Wolfrum, E. and May, E. K. (1997) Material issues in solar detoxification of air and water, in Optical
Materials Technology for Energy Efficiency and Solar Energy Conversion XV, The Int. Society for Optical Engineering, pp154162
Blanco, J. and Malato, S. (2003) Solar Detoxification, UNESCO Publishing, France
Blanco, J., Malato, S., Maldonado, M.I., Fallmann, H., Krutzler, T. and Bauer, R. (1999) Techno-economical assessment of
solar detoxification systems with compound parabolic collectors, J. Phys., IV, Pr 3, pp259264
Blanco, J., Malato, S., Fernndez, P., Vidal, A., Morales, A., Trincado, P., de Oliveira, J.C., Minero, C., Musci, M., Casalle, C.,
Brunotte, M., Tratzky, S., Dischinger, N., Funken, K-H., Sattler, C., Vincent, M., Collares-Pereira, M., Mendes, J. F. and
Rangel, C. M. (2000) Compound parabolic concentrator technology development to commercial solar detoxification
applications, Solar Energy, no 67, pp317330
Bousselmi, L., Geissen, S. U. and Schroeder, H. (2004) Textile wastewater tratment and reuse by solar photocatalysis: results
from a pilot plant in Tunisia, Wat. Sci. Technol., vol 49, no 4, pp331337
Braslavsky, S. E. and Houk, K. N. (1988) Glossary of terms used in photochemistry, Pure Applied Chem., no 60,
pp10551106
Carey, J. H., Lawrence, J. and Tosine, H. M. (1976) Photodechlorination of PCBs in the presence of TiO2 in aqueous
suspensions, Bull. Environ. Contamination & Toxicol., vol 16, no 6, pp697701
Cassano, A. E. and Alfano, O. E. (2000) Reaction engineering of suspended solids heterogeneous photocatalytic reactors,
Catal. Today, no 58, pp167197
Chen, T., Doong, R. and Lei, W. (1998) Photocatalytic degradation of paration in aqueous TiO2 dispersion, the effect of
hydrogen peroxide and light intensity, Wat. Sci. Technol., vol 37, no 8, pp187194
Collares-Pereira, M. (1995) The relevance of non-imaging optics for solar energy: A review, Revista de Engenharia Tcnica, 3
Cunningham, J. and Sedlak, J. (1996) Kinetic studies of depollution process in TiO2 slurries: Interdependences of adsorption
and UV intensity, Catalysis Today, no 29, pp309315
Curc, D., Malato, S., Blanco, J. and Gimnez, J. (1996) Photocatalysis and radiation absorption in a solar plant, Solar Energy
Materials and Solar Cells, no 44, pp199217
Dillert, R., Cassano, A.E., Goslich, R. and Bahnemann, D. (1999a) Large scale studies in solar catalytic wastewater
treatment, Catalysis Today, no 54, pp267282
Dillert, R., Vollmer, S., Gross, E., Schober, M., Bahnemann, D., Wienefeld, J., Pahlmann, K., Schmedding, T., Arntz, H. J. and
Sager, G. (1999b) Solar-catalytic treatment of an industrial wastewater, Z. Phys. Chem., no 213, pp141147
Doong, R. and Chang, W. (1997) Photoassisted titanium dioxide mediated degradation of organophosphorus pesticides by
hydrogen peroxide, J. Photochem. Photobiol., A: Chem., no 107, pp239244
Fallmann, H., Krutzler, T., Bauer, R., Malato, S. and Blanco, J. (1999) Applicability of the photo-Fenton method for treating
water containing pesticides, Catalysis Today, no 54, pp309319
Feitz, A. J., Boyden, B. H. and Waite, T. D. (2000) Evaluation of two solar pilot scale fixed-bed photocatalytic reactors, Water
Res., no 34, pp39273932
Fernndez-Alba, A. R., Hernando, L., Daz, G. and Chisti, Y. (2001) Toxicity of pesticides in wastewater: a comparative
assessment of rapid bioassays, Anal. Chim. Acta, no 426, pp289301
Fernndez-Alba, A. R., Hernando, D., Agera, A., Cceres, J. and Malato, S. (2002) Toxicity assays: a way for evaluating
AOPs efficiency, Wat. Res., no 36, pp42554262
Fernndez-Ibez, P., Blanco, J., Malato, S. and de las Nieves, F. J. (2003) Application of the colloidal stability of TiO2 particles
for recovery and reuse in solar photocatalysis, Wat. Res., no 37, pp31803188
Franch, M., Aylln, J.A., Peral, J. and Dom`enech, X. (2002) Photocatalytic degradation of short-chain organic diacids,
Catalysis Today, no 76, pp221235
Freudenhammer, H., Bahnemann, D., Bousselmi, L, Geissen, S. U., Ghrabi, A., Saleh, F., Si-Salah, A., Siemon, U. und
Vogelpohl, A. (1997) Detoxification and recycling of wastewater by solar-catalytic treatment, Wat. Sci. Tech., no 35,
pp149156
ch-05
2/7/07
2:25 PM
Page 165
165
Funken, K-H., Sattler, C., Milow, B., de Oliveira, L, Blanco, J., Fernndez, P., Malato, S., Brunotte, M., Dischinger, N., Tratzky,
S., Musci, M. and de Oliveira, J. C. (2001) A comparison of prototype compound parabolic collectorreactors (CPC) on
the road to SOLARDETOX technology, Wat. Sci. Technol., vol 44, no 5, pp271278
Gernjak, W., Krutzler, T., Glaser, A., Malato, S., Cceres, J., Bauer, R. and Fernndez-Alba, A. R. (2003) Photo-Fenton
treatment of water containing natural phenolic pollutants, Chemosphere, no 50, pp7178
Gimnez, J., Curc, D. and Queral, M. A. (1999) Photocatalytic treatment of phenol and 2,4-dichlorophenol in a solar plant in
the way to scaling-up, Catalysis Today, no 54, pp229244
Gogate, P. R. and Pandit, A. B. (2004a) A review of comparative technologies for wastewater treatment. I: Oxidation
technologies at ambient conditions, Adv. Environ. Res., no 8, pp501551
Gogate, P. R. and Pandit, A. B. (2004b) A review of comparative technologies for wastewater treatment. II: Hybrid methods,
Adv. Environ. Res., no 8, pp553597
Goslich, R., Dillert, R. and Bahnemann, D. (1997a) Solar water treatment: Principles and reactors, Wat. Sci. Tech., vol 35, no
4, pp137148
Goslich, R., Bahnemann, D., Schumacher, H. W., Benz, V. and Mller, M. (1997b) Photocatalytic solar wastewater treatment:
New reactor concepts, proceedings of 8th Int. Symp. Solar Thermal Concentrating Technologies, Kln, Germany,
October, pp13371353
Goswami, D. Y. (1995) Engineering of solar photocatalytic detoxification and disinfection processes. Advances in solar
energy, Am. Solar En. Soc., pp165209
Goswami, D. Y. (1997) A review of engineering developments of aqueous phase solar photocatalytic detoxification and
disinfection processes, J. Sol. En. Eng., no 119, pp101107
Goswami, D. Y. and Blake, D. M. (1996) Cleaning up with sunshine, Mech. Engin., August, pp5659
Goswami, D. Y., Sharma, S. K., Mathur, G. D. and Jotshi, C. K. (1997) Techno-economic analysis of solar detoxification
systems, J. Solar En. Eng., no 119, pp108113
Goswami, D. Y., Vijayaraghavan, S., Lu, S. and Tamm, G. (2003) New and emerging developments in solar energy, Solar
Energy, no 76, pp3343
Guillard, C., Disdier, J., Herrmann, J. M., Lehaut, C., Chopin, T., Malato, S. and Blanco, J. (1999) Comparison of various titania
samples of industrial origin in the solar photocatalytic detoxification of water containing 4-chlorophenol, Catalysis Today,
no 54, pp217228
Guillard, C., Disdier, J., Monnet, Ch., Dussaud, J., Malato, S., Blanco, J., Maldonado, M. I. and Herrmann, J. M. (2003) Solar
efficiency of a new deposited titania photocatalyst: Chlorophenol, pesticide and dye removal applications, Appl. Catal. B:
Environ., no 46, pp319332
Hayo, M. G. (1996) Assessing the impact of pesticides on the environment, Agric. Ecosys. Environ., no 60, pp8196
Harada K., Hisanaga, T. and Tanaka, K. (1990) Photocatalytic degradation of organophosphorous insecticides in aqueous
semiconductor suspensions, Wat. Res., vol 24, no 11, pp14151417
Herrmann, J. M. (1999) Heterogeneous photocatalysis: Fundamentals and applications to the removal of various types of
aqueous pollutants, Catalysis Today, no 53, pp115129
Hofstadler, K., Bauer, R., Novalic, S. and Heisler, G. (1994) New reactor design for photocatalytic wastewater treatment with
TiO2 inmobilized on fused silica glass fibers: Photomineralization of 4-chlorophenol, Environ. Sci. Technol., vol 28, no 4, pp
670674
Jubran, B. A., Ismail, A. F. and Pervez, T. (1999) Prediction of the performance of a solar water detoxification system under
Malaysian climatic conditions, Energy Convers. Manage., no 41, pp112
Konstantinou, I. K. and Albanis, T. A. (2003) Photocatalytic transformation of pesticides in aqueous titanium dioxide
suspensions using artificial and solar light: Intermediates and degradation pathways, Appl. Catal. B: Environ., no 42,
pp319335
ch-05
2/7/07
2:25 PM
Page 166
166 S. MALATO et al
Lee, C. and Yoon, J. (2004) Temperature dependence of hydroxyl radical formation in the hv/Fe3+/H2O2 and Fe3+/H2O2
systems, Chemosphere, no 56, pp923934
Legrini O., Oliveros, E. and Braun, A. M. (1993) Photochemical processes for water treatment, Chem. Rev., no 93, pp671698
Malato, S. and Agera, A. (2003) Detoxification of pesticides in water using solar photocatalysis, In Pesticide
Decontamination and Detoxification, ACS Symposium Series 863, Am. Chem. Soc., Washington DC, pp113126
Malato, S., Richter, C., Blanco, J. and Vincent, M. (1996) Photocatalytic degradation of industrial residual waters, Solar
Energy, vol 56, no 5, pp401410
Malato, S., Blanco, J., Richter, C., Braun, B. and Maldonado, M. I. (1998) Enhancement of the rate of solar photocatalytic
mineralization of organic pollutants by inorganic species, Appl. Catal. B: Environ., no 17, pp347356
Malato, S., Blanco, J., Maldonado, M. I., Fernndez-Ibez, P. and Campos, A. (2000a) Optimizing solar photocatalytic
mineralization of pesticides by adding inorganic oxidising species: Application to the recycling of pesticide containers,
Appl. Catal. B: Environ., no 28, pp163174
Malato, S., Blanco, J., Richter, C. and Maldonado, M. I. (2000b) Optimization of pre-industrial solar photocatalytic
mineralization of commercial pesticides; application to pesticide container recycling, Appl. Catal. B: Environ., no 25,
pp3138
Malato, S., Blanco, J., Richter, C., Fernndez, P. and Maldonado, M. I. (2000c) Solar photocatalytic mineralization of
commercial pesticides: Oxamyl, Solar En. Mat. Solar Cells, no 64, pp114
Malato, S., Blanco, J., Vidal, A. and Richter, C. (2002a) Photocatalysis with solar energy at a pilot-plant scale: An overview,
Appl. Catal. B: Environ., no 37, pp115
Malato, S., Blanco, J., Cceres, J., Fernndez-Alba, A. R., Agera, A. and Rodrguez, A. (2002b) Photocatalytic treatment of
water-soluble pesticides by photo-Fenton and TiO2 using solar energy, Catalysis Today, no 76, pp209220
Malato, S., Blanco, J., Vidal, A., Alarcn, D., Maldonado, M. I., Cceres, J. and Gernjak, W. (2003a) Applied studies in solar
photocatalytic detoxification: An overview, Solar Energy, no 75, pp329336
Malato, S., Blanco, J., Campos, A., Cceres, J., Guillard, C., Herrmann, J. M. and Fernndez-Alba, A. R., (2003b) Effect of
operating parameters on the testing of new industrial titania catalysts at solar pilot plant scale, Appl. Catal. B: Environ.,
no 42, pp349357
Malato, S., Cceres, J., Fernndez-Alba, A. R., Piedra, L., Hernando, M. D., Agera, A. and Vial, J. (2003c) Photocatalytic
treatment of diuron by solar photocatalysis: Evaluation of main intermediates and toxicity, Env. Sci. Technol., no 37,
pp25162524
May, E. K., Gee, R., Wickham, D. T., Lafloon, L. A. and Wright, J. D. (1991) Design and fabrication of a prototype solar
receiver/reactors for the solar detoxification of contaminated water, NREL Report, Industrial Solar Technology Corp.,
Golden, Colorado, US
Mazellier, P., Jirkovsky, J. and Bolte, M. (1997) Degradation of diuron photoinduced by iron III in aqueous solution, Pestic.
Sci., no 49, pp259267
Mehos, M., Turchi, C., Pacheco, J., Bogel, A. J., Merrill, T. and Stanley, R. (1992) Pilot-scale study of the solar detoxification
of VOC-contaminated groundwater, National Renewable Energy Laboratory, Golden, Colorado, US
Minero, C. (1999) Kinetic analysis of photoinduced reactions at the water semiconductor interface, Catalysis Today, no 54,
pp205216
Minero, C., Pelizzetti, E., Malato, S. and Blanco, J. (1993) Large solar plant photocatalytic water decontamination:
Degradation of pentachlorophenol, Chemosphere, vol 26, no 12, pp21032119
Minero, C., Pelizzetti, E., Malato, S. and Blanco, J. (1996a) Large solar plant photocatalytic water decontamination: Effect of
operational parameters, Solar Energy, no 56, pp421428
Minero, C., Pelizzetti, E., Malato, S. and Blanco, J. (1996b) Large solar plant photocatalytic water decontamination:
degradation of atrazine, Solar Energy, no 56, pp411419
ch-05
2/7/07
2:25 PM
Page 167
167
Mukherje, P. S. and Ray, A. K. (1999) Major challenges in the design of a large-scale photocatalytic reactor for water
treatment, Chem. Eng. Technol., no 22, pp253260
Pacheco, J. E. and Tyner, C.E. (1990) Enhancement of processes for solar photocatalytic detoxification of water, Solar
Engineering, no 1990, pp163169
Pacheco, J., Prairie, M., Evans, L. and Yellowhorse, L. (1990) Engineering-scale experiments of solar photocatalytic oxidation
of trichloroethylene, proceedings of 25th Intersociety Energy Conversion Engineering Conference, Reno, Nevada, US
Parent, Y., Blake, D., Magrini-Bair, K., Lyons, C., Turchi, C., Watt, A., Wolfrum, E. and Prairie, M. (1996) Solar photocatalytic
process for the purification of water: State of development and barriers to commercialization, Solar Energy, vol 56, no 5,
pp429438
Pera-Titus, M., Garca-Molina, V., Baos, M. A., Gimnez, J. and Esplugas, S. (2004) Degradation of chlorophenols by means
of advanced oxidation proceses: A general review, Appl. Catal. B: Environ., no 47, pp219256
Pignatello, J. J. (1992) Dark and photoassisted Fe3+ catalyzed degradation of chlorophenoxy herbicides by hydrogen
peroxide, Environ. Sci. Technol., no 26, pp944951
Poulios, I., Kositzi, M. and Kouras, A. (1998) Photocatalytic decomposition of triclopyr over aqueous semiconductor
suspensions, J. Photochem. Photobiol. A: Chem., no 115, pp175183
Pozzo, R. L., Baltans, M. A. and Cassano, A. E. (1997) Supported titanium oxide as photocatalyst in water decontamination:
State of the art, Catal. Today, no 39, pp219231
Prairie, M. R., Pacheco, J. E. and Evans, L. R. (1992) Solar detoxification of water containing chlorinated solvents and heavy
metals via TiO2 photocatalysis, Solar Engineering, no 1992, pp18
Robert, D., Piscopo, A., Heintz, O. and Weber, V. (1999) Photocatalytic detoxification with TiO2 supported on glass-fibre by
using artificial and natural light, Catalysis Today, no 54, pp291296
Romero, M., Vidal, A., Senhaji, F. and El-Hraiki, A. (1999a) Application of a CPC-based photocatalytic reactor to detoxification
of washing waters from agricultural industries, J. Phys., vol IV, no 9, pp283288
Romero, M., Blanco, J., Snchez, B., Vidal, A., Malato, S., Cardona, A. and Garca, E. (1999b) Solar photocatalytic
degradation of water and air pollutants: Challenges and perspectives, Solar Energy, vol 66 no 2, pp169182
Sagawe, G., Lehnard, A., Lbber, M. and Bahnemann, D. (2001) The insulated solar Fenton hybrid process: Fundamental
investigations, Helv. Chem. Acta, no 86, pp37423758
Sagawe, G., Brandi, R. J., Bahnemann, D. and Cassano, A. E. (2003a) Photocatalytic reactors for treating water pollution with
solar illumination. I: A simplified analysis of batch reactors, Chem. Eng. Sci., no 58, pp25872599
Sagawe, G., Brandi, R. J., Bahnemann, D. and Cassano, A. E. (2003b) Photocatalytic reactor for treating water pollution with
solar illumination. II: A simplified analysis for flow reactors, Chem. Eng. Sci., no 58, pp26012615
Sarria, V., Kenfack, S., Guillod, O. and Pulgarn, C. (2003) An innovative coupled solar-biological system at field pilot scale for
the treatment of biorecalcitrant compounds, J. Photochem. Photobiol. A: Chem., no 159, pp8999
Serpone, N. and Emeline, A. V. (2002) Suggested terms and definitions in photocatalysis and radiocatalysis, Int. J.
Photoenergy, no 4, pp91131
Serpone, N., Sauv, G., Koch, R., Tahiri, H., Pichat, P., Piccini, P., Pelizzetti, E. and Hidaka, H. (1996) Standardization protocol
of process efficiencies and activation parameters in heterogeneous photocatalysis: Relative photonic efficiencies, J.
Photochem. Photobiol. A: Chem., no94, pp191203
Srinivasan, M., Klausner, J. F., Jotshi, C. K. and Goswami, D. Y. (1997) Solar photocatalytic treatment of BTEX contaminated
groundwater, proceedings of the ISES 1997 Solar World Congress, August 1997
Srinivasan, M., Goswami, D. Y., Klausner, J. F. and Jotshi, C. K. (1998) Design, construction and testing of a solar
photocatalytic treatment facility for BTEX contaminated groundwater. Solar engineering 1998, proceedings of the ASME
Intern. Solar Energy, Albuquerque, NM, US, pp299304
ch-05
2/7/07
2:25 PM
Page 168
168 S. MALATO et al
Tothill, I. E. and Turner, A. P. F. (1996) Developments in bioassay methods for toxicity testing in water treatment, Trends Anal.
Chem., vol 15, no 5, pp178186
Welford, W. T. and Winston, R. (1978) The Optics of Non-Imaging Concentrators: Light and Solar Energy, Academic Press Inc.,
New York
Well, M., Dillert, R. H. G., Bahnemann, D. W., Benz, V. W. and Mueller, M. A. (1997) A novel nonconcentrating reactor for
solar water detoxification, J. Sol. En. Eng., no 119, pp114119
Wolfrum, E. I. and Turchi, S. (1992) Comments on reactor dynamics in the evaluation of photocatalytic oxidation kinetics, J.
Catal., no 136, pp626628
Wyness, P., Klausner, J. F., Goswami, D. Y. and Schanze, K. S. (1994a) Performance of nonconcentrating solar photocatalytic
oxidation reactors, Part I, J. Solar Energy Engineering, no 116, pp27
Wyness, P., Klausner, J. F., Goswami, D. Y. and Schance, K. S. (1994b) Performance of non-concentrating solar photocatalytic
oxidation reactors, Part II: Shallow pond configuration, J. Sol. En Eng., no 116, pp813
Zaidi, A. H., Goswami, D. Y. and Wilkie, A. C. (1995) Solar photocatalytic post-treatment of anaerobically digested distillery
effluent, proceedings of Am. Solar Energy Soc. Annual Conf., Minneapolis, MN, US, July, pp5156
Ch-06
2/6/07
7:06 PM
Page 169
Solar-Hydrogen:
A Solid-State Chemistry
Perspective
J. Nowotny, T. Bak, L. R. Sheppard and C. C. Sorrell
Abstract
The present work considers the solid-state chemistry-related issues relevant to the
photo-electrochemical generation of hydrogen from water using solar energy. The focus
is on the properties of semiconducting photo-electrodes, which comprise the critical
component for increasing the efficiency of the conversion of solar energy into chemical
energy. The most important parameters of the photo-electrodes, which are essential for
the conversion and to be modified to this end, are the band gap and the Fermi energy.
The relationships between these properties and the materials properties that can be
experimentally determined conveniently (electrical conductivity, thermoelectric power
and work function) are outlined.
The present paper brings together the concepts of photo-electrochemistry with the
concepts of defect chemistry and solid-state electrochemistry. It is considered that the
Fermi energy (the chemical potential of electrons) at the surface of the photo-electrode
is the key quantity in the assessment of the reactivity of photo-electrode materials with
water. Consequently, the desired reactivity between the photo-electrode and water may
be achieved through appropriate modification of this quantity during processing or
subsequent treatment of the photo-electrode material. Further, the optimal photoelectrode for water photolysis with high solar energy conversion efficiency should
exhibit a band gap of ~2eV and be resistant to corrosion in water. Finally, the reactivity
between the photo-electrode and water must be considered in terms of both collective
properties, such as electronic structure, and local properties related to photocatalytic
active surface sites at which the photocatalytic reaction between water and the photoelectrode takes place. The leading candidate for such photo-electrodes is titania (TiO2-x),
which exhibits outstanding resistance to corrosion and photo-corrosion in aqueous
environments.
The optimal reactivity of titania with water, leading to maximal solar energy
conversion efficiencies, can be achieved through maximization of solar energy
absorption and minimization of energy losses due to recombination and charge transport.
This may be achieved through the modification of defect disorder and related properties
including charge transport, charge separation and electronic structure. Progress in
research on the determination of the relevant properties of titania essential to the
2/6/07
7:06 PM
170
Page 170
performance of photo-electrodes, including defect chemistry and related electroactivity, is discussed. Significance, benefits and advantages of the solar-hydrogen
technology and cost-related estimates are briefly considered.
Keywords solar-hydrogen; water photolysis; titanium dioxide; photo-electrode; charge transfer; defect
chemistry; defect disorder; photo-electrochemistry; solar energy conversion
6.1 INTRODUCTION
6.1.1 GLOBAL WARMING
Increasing environmental pollution is becoming a vital global concern (Ooki, 1998). Figure
6.1, showing the increase in CO2 emissions over the period 17002000, provides a
dramatic illustration of the extent of the problem (Neftel et al, 1985; Friedli et al, 1986;
Russ, 1994). In order to reverse the global warming associated with CO2 emissions, it will
be imperative to reduce emissions of greenhouse gases, which are formed as a result of
the burning of fossil fuels. Therefore there is an urgent need to increase the production
of energy which is environmentally clean. This may be achieved through the development
of novel materials required for the conversion of renewable energy, such as solar energy,
into other types of energy, such as chemical energy (for example hydrogen). It is believed
that an understanding of materials interfaces is essential for the development of such
materials through interface engineering (Hirano et al, 1999).
Centre for Materials Research in Energy Conversion, UNSW
Ch-06
340
ATMOSPHERIC MEASUREMENTS
ICE PROBING
320
300
280
1800
1850
1900
1950
2000
YEAR
Sources: Neftel et al (1985), Friedli et al (1986) and Russ (1994)
FIGURE 6.1 Emission of carbon dioxide over the period 17002000 according to direct atmospheric
measurements and ice probing
Ch-06
2/6/07
7:06 PM
Page 171
171
min
i
n
p
(e)
o(e)
2/6/07
172
R
s
[ ]
7:06 PM
Page 172
Frequency [Hz]
Work function [eV]
Work function of reference electrode [eV]
Work function component related to the surface charge-induced potential barrier
[eV]
Thermovoltage (electrical potential) [V]
Concentration [atomic ratio]
The symbols for point defects are according to the Kroger-Vink notation (Kroger, 1974)
ISS
ION
480
TU
AL
EM
460
440
T
RGE
AC
Ch-06
A
TO T
KYO
420
1990
1995
2000
2005
2010
YEAR
FIGURE 6.2 Emission of carbon dioxide in Australia over the period 19902000 and the emission level
according to the Kyoto target, according to the United Nations Environmental Programme, 2002
The list of problems related to pollution is long and has resulted in the understanding
that a reduction in greenhouse gas emissions must be addressed immediately.
Consequently, there is an increasing need to scale down the combustion of fossil fuels
Ch-06
2/6/07
7:06 PM
Page 173
173
and to develop alternative sources of energy that do not result in the emission of
greenhouse, pollutant or toxic gases and that are, therefore, environmentally friendly.
While there is a widespread perception that hydrogen is such form of energy, one must
ask the question: Is this perception correct?
increase.
Energy capacity: Hydrogen is the fuel with the highest energy capacity.
Amount of pollution: The absence of pollution during combustion (except nitrogen
Ch-06
2/6/07
174
7:06 PM
Page 174
that urban pollution will be reduced by removing power generation plants to less
populated areas.
Security of resources: Fossil fuel reserves are limited in quantity and they are
concentrated in politically uncertain regions.
Hydrogen generated from water using renewable sources of energy can resolve the issue
of the availability of a fuel that is 100 per cent environmentally safe.1 Consequently, the
development of hydrogen technologies using renewable sources of energy will reduce
the dominance of resource-rich nations in international energy markets. Such
technologies will provide these nations with increased energy security.
Substantial seed funding already has been allocated to research and development
specifically for solar-hydrogen technologies in the US, Japan and the European Union
(EU). These countries have recognized that it is in their best national interests not to be
left behind in the race to develop these technologies.
Recently, the Australian Government has entered the race to develop hydrogen-related
technologies by allocating funding through the Department of Industry, Tourism and
Resources to study the potential for an Australian hydrogen industry (Rand and Badwal,
2005) and to organize an international meeting on The Hydrogen Economy (1821 May
2003). The meeting aimed at mapping the future of hydrogen as an important potential
long-term source of energy for Australia and identifying Australias contribution to world
hydrogen developments (Hartley Management Group, 2003). The location of the meeting
site in Broome was not incidental since the Pacific coastline in this area is exposed to the
worlds largest tide (up to 11m) and so represents a potential for utilizing a substantial
amount of environmentally clean tidal energy.
The most critical issue in the development of hydrogen energy, and specifically the
technologies for the conversion of renewable energy into other forms of energy, such as
chemical energy (hydrogen), is the development of a special class of materials required
for efficient and clean conversion of energy. Development of these materials, which will
need to exhibit sophisticated functional properties, will require applying the most recent
progress in the science of materials interfaces and solid-state science. These issues were
addressed at the First International Conference on Materials for Hydrogen Energy held
in Sydney on 27 August 2004.
6.2 SOLAR-HYDROGEN
One of the most promising renewable energy technologies is the production of hydrogen
by water photolysis (Veziroglu, 1998 and 2000; Bockris et al, 1991; Bockris 1999; Bak
et al, 2002). The current predictions indicate that the production of hydrogen will
skyrocket by 2010 (Thomas et al, 1998). According to a recent comprehensive review the
method of photo-electrochemical water decomposition using solar energy is the most
promising method for the generation of hydrogen (Bak et al, 2002). The key functional
element in photo-electrochemical devices which allows the exploitation of this technology
is the photo-electrode (PE). However, there is a need to enhance the performance of this
element in order to achieve the level required for commercialization of the device
Ch-06
2/6/07
7:06 PM
Page 175
175
technology. Therefore the primary goal of research and development activities in many
research groups is the development of photo-electrodes with maximal efficiencies of
conversion of solar energy into chemical energy, viz hydrogen.
Ch-06
2/6/07
176
7:06 PM
Page 176
O2
H2
H2O 2e + O2 + 2H
2H + 2e H2
PHOTO-ANODE
CATHODE
+
H2O H + OH
AQUEOUS ELECTROLYTE
FIGURE 6.3 Electrochemical cell showing the principle of electrochemical hydrogen generation through
water photolysis using solar energy and related reactions
R
e
h
EF
+
H /H2
EC
1.23 eV
Eg
O2 /H2O
EV
H2O + 2 h 2 H + O2
PHOTO-ANODE
(SEMICONDUCTOR)
H + e H2
AQUEOUS
ELECTROLYTE
CATHODE
(METAL)
FIGURE 6.4 Electrical circuit representing the photo-electrochemical device formed by a semiconducting
photo-anode and metallic cathode
Ch-06
2/6/07
7:06 PM
Page 177
177
O2
h
1
2
e
Pt
H2
TiO2-x
H+
AQUEOUS SOLUTION
Pt
e
H2O + 2h 2H+ + O2
2H+ + 2e H2
R
V
FIGURE 6.5 The electrochemical chain of TiO2-x-based photo-electrochemical device and related
charge transfer
An essential part of the device for hydrogen generation using solar energy is the
semiconducting photo-electrode, which consists of an n-type semiconductor. Exposure to
light results in intrinsic ionization over the band gap, leading to the formation of an
electronhole pair:
(1)
The excess minority charge carriers (electron holes) gives rise to a photo-voltage, resulting
in splitting of the water molecule into hydrogen ions and gaseous oxygen:
(2)
Gaseous oxygen evolves at the photo-anode and the hydrogen ions migrate to the
cathode through the internal circuit (electrolyte), where the reduction of hydrogen ions to
gaseous hydrogen takes place:
(3)
The overall reaction of the photo-electrochemical process is as follows:
(4)
The preceding reaction takes place when:
the electromotive force (EMF) of the photo-electrochemical cell is 1.23V (in practice
Ch-06
2/6/07
7:06 PM
178
Page 178
the energy of the photons is equal to or greater than the band gap of the
photo-electrode.
The economic feasibility of photo-electrochemical hydrogen generation technology
depends principally on meeting the following criteria:
the ECE must be 10 per cent, according to the USDOE (Service, 2002), where ECE
is defined as the ratio of the amount of the energy output, equal to the generated
chemical or electrical energy, to the energy input, equal to the solar energy striking
the photo-electrode;
the photo-electrode must be durable and resistant to corrosion and photo-corrosion
in aqueous environments;
the overall cost of the device and infrastructure must be at least comparable to the
value of the produced hydrogen, if not lower or substantially lower; and
the device should be either maintenance free or low maintenance.
properties of PEs. In the case of water photolysis, the band gap should remain
between 1.23eV (theoretical lower limit) and ~2eV (practical upper limit). The
electronic structure and related band gap of semiconductors based on
nonstoichiometric compounds, such as TiO2-x, are very sensitive to defect disorder
Ch-06
2/6/07
7:06 PM
Page 179
179
(Khan et al, 2002). On the other hand, in the case of single-crystal Si, used for
photovoltaic (PV) cells, the band gap is fixed at 1.1eV and it is a material property.
Therefore, the electronic structure of nonstoichiometric compounds may be modified
through the control of the defect chemistry.
Electrical resistance: Efficient charge transfer within the circuit of a
photo-electrochemical cell requires the electrical resistance to be at its minimal level.
Consequently, this requires the resistance of all elements of the circuit, including the
photo-electrode, to be minimized. The resistivity of the photo-electrode can be
minimized through the imposition of suitable defect disorder. Concurrently, the
materials engineering must be undertaken so as to retain or achieve the optimal band
gap.
Schottky barrier: The Schottky-type barrier is needed for effective charge separation
(prevention of recombination of the electron-hole pair). The consequence of this
separation is that electrons are transported into the n-type solid while electron holes
are transported toward the photo-electrode/electrolyte interface, resulting in the
formation of a p-type surface layer (Bak et al, 2002). In the case of nonstoichiometric
compounds used as photo-electrodes, the potential barrier may be formed either by
the imposition of a defect concentration gradient during processing or through the
effect of segregation leading to enrichment of the surface layer. These have the
effect of imposing a surface Fermi energy at a level different from that of the bulk
phase.
Microstructure: The properties of polycrystalline solids are well known to be very
sensitive to microstructure. The effects of the microstructure on properties may be
considered in terms of, inter alia, grain size and the concentration of grain
boundaries. The effect of the surface microstructure may be considered in terms of
the local properties of the linear defects formed at the intersections between the
grain boundaries and the external surface. So far, little is known of the effects of
these defects on the ECE. Further, the effect of the grain size (bulk/surface) may be
considered in terms of the interfacial energy and its impact on the local defect
chemistry and related charge transfer across the photo-electrode/electrolyte
interface.
Flat band potential: The flat band potential should be higher than the redox
potential of the H+/H2 couple. This potential may be engineered by the imposition of
specific surface-to-bulk composition gradient and related potential gradient. The flat
band potential was described by Bak et al (2002).
Helmholtz potential barrier: The Helmholtz potential barrier, which is formed at the
interface between the photo-electrode and the electrolyte, plays an important role in
retarding/accelerating the charge transfer within the electrochemical chain shown in
Figure 6.5 (Bak et al, 2002). Therefore its value is determined by the surface
compositions of the photo-electrode and the composition of the electrolyte.
Nonstoichiometry and related defect disorder: Although it is well known that
nonstoichiometry and point defects have a substantial impact on material properties,
so far little is known on the effect of defect disorder on electroactivity and
photo-sensitivity. Awareness is growing that properties of photo-electrodes based on
Ch-06
2/6/07
180
7:06 PM
Page 180
Ch-06
2/6/07
7:06 PM
Page 181
181
The key energy quantities describing the electronic structure of solids and having an
impact on the photo-sensitivity include:
It is well known that the width of the forbidden gap (Eg = EV EC) is the key property
controlling ECE (Bak et al, 2002; Chandra, 1985; Gerischer, 1977 and 1997; Seraphin,
1979). Another quantity that has a crucial impact on the reactivity of photo-electrodes is
the Fermi energy.
There is a close relationship between the Fermi energy and the lattice defect disorder
of semiconducting oxides. This involves point defects that form either donors, such as
oxygen vacancies and cation interstitials, or acceptors, such as cation vacancies. The
Fermi energy is determined by the location of the associated energy levels within the
band gap and the related density of states. Doping of metal oxides with aliovalent ions
results in the formation of either donors or acceptors located within the band gap. The
positions of these levels are determined by the incorporation mechanisms of the specific
ions into the lattice of the host semiconducting oxides. The densities of states are
controlled by the concentration of dopants within their solubility ranges.
The mobility of electronic charge carriers within the bands or mid-gap levels is
determined by the density of allowed states forming the bands and the donor and acceptor
states. Specifically, at high and low density of states, the charge carriers are quasi-free or
localized respectively. At low density of states, when these levels are isolated, the charge
transport occurs by the hopping mechanism. Increase of the density of states to very high
values may, in extreme cases, lead to metallic conduction (Sheppard, 2004).
Detailed analysis of the electronic structure of semiconducting solids is beyond the
scope of the present work, which is limited to a brief analysis of the following two
quantities:
1 band gap; and
2 Fermi energy.
These two quantities are selected for consideration for the following reasons:
They are the most important in the assessment of photo-sensitivity and reactivity at
Ch-06
2/6/07
182
7:06 PM
Page 182
of the most critical factors that control the properties of intrinsic semiconductivity in
general and photo-sensitivity in particular. The band gap is the energy difference through
which heat- and photon-induced ionization occur. This ionization results in the formation
of quasi-free charge carriers (electrons and electron holes) that take part in the charge
transfer within the electrode and its electroactivity (certain density of states is required for
the charge carriers to be quasi-free, alternatively they are isolated). In other words, only
the photons of the energy equal to or larger than that of the band gap may be absorbed
and be available for conversion of the electromagnetic energy into another form of energy,
such as electrical and chemical. The critical impact of Eg on photo-sensitivity and ECE has
been discussed previously (Bak et al, 2002).
E0=0
O
1
EC
EF
Eg
S=-EF
EC
EF
O
O
EV
EV
fn ( E ) = 1 + exp ( Ei - E F ) / kT
-1
(5)
where fn(E) is the probability of occupation of the energy level Ei. The similar
expression for fp(E) describes the distribution of electron holes:
(6)
Ch-06
2/6/07
7:06 PM
Page 183
183
(9)
(10)
where n and p denote the concentrations of electrons and electron holes respectively, and
Nn and Np denote the densities of states of electrons and electron holes respectively.
2/6/07
184
7:06 PM
Page 184
Equations 9 and 10 may be used for the determination of the basic electronic structure
when the concentrations of the electronic carriers are known (or vice versa). In the case
of a flat-band structure, corresponding to semiconductors that are free of surface charge,
the value of EF is independent of the distance from the surface. Alternatively, the presence
of a surface charge, caused, for example, by the adsorption of charged species, results in
a difference in the Fermi energies of the surface and the bulk by the energy component
s, as shown on the right side of Figure 6.6.
There is a growing body of empirical data that indicate that the EF in the bulk is quite
different from that at the surface (Nowotny, 1991). This difference is due to segregationinduced concentration gradients at interfaces (surfaces and grain boundaries) and the
resultant electrical potential gradients (Adamczyk and Nowotny, 1986). These gradients
have a significant impact on the functional properties of solids, including photo-sensitivity.
Therefore quantitative analysis of surface versus bulk properties is a critical issue in the
interpretation of photo-induced effects. Accordingly, the imposition of specific surface
versus bulk properties during the processing of photo-sensitive materials is required for
the achievement of desired photo-sensitivity.
Centre for Materials Research in Energy Conversion, UNSW
ENERGY
Ch-06
PHOTON
PENETRATION
THICKNESS
(EF)n
EC
EF
(EF)p
EV
FIGURE 6.7 Effect of light in splitting of the Fermi energy level into two quasi-levels corresponding to
electrons and electron holes for an n-type semiconductor
Ch-06
2/6/07
7:06 PM
Page 185
185
(11)
(12)
Therefore:
(13)
(14)
where EF is the change of the Fermi energy induced by light and no and po denote the
concentration of the electronic charge carriers before the light is applied.
The effect of light on (EF)n in an n-type semiconductor such as reduced TiO2 is
relatively small because the component no is relatively large compared to n. However,
the effect of light on (EF)p in the n-type material is substantial because the component p
is large compared to po, which, in the dark, corresponds to the minority charge carriers.
Consequently, application of the light results in a substantial increase of the oxidation
potential represented by (EF)p leading to the shift of the equilibria represented by
Equations 24 to the right. Then the rate of these reactions is determined by the light
intensity and the resulting (EF)p level.
Ch-06
2/6/07
7:06 PM
186
Page 186
electrical conductivity;
thermoelectric power (Seebeck coefficient);
electrical conductivity and thermoelectric power measured simultaneously; and
work function.
The following sections discuss the relationships between these three electrical
properties, which can be determined experimentally, and the quantities describing the
electronic structure, Eg and EF.
(18)
where e is the electron charge, is the mobility, n, p and [i] are the concentrations of
electrons, electron holes and ions respectively, z is the charge of ionic species, and l is
their number. Therefore:
(19)
The temperature dependence of the electrical conductivity can be used to determine the
electronic structure of a solid (Kofstad, 1972). Becker and Frederikse (1962) showed that
there is an explicit relationship between the Eg of a semiconductor that exhibits an np
transition, such as titania, and the temperature dependence of the minimal value of the
at the transition:
(20)
Ch-06
2/6/07
7:06 PM
Page 187
187
where min is at the np transition point, which then assumes a minimum. Solids,
particularly nonstoichiometric compounds, can be well defined when they are in a state
of gas/solid equilibrium. Consequently, well-defined data for the should be determined
at elevated temperatures and under controlled gas phase compositions. In the case of
oxide materials, oxygen is the most important component of the gas phase under which
the material is annealed or equilibrated since oxygen is a component of the lattice. Thus
its activity in the gas phase results in the imposition of its activity in the lattice and the
related defect disorder. The determination of the electrical conductivity of oxide materials
has been described elsewhere (Nowotny, 1997).
As seen in Equations 1619, the determination of the concentration of charge carriers
from the electrical conductivity data is possible when the mobility terms are known (or
vice versa). The evaluation of the mobility terms corresponding to electronic charge
carriers requires the application of more sophisticated experimental and theoretical
approaches (Kofstad, 1972). The electrical conductivity measurements may also be
applied for monitoring the gas/solid equilibration kinetics for a metal oxide/oxygen system
(Kofstad, 1972).
While the electrical conductivity is the most commonly measured electrical property,
its physical meaning is complicated because it incorporates both concentration and
mobility terms (Kofstad, 1972). The commonly assumed interpretation of the electrical
properties of materials is based on the assumption that changes in the electrical
conductivity are determined by the concentration term alone, while the mobility term
remains independent of concentration. Unfortunately, this assumption frequently is not
valid, which is the reason why studies based solely on electrical conductivity may lead to
a misleading picture of semiconducting properties.
2/6/07
188
7:06 PM
Page 188
(23)
where Sn and Sp denote the Seebeck coefficient components corresponding to electrons
and electron holes respectively, and An and Ap are kinetics constants related to the
scattering of electrons and electron holes respectively.
Experimental determination of Sn and Sp is straightforward when one type of charge
carrier is predominant. The separation of S into the two components Sn and Sp within the
np transition regime is more complicated. Methods for the determination of the
components of S in terms of different charge carriers are described elsewhere (Wagner,
1972).
Combination of Equations 9 and 10 with Equations 22 and 23 may be used to establish
a direct interrelationship between S, which is a measurable quantity, and the
concentration of electronic charge carriers, which are essential to the assessment of the
charge transfer.
Experimental determination of thermoelectric power at elevated temperatures and
under controlled gas phase compositions has been described previously (Nowotny, 1997).
min
eB
k ln10
lg 2
k ln
4e B - 1
2B -
e
k
2B
Sp
THERMOPOWER, S
Ch-06
Sn
log
Source: Jonker, 1968; Nowotny, 1997
FIGURE 6.8 The schematic Jonker plot of thermoelectric power vs log showing the critical parameters
(meanings of the symbols are explained in text)
Ch-06
2/6/07
7:06 PM
Page 189
189
(24)
where:
(25)
and
(26)
where n and p are the mobilities of the electrons and electron holes respectively. The
parameters B and D, which can be determined from the Jonkers formalism, are directly
related to the band gap and the ratio of the mobility terms respectively.
Since the Jonker analysis allows determination of the mobilities, then Equation 19 may
be used for the determination of the concentration of charge carriers from the electrical
conductivity data. Also Equation 25 allows direct determination of the Eg when the
kinetics terms An and Ap are known.
A specific feature of the analysis described by Equations 2426 is that knowledge of
the p(O2) under which the oxide specimen is equilibrated is not required. However, it is
critical that both and S are determined in equilibrium under identical p(O2).
The simultaneous determination of both the and S of oxide semiconductors is
possible using a hightemperature Seebeck probe, which has been described previously
(Nowotny, 1991). A prototype of this instrument is available in the Centre for Materials
Research in Energy Conversion, University of New South Wales (UNSW).
Ch-06
2/6/07
7:06 PM
190
Page 190
widely applied method for the determination of work function changes of compounds at
elevated temperatures and under controlled gas phase environment is based on the
determination of the contact potential difference (CPD) (Nowotny, 1997):
(29)
where R is the work function of the reference material. Pt has been used for the
reference material in measurements of work function changes as a function of p(O2).
Therefore knowledge of the reference energy level allows determination of the EF at the
surface:
(30)
The changes in oxide materials at elevated temperatures (up to 1300K) in gas phases of
controlled oxygen activities using a high-temperature Kelvin probe have been described
previously (Nowotny, 1991). A prototype of this instrument is available in the Centre for
Materials Research in Energy Conversion, UNSW. This unique surface-sensitive tool may
be used for:
determination of the chemical potential of electrons; and
in situ monitoring of the changes in the chemical potential of electrons during the
with which the Ti ions (Ti3+ and Ti4+) alter their valence (Kofstad, 1972; Matzke,
1981).
TiO2-x has excellent resistance to chemical and photochemical corrosion in aggressive
aqueous environments (Bak et al, 2002).
The properties of TiO2-x can be altered by varying the defect chemistry and associated
electronic structure through the introduction of aliovalent ions and the alteration of
the concentration of intrinsic defects.
TiO2-x is substantially less expensive than other photo-sensitive materials.
Ch-06
2/6/07
7:06 PM
Page 191
191
6.6.1 SENSITIZATION
6.6.1.1 Dye sensitization
The function of dye sensitizers is to increase the absorption of solar energy through their
application as coatings to the external surface of the TiO2-x electrode (El Zayat et al, 1998;
Memming, 1980; ORegan and Gratzel, 1991; Bach et al, 1998). However, they have the
significant disadvantage that they have limited resistance to corrosion and photocorrosion in water compared to that of TiO2-x (Bak et al, 2002). The second problem is that
only the dye molecules in close proximity to the interface, corresponding to approximately
one monolayer, are efficient.2 This type of sensitizing has been already applied
commercially (AGO, 1999).
2/6/07
Page 192
NUMBER OF PHOTONS [s m eV ]
192
7:06 PM
21
410
21
310
21
210
J (RBGT)
J (TiO2)
21
110
1.0
0.5
Centre for Materials Research in Energy Conversion, UNSW
5.0 3.0
1.23 eV
1.5
-2
2
1.0
1 - Ir (TiO2)
2 - Ir (RBGT)
1
0.5
0.0
VISIBLE
Ch-06
0.5
1.0
1.5
2.0
2.5
3.0
WAVELENGTH [m]
FIGURE 6.10 Solar energy spectrum in terms of the radiation energy vs wavelength where Ir(TiO2) and
Ir(RBGT) denote the incidence of solar irradiance for commercial TiO2 and reduced-band-gap TiO2 respectively
Ch-06
2/6/07
7:06 PM
Page 193
193
disorder;
understanding of the relationship between defect disorder and electronic
Ch-06
2/6/07
7:06 PM
194
Page 194
that retard the charge transport across this structure (Nowotny, 1991); and
insignificant overlap of the electron wave functions of the dopants in rutile
(Tang, 1994).
The research and development programme on solar-hydrogen at the University of New
South Wales aims to engineer the electronic structure of TiO2-x and that of other Ti-based
compounds through the imposition of controlled defect chemistry. An important part of
this research is the derivation of defect diagrams and the establishment of the
relationships between defect disorder, electronic structure and the semiconducting
properties of undoped TiO2 as well as TiO2 doped with aliovalent ions. Preliminary
progress in this research has been reported recently by the authors (Nowotny et al, 1997
and 2005; Bak et al, 2002, 2003a and 2003b).
The present work considers Ti-based compounds, mainly TiO2x and its solid solutions,
which are considered to be primary candidates for PEs. The following sections overview
several key properties of TiO2x that are essential for its performance as a photo-electrode:
nonstoichiometry and related defect chemistry;
charge transfer; and
segregation.
6.8.2 NONSTOICHIOMETRY
Titania is known to be an oxygen-deficient material with its nonstoichiometry being
determined by the apparent deficit of oxygen (Kofstad, 1972). Accordingly, the correct
formula of titanium dioxide is TiO2-x, where x is determined by the concentration of ionic
lattice defects. However, the present state of understanding is such that the specific value
of x is not limited solely to oxygen vacancies; it should be considered in terms of the
vacancies in both the titanium and oxygen sublattices as well as titanium interstitials.
Taking into account all these defects and also assuming the presence of donor- and
acceptor-type foreign ions (impurities and dopants) the formula of TiO2-x may be
expressed in Krger-Vink notation (Kroger, 1974) as follows:
(31)
Ch-06
2/6/07
7:06 PM
Page 195
195
where ATi and DTi denote singly ionized acceptor- and donor-type ions. The law of
conservation of the number of lattice sites requires that:
(32)
The lattice charge neutrality condition requires that:
(33)
where [e] and [h] denote the concentration of electrons and electron holes respectively.
Equation 33 represents the general charge neutrality condition. This condition assumes a
simple form when only the predominant defects are involved (Kofstad, 1972).
Ch-06
2/6/07
196
7:06 PM
Page 196
sublattices (Yahia, 1963; Baumard and Tani, 1977; Balachandran and Eror, 1988; Son and Yu,
1996; Marucco et al, 1981; Blumenthal et al, 1967; Sawatari et al, 1982) that have been
taken into account. These disagreements are the reason why the authors of the
present work have undertaken extensive studies in the determination of the defect disorder
of titania.
(34)
(35)
(36)
(37)
(38)
(39)
It has recently been shown that prolonged oxidation results in oxygen incorporation into
the TiO2 leading to the formation of Ti vacancies (Nowotny et al, 2005):
(40)
The equilibrium constants for the preceding six reactions may be expressed according to
the following equations respectively:
(41)
(42)
(43)
(44)
(45)
(46)
where the square brackets denote the concentrations of defects, expressed in atomic
ratio. The preceding equilibrium constants can be related to the standard state
thermodynamic quantities entropy So and enthalpy Ho:
Ch-06
2/6/07
7:06 PM
Page 197
197
(47)
where both So and Ho are the specific materials properties. The thermodynamic
quantities related to the equilibrium constants have been reported elsewhere (Kofstad,
1972; Bak et al, 2003a).
A strongly reduced regime is characterized by the p(O2) exponent, 1/m = 1/4. This
exponent may be explained by the defect disorder based on Ti interstitials as the
predominant defects. Then simplified charge neutrality assumes the form:
(48)
The combination of Equations 42 and 48 leads to the following relation:
p(o2)
(49)
A reduced regime corresponds to the p(O2) exponent, 1/m = 1/4. This exponent is
consistent with the defect disorder based on doubly ionied oxygen vacancies as the
predominant defects. The simplified charge neutrality is:
(50)
The combination of Equations 41 and 50 results in:
p(o2)
(51)
An oxidized regime involves three sub-regimes including n-type, p-type and mixed
np sub-regimes in which the p(O2) exponent assumes the following respective values:
1/m = 1/4 (in the n-type regime);
1/4 <1/m <1/4 (in the np transition regime); and
1/m = 1/4 (in the p-type regime).
Ch-06
2/6/07
198
7:06 PM
Page 198
(52)
where A denotes an effective concentration of acceptors, which includes the
concentration of extrinsic acceptor-type impurities, [A], donor-type defects, [D], as well
as Ti vacancies:
(53)
where A and D denote the singly ionized acceptors and donors respectively. While the
concentration of acceptor-type impurities is statistically higher than that of donors, it will
be shown below that the major component of the parameter A are Ti vacancies. It will also
be shown that the component A should be considered as quenched under usual
experimental conditions. Combining Equations 41 and 52 results in:
p(o2)
(54)
It has been recently shown that TiO2 may also exhibit p-type properties. Assuming that the
predominant defects in this regime are acceptor-type defects and oxygen vacancies, the
concentration of electronic defects is the following function of the p(O2):
p(o2)
(55)
2/6/07
7:06 PM
Page 199
199
TiO2x
1273 K, A = 0
2
VO
4+
Ti i
e
3+
Ti i
log [ ]
Ch-06
10
12
14
25
20
15
10
FIGURE 6.11 Defect diagram showing the concentration of defects vs oxygen partial pressure for undoped
TiO2 at 1273K in absence of foreign elements forming donors or/and acceptors
defects are electrons within the entire range of p(O2). In the case where the mobility terms
are independent of p(O2), the slope of the log versus log p(O2) plot (where is electrical
conductivity) is the same as the slope of the log n versus log p(O2) plot. Accordingly,
reduction of undoped TiO2-x, through the reduction of p(O2) during equilibration, results in
an increase of both the concentration of electrons and the electrical conductivity. As seen
in Figure 6.11, the n-type conductivity of undoped TiO2-x at high and low p(O2) is
determined by oxygen vacancies and tri-valent Ti interstitials respectively. Similar defect
diagrams may be derived for acceptor- and donor-doped TiO2-x (Bak et al, 2003a and 2003c).
One should emphasize that in the temperature range usually applied for the
characterization of electrical properties (10001500K) the concentration of titanium
Ch-06
2/6/07
200
7:06 PM
Page 200
vacancies does not assume the equilibrium concentration because the mobility of these
defects is too low (Baumard and Tani, 1977; Bak et al, 2003a and 2003c). This is also
the case for BaTiO3 (Nowotny and Rekas, 1994). In other words, titanium vacancies attain
equilibrium concentrations at substantially higher temperatures, typically above 2000K
(Bak et al, 2003c). At lower temperatures, the concentrations of titanium vacancies cannot
reach an equilibrium state, so their concentrations may vary depending on specific
processing conditions. This is the main reason why there is a substantial discrepancy
between the reported data for electrical properties and the related defect diagrams.
According to the derived defect diagrams, undoped TiO2-x exhibits n-type properties within
the entire range of p(O2) and does not exhibit an np transition (Bak et al, 2003). This is the
case when titanium vacancies are absent or present at very low concentrations. The fact that
the majority of the reported electrical conductivity data reveal minima indicates that titanium
vacancies are formed during either a) processing of the specimen or b) the experiment if
performed over a prolonged period of time (Bak et al, 2003c; Nowotny et al, 2005).
The defect diagrams for undoped TiO2-x may be considered to be divisible into two
p(O2) regimes, where the predominant ionic defects are:
1 tri-valent titanium interstitials under extremely reducing conditions; or
2 oxygen vacancies at intermediate and high p(O2) values.
Presence of aliovalent ions has a substantial impact on the electrical properties of TiO2-x.
Therefore the characterization of TiO2-x specimens should include the determination of the
concentrations of impurities that act as donors and acceptors. The defect diagrams reported
in this study (Bak et al, 2003) provide the first comprehensive representation of the defect
chemistry of TiO2-x containing both intrinsic and foreign defects, the latter including both
acceptor- and donor-type elements in TiO2-x, for a wide range of processing conditions.
Ch-06
2/6/07
7:06 PM
Page 201
201
EC
Eact
Eact
EC
EF
EF
EV
EV
a)
b)
FIGURE 6.12 Effect of Fermi energy on the charge transfer between the surface of semiconducting solid
and adsorbed species forming either acceptors or donors
When the collective electronic properties control the reactivity, the charge transfer is
determined by the difference between the Fermi energy and the electron affinity or
ionization potential of the adsorbed species, as shown schematically in Figure 6.12. As
can be seen, the adsorbed molecules may be either donors or acceptors, depending on
the Fermi energy of the solid and the electronic structure of the adsorbed species. In this
case one should expect a linear dependence between the activation energy of the charge
transfer and the Fermi energy of the solid (Nowotny, 1991).
The preceding model does not apply when the local electrostatic interactions between
the adsorbed species and surface-active sites control the reactivity. These local interactions,
which are determined by the electronic structure of the specific surface site, are independent
of the Fermi energy which may be considered as a collective factor of the lattice.
A study by Kowalski et al (1980) indicates that the reduction of the TiO2-x surface
results in the formation of surface states that enhance the dissociation of water
molecules.
Hepel et al (1982) reported that photo-etching of TiO2 in acid solutions results in
dissolution of surface donor states, such as tri-valent Ti ions, and increase in the surface
area of the electrode. In consequence, the acid etching is expected to enhance the ECE
of the TiO2-based photo-electrode.
Ch-06
2/6/07
202
7:06 PM
Page 202
TABLE 6.1 Effect of Cr on the energy conversion efficiency (ECE) of a photo-electrochemical cell involving
photo-anode made of Cr-doped TiO2
EFFECT ON SPECIFIC PROPERTY
Reduction of band gap
Increase of stability of chemisorbed
oxygen species
Reduction of the electron-hole pair lifetime
Formation of segregation-induced retarding
potential barrier
EFFECT ON ECE
Increase
Decrease
REFERENCE
Wilke and Brauer, 1999
Sharma et al, 1997
Decrease
Decrease
Decrease
Decrease
Decrease
Decrease
Decrease
Ch-06
2/6/07
7:06 PM
Page 203
203
ENERGY
ATOM
MOLECULE
(2 ATOMS)
SMALL
CLUSTER
(10 ATOMS)
CLUSTER
PARTICLE
(~200 ATOMS)
SEMICONDUCTOR
(>200 ATOMS)
VACUUM
CONDUCTION
BAND
Eg
Eg
VALENCE
BAND
FIGURE 6.13 Effect of cluster and grain size on electronic structure and related band gap width according
to Hoffmann et al (1995)
to have an effect on the electronic structure and the band gap as well. This effect has
been considered by Hoffmann et al (1995) in terms of a quantum size effect, where finegrained materials exhibit larger Eg values than those of coarse-grained materials, as
shown in Figure 6.13. Therefore it is to be expected that a reduction in the grain size of
TiO2-x will result in a decrease in its photo-sensitivity.
and
titanium vacancies are the surface active sites for water splitting.
Ch-06
2/6/07
7:06 PM
204
Page 204
et al, 1999). The in status nascendi atomic oxygen formed on the surface of TiO2-x is
extremely reactive. Therefore even traces of TiO2-x present in water result in the rapid
oxidation of toxic and pollutant compounds.
Decontamination of water containing viruses and bacteria (Fujishima et al, 1999).
The in status nascendi atomic oxygen also is known to oxidize these species,
resulting in the sterilization of water.
Generation of electricity using so-called electric windows (Bach et al, 1998;
Nakato et al, 1982). Dye-sensitized TiO2-x, deposited on the surface of glass windows,
forms the PE component of a solar cell to produce electricity (AGO, 1999).
Inhibition of water condensation on ceramic tiles, glass and other building
materials (Fujishima et al, 1999). It has been observed that thin films of TiO2-x exhibit
self-cleaning due to their strongly oxidizing properties.
Anti-tumour effect on gastric cancers (Fujishima et al, 1999).
Anti-fogging coatings on glasses due to superhydrophilicity (Fujishima et al,
1999).
Coatings for enhanced cleanability of car bodies and monuments due to
hydrophilicity (Fujishima et al, 1999).
It is clear that there is a wide range of spin-off applications for TiO2-x with reduced band
gap and increased photo-sensitivity. Ultimately, it is likely that it has the potential to
replace silicon as a photovoltaic material in the future (Bak et al, 2003).
Ch-06
2/6/07
7:06 PM
Page 205
205
h
1
e
(1)
n-Si
p-Si
O2
H2
4
e
TiO2-x
e
(2)
H2 O + M X
(3)
e
H2O + 2h 2H+ + O2
2H+ + 2e H2
V
FIGURE 6.14 Circuit of the hybrid cell of Morisaki et al (1976) and related electrochemical chain
Ch-06
2/6/07
7:06 PM
206
Page 206
LASER
GENERATOR
SUN
SPACE-BASED
EQUIPMENT
SOLAR
CONCENTRATOR
EARTH-BASED
EQUIPMENT
O2
H2
H
TiO2
H2 2H + 2e
TiO2-BASED
ELECTROCHEMICAL CELL
2H + 2e + O2 H2O
HYDROGEN
FUEL CELL
FIGURE 6.15 Schematic illustration of the solar-hydrogen generation system using space solar energy
collector
cycle; and
solar energy in space is available independent of weather conditions.
It has been claimed that the cost of hydrogen manufactured using this technology will
be ~20 Japanese Yen for an amount of hydrogen equivalent to one litre of gasoline. It is
expected that the first experimental satellite will be launched by 2010 and the entire
system will be completed by 2020. A vision of this technology, using the solar energy
collector outside the atmosphere, was described by Bockris (1999).
Ch-06
2/6/07
7:06 PM
Page 207
207
O2
H2
H
+
H2 2H + 2e
TiO2
TiO2-BASED
ELECTROCHEMICAL CELL
2H + 2e + O2 H2O
HYDROGEN
FUEL CELL
FIGURE 6.16 Schematic illustration of the use of both solar-hydrogen and solar-oxygen as input gases for
the production of electricity using a fuel cell
6.13 SOLAR-OXYGEN
Although the principal objective of the photo-electrochemical cells technology is the
production of hydrogen, the by-product of the photolysis of water is oxygen, whose wide
range of applications includes medical applications and metallurgy. Such solar-oxygen
also has the potential to be used in high-performance fuel cells instead of air. This
application would be advantageous because the combustion of any fuel, even hydrogen,
has the potential to result in NOx pollution owing to the fact that combustion usually takes
place in the presence of air, which contains ~80 per cent nitrogen. The NOx pollution at
the oxidation electrode could, therefore, be eliminated if the oxidation electrode is
exposed to solar-oxygen as the oxidizing agent instead of air as shown in Figure 6.16.
1998).
Solar-hydrogen is the leading candidate for a renewable and clean energy carrier of
the future.
Hydrogen generated from water using solar energy constitutes a renewable form of
energy.
Hydrogen generated from water using solar energy constitutes a clean energy carrier
as neither its production nor its combustion produces greenhouse or pollutant gases.
Solar-hydrogen ultimately will reduce our total reliance on coal, gasoline and natural
Ch-06
2/6/07
7:06 PM
208
Page 208
6.14.2 BENEFITS
Many areas are subject to abundant sunlight and so are well placed to commercialize
solar-hydrogen technologies. For example, the northern parts of Australia have some
of the highest levels of sunlight in the world. When this energy is rationally utilized
for the production of fuel Australia could become an exporter of solar energy: a
singular concept.
Solar-hydrogen will allow reduction in emissions of greenhouse, pollutant and toxic
gases.
Solar-hydrogen encompasses both the production and utilization of a fuel that is
100 per cent environmentally clean.
Countries that possess the infrastructure for solar-hydrogen will be able to reduce
their reliance on fossil fuels.
Countries that produce excess solar-hydrogen can export solar energy, which is a
singular concept.
The technologies created for solar-hydrogen production will have far-reaching applications
in both domestic and industrial markets. For example, a solar-hydrogen panel of
approximately 40 40km, located in the Pilbara region of Western Australia, will be
sufficient to meet all of Australias current energy requirements.
6.14.3 ADVANTAGES
The fuel may be generated anywhere.
The process requires only water and solar energy;
The hydrogen-generating device does not have any moving parts, so maintenance on
Ch-06
2/6/07
7:06 PM
Page 209
209
Assuming that there are 300 sunny days a year, under the preceding conditions, the
annual production of solar-hydrogen would be 1470kg. At the present hydrogen price of
US$1.3/kg, the value of this annual hydrogen production would be US$1911. Taking into
account the impact of pollution when using the hydrogen generated from natural gas and
the associated costs (see section 6.14.4), and the lack of pollution when using solarhydrogen, the real value of the hydrogen generated using solar energy should be elevated
to approximately US$8800 p.a.
At the above assumptions it is calculated (using 2001 data) that a 40 x 40km panel
would produce sufficient hydrogen to meet all of the energy needs of Australia. The above
surface area is equivalent to the surface of the individual household units if all homes in
Australia are equipped with such units.
6.16 SOLAR-METHANOL
Several Australian industrial organizations, including DUT, CC Energy Pty. Ltd and Isentropic
Systems Ltd., have proposed the integration of the UNSW solar-hydrogen technology with
CC Energys proposed coal-solar system for the production of zero-CO2-emission methanol
(Cummings, 2003). This initiative aims at the use of coal and solar-hydrogen for the
production of solar-methanol, with the surplus of CO2 entering into reaction with solarhydrogen, thus forming another production line of methanol. Further, the oxygen produced
along with the solar-hydrogen can be used for coal gasification, as shown in Figure 6.17.
Ch-06
2/6/07
7:06 PM
210
Page 210
COAL
GASIFIER
2 C + O2 2 CO
H2O
CO + H2O CO2 + H2
REACTOR-1
2 H2 + CO CH3OH
O2
CO2
WATER PHOTOLYSIS
H2O H2 + O2
H2O
REACTOR-2
CO2 + 3 H2 H2O + CH3OH
METHANOL
FIGURE 6.17 The concept of the Australian initiative of the solar-methanol technology
The system, in effect, sequesters surplus CO2, generated as a result of coal gasification,
and uses it to produce a low-emission fuel in the form of solar-methanol. Methanol has the
advantage that it is substantially easier to transport than hydrogen. Although the
combustion of methanol still results in the emission of greenhouse gases, the fuels
principal component is the solar energy. Therefore, this fuel should be considered to be a
substantial step forward, compared to fossil fuels, in the reduction of global warming.
According to Cummings (2003), solar-methanol is expected to be the most feasible
carrier of solar-hydrogen to be used for hydrogen transportation over large distances.
6.17 CONCLUSIONS
There has been growing awareness that the so-called hydrogen economy will soon be
part of our life. Introduction of this technology means that hydrogen will replace fossil
fuels. The move towards the hydrogen economy is rapid and it is being driven by a general
perception that hydrogen is an efficient and clean fuel. The perception of the latter must
be viewed with caution. Specifically, it must be appreciated that only hydrogen generated
using renewable energy leads to zero emissions of greenhouse gases and thus is
environmentally clean.
It is expected that the commencement of the hydrogen era will unfold a market
dominated by hydrogen generated from natural gas for economic reasons. However, the
Ch-06
2/6/07
7:06 PM
Page 211
211
issue of global warming will not go away and its increasingly damaging effects will
impose the need to generate hydrogen from renewable energy. In the present work, it has
been argued that generation from water decomposition using solar energy is one of the
most promising means to produce hydrogen.
The principle of hydrogen generation through water photolysis using TiO2-x as the
photo-electrode has been overviewed. The present work brought together the concepts
of photo-electrochemistry and the concepts of defect chemistry using TiO2 as an
example of photo-electrode. It was shown that photo-sensitivity of oxide materials is
closely related to their defect disorder. Therefore defect chemistry may be used as
a framework for engineering novel photo-sensitive materials. The defect chemistry of
TiO2-x and its impact on the properties that are critical for its performance as a
photo-electrode, including the chemical potential of electrons and the band gap, have
been emphasized.
It may be expected that, during the initial stage of commercialization, the price of
solar-hydrogen will be higher than that of hydrogen from methane. However, this cost
disadvantage will be compensated by reduction of pollution and global warming.
According to Bockris (2002), the price of pollution is so high that we cannot afford it. Solarhydrogen may be an ultimate remedy for pollution and global warming. This is the reason
why the discovery of solar-hydrogen by Fujishima and Honda (1972) may be considered
as one of the major discoveries of the 20th century.
NOTES
1 While in theory the combustion of hydrogen in oxygen results solely in the formation of
water, in practice an internal combustion engine uses air rather than oxygen. Since air
contains 80 per cent nitrogen then it is likely that the combustion in air may result in the
formation of a certain amount of NOx, irrespectively of the fuel used. However, this kind
of NOx pollution does not apply to the combustion of hydrogen in low temperature
hydrogen fuel cells.
2 This comment was made by the reviewer of the present paper.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge financial support from The University of New South
Wales, the Australian Research Council, Rio Tinto Ltd and Sialon Ceramics Pty. Ltd. The
authors also wish to acknowledge the extensive comments on hydrogen economy-related
data kindly provided by Professor John OM. Bockris.
REFERENCES
Adamczyk, Z. and Nowotny, J. (1986) Effect of segregation on near-surface and bulk transport phenomena in ionic crystals,
J. Phys. Chem. Solids, no 47, pp1124
Ch-06
2/6/07
7:06 PM
212
Page 212
AGO (1999) Australian Greenhouse Office, Renewable Energy Commercialisation Program, 9 July 1999,
www.greenhouse.gov.au/renewable/renew3.html
Armstrong, P. (1996) Hydrogen power science facts or science fiction, Borderland Science Research Foundation,
www.borderlands.com/journal/h2.htm
Bach, U., Lupe, D and Gratzel, M. (1998) Solid-state dye-sensitized mesoporas TiO2 solar cells with high photon-to-electron
conversion efficiencies, Nature, no 395, pp583585
Bak, T., Nowotny, J., Rekas, M. and Sorrell, C. C. (2002) Photo-electrochemical hydrogen generation from water using solar
energy. Materials-related aspects, Int. J. Hydrogen. Energy, no 27, pp9911022
Bak, T., Nowotny, J., Rekas, M. and Sorrell, C. C. (2003) Defect chemistry and semiconducting properties of titanium dioxide.
I. Intrisic electronic equilibrium, J. Phys. Chem. Solids, no 64, pp10431056
Bak, T., Nowotny, J., Rekas, M. and Sorrell, C. C. (2003a) Defect chemistry and semiconducting properties of titanium
dioxide. II. Defect diagrams, J. Phys. Chem. Solids, no 64, pp10571067
Bak, T., Nowotny, J., Rekas, M. and Sorrell, C. C. (2003b) Defect chemistry and semiconducting properties of titanium
dioxide. III. Mobility of electronic charge carriers, J. Phys. Chem. Solids, no 64, pp10691087
Bak, T., Burg, T., Kang, S-J. L., Nowotny, J., Rekas, M., Sorrell, C. C., Vance, E. R., Yoshida, Y. and Yamaguchi, K. (2003c)
Charge transfer in polycrystalline titanium dioxide, J. Phys. Chem. Solids, no 64, pp10891095
Bak, T., Nowotny, J., Rekas, M. and Sorrell, C. C. (2003d) Application of the high temperature Fermi probe in studies of the
charge transfer at the gas/solid interface of electrochemical systems. Example of the oxygen/zirconia system, in
K. Uematsu (ed) Ceramic Interfaces, Trans Tech Publications, Zurich, Switzerland, pp151178
Balachandran, U. and Eror, N. G. (1988) Electrical conductivity in non-stoichiometric titanium dioxide at elevated
temperatures, J. Mater. Sci., no 23, pp26762682
Baumard, J. F. and Tani, E. (1977) Electrical conductivity and charge compensation in Nb doped rutile, J. Chem. Phys., no 67,
pp857860
Becker, J. H. and Frederikse, H. P. R. (1962) Electrical properties of nano-stoichiometric semiconductors, J. Appl. Phys.,
no 33, pp447453
Blumenthal, R. N., Kirk, Jr. J. C. and Hirthe, W. M. (1967) Electronic mobility in rutile (TiO2) at high temperatures, J. Phys.
Chem. Solids, no 28, pp10771079
Bockris, J. OM. (1980) Energy Options. Real Economics and the Solar-Hydrogen System, Australia & New Zealand Book
Company, Sydney, Australia, pp325364
Bockris, J. OM. (1999) Hydrogen economy in the future, Int. J. Hydrogen Energy, no 24, pp115
Bockris, J. OM. (2002), private communication to J. Nowotny, 6 June
Bockris, J. OM. (2003) On hydrogen futures: Towards a sustainable energy system, Int. J. Hydrogen Energy, no 28,pp131133
Bockris, J. OM. and Uosaki, K. (1976) The theory of hydrogen production in a photoelectrochemical cell, Heliotechnique and
Development, no 1, pp435470
Bockris, J. OM. and Uosaki, K. (1976a) Photoproduction of hydrogen: Potential dependence of the quantum efficiency as a
function of wavelength, Energy, no 1, pp143145
Bockris J. OM., Uosaki, K. and Kita, H. (1981) Interfacial electron transfer as a significant step in photo-electrochemical
reactions on some semiconductors, J. Appl. Phys., no 52, pp808810
Bockris, J. OM., Veziroglu, T. N. and Smith, D. (1991) Solar Hydrogen Energy. The Power to Save the Earth, Optima Series,
The Guernsey Press Co Ltd., UK, pp7789
Chandra, S. (1985) Photoelectrochemical Solar Cells, Gordon & Breach, New York, pp1619 and 63203
Cummings, R. (2003) The solar-methanol production system, communication to J. Nowotny, 8 May
El Zayat, M. Y., Saed, A. O. and El-Dessouki, M. S. (1998) Photoelectrochemical properties of dye sensitized Zr-doped SrTiO3
electrodes, Int. J. Hydrogen Energy, no 23, pp259266
Ch-06
2/6/07
7:06 PM
Page 213
213
Fievet, F., Germi, P., Bergevin, F. and Figlarz, M. (1979) Lattice parameter, microstrains and nonstoichiometry in nickel (II) oxide.
Comparison between mosaic microcrystals and quasi-perfect single microcrystals, J. Appl. Crystal., no 12, pp387394
Friedli, H., Lotscher, H., Oescher, H. and Siegenthaler, U. (1986) Ice core record of the 13C/12C ratio of atmospheric CO2 in the
past two centuries, Nature, no 324, pp237239
Fujishima, A. and Honda, K. (1972) Electrochemical photolysis of water at a semiconductor electrode, Nature, no 238, pp3738
Fujishima, A., Kohayakawa, K. and Honda, K. (1975) Hydrogen production under sunlight with an electrochemical photocell,
J. Electrochem. Soc., no 122, pp14871489
Fujishima, A., Hashimoto, K. and Watanabe, T. (1999) TiO2 Photocatalysis. Fundamentals and Applications, BKC, Inc., Tokyo,
pp14176
Gerischer, H. (1972) Electrochemical techniques for the study of photosensitization, Photochemistry & Photobiology, no 16,
pp243260
Gerischer, H. (1977) The electrochemical cell, in A. Heller (ed) Semiconductor Liquid-Junction Solar Cells, The
electrochemical Society, Inc., Princeton, NJ, US, pp119
Gerischer, H. (1997) Principles of electrochemistry, in P. J. Gellings and H. J. M. Bouwmeester (eds) The CRC Handbook of
Solid State Electrochemistry, CRC Press, Boca Raton, FL, US, pp1072
Green, M. (1986) Solar Cells. Operating Principles, Technology and System Applications, University of New South Wales,
Sydney, Australia, pp1731
Gulino, A., Condorelli, G. G., Fragalh, I. and Egdell, R. G. (1995) Surface segregation of Sb in doped TiO2 rutile, Appl. Surf.
Sci., no 90, pp289295
Hartley Management Group Pty Ltd (2003) The hydrogen economy. Challenges and strategies for Australia, including the tidal
energy link, www.hartleymgt.com.au/hydrogenbroome
Hepel, T., Hepel, M. and Osteryoung, R. A. (1982) Thermodynamic and photo-electrochemical behavior of n-TiO2 electrode in
fluorite-containing solutions, J. Electrochem. Soc., vol 129, no 9, pp12311241
Hirano, S. I., Kang, S-J. L., Nowotny, J., Smar, R. St. C., Sorrell, C. C., Sugihara, S., Taniguchi, T., Yamawaki, M. and Yoo, H-I.
(1999) Development of novel materials for reduction of greenhouse gases and environmental monitoring through interface
engineering, Korean Journal of Materials Science, no 9, pp635653
Hoffman, M. R., Martin, S. T., Choi, W. and Bahneman, D. W. (1995) Environmental applications of semiconductor
photocatalysis, Chem. Rev., no 95, pp6996
Ikeda, J. A. and Chian, Y. M. (1991) Space charge segregation at grain boundaries in titanium dioxide: I. Relationship
between lattice defect chemistry and space charge potential, J. Am. Ceram. Soc., no 76, pp24372446
Jonker, G. H. (1968) The application of combined conductivity and Seebeck effect plots for the analysis of semiconductor
properties, Philips Res. Rep., no 23, pp131138
Khan, S. U. M., Al-Shahry, M. and Ingler, W. B. (2002) Efficient photochemical water splitting by a chemically modified
n-TiO2, Science, no 297, pp22432245
Kofstad, P. (1972) Nonstoichiometry, Electrical Conductivity and Diffusion of Binary Metal Oxides, Wiley, New York
Kowalski, J. M., Johnson, K. H. and Tuller, H. L. (1980) Models for the photo-electrocatalytic decomposition of water at
semiconducting oxide anodes, J. Electrochem. Soc., no 127, pp19691973
Kowalski, K., Ijjaali, M., Bak, T., Dupre, B., Gleitzer, Ch., Nowotny, J., Rekas, M. and Sorrell, C. C. (2001) Semiconducting
properties of CoO thin films, Intern. J. Ionics, nos 46, pp394399
Kroger, F. A. (1974) The Chemistry of Imperfect Crystals, vol 3, North Holland, Amsterdam, p275
Licht, S., Wang, B., Mukerji, S., Soga, T., Umeno, M. and Tributsch, H. (2000) Efficient solar water splitting, exemplified by
ReO2-catalyzed AlGaAs/Si photoelectrolysis, J. Chem. Phys. B, no 104, pp89208924
Lo, W. J., Chung, Y. W. and Samorjai, G. A. (1978) Electron spectroscopy studies of the chemisorption of O2, H2 and H2O on
the TiO2 (100) surface with varied stoichiometry: Evidence for the photogeneration of Ti+3 and its importance in
chemisorption, Surf. Sci., no 71, p199
Ch-06
2/6/07
7:06 PM
214
Page 214
Marucco, J. F., Gautron, J. and Lemasson, P. (1981) Thermodynamic properties of titanium dioxide, niobium dioxide and their
solid solutions at high temperature, J. Phys. Chem. Solids, no 46, pp709717
Matzke, Hj. (1981) Diffusion in nonstoichiometric oxides, in O. T. Sorensen (ed) Nonstoichiometric Oxides, Academic Press,
New York, pp155230
Mavroides, J. G., Tchernev, D. I., Kafalas, J. A. and Kolesar, D. E. (1975) Photoelectrolysis of water in cells with TiO2 anodes,
Mater. Res. Bull., no 10, pp10231030
Memming, R. (1980) Solar energy conversion by photoelectrochemical process, Electrochem. Acta, no 25, pp7788
Momirlan, M. and Veziroglu, T. N. (2002) Current status of hydrogen energy, Renewable and Sustainable Energy Reviewes,
no 6, pp141179
Morisaki, H., Watanabe, T., Iwase, M. and Yazawa, K. (1976) Photoelectrolysis of water with TiO2-covered Solar-cell
electrodes, Appl. Phys. Lett., no 28, pp338340
Nakato, Y., Shioji, M., Tsubomura, H. (1982) Photoeffects on the potentials of thin metal films on a n-TiO2 crystal wafer. The
mechanism of semiconductor photocatalysis, Chem. Phys. Lett., no 99, pp453456
Neftel, A., Moore, E., Oeschger, H. and Stauffer, B. (1985) Evidence from polar ice cores for the increase in atmospheric CO2
in the past two centuries, Nature, no 315, pp4547
Nowotny, J. (1991) Interface defect chemistry and its impact on properties of oxide ceramic materials, in J. Nowotny (ed)
Science of Ceramic Interfaces, Elsevier, Amsterdam, pp79204
Nowotny, J. (1997) Interface electrical phenomena in ionic solids, in P. J. Gellings and H. J. M. Bouwmeester (eds) The CRC
Handbook of Solid-State Electrochemistry, CRC Press, Boca Raton, FL, US, pp121159
Nowotny, J. and Rekas, M. (1994) Defect structure, electrical properties and transport in barium titanate, VII. Chemical
diffusion, Ceramics International, vol 20, no 4, pp265275
Nowotny, J., Radecka, M. and Rekas, M. (1997) Semiconducting properties of undoped TiO2 J. Phys. Chem. Solids, no 33,
pp927937
Nowotny, J., Radecka, M., Rekas, M., Sugihara, S., Vance, E. R. and Weppner, W. (1998) Electronic and ionic conductivity of
TiO2 single crystals within the n-p transition range, Ceram. Intern., no 24, pp571577
Nowotny, J., Sorrell, C. C., Sheppard, L. R. and Bak, T. (2005) Solar-hydrogen: Environmentally safe fuel for the future, Int.
J. Hydrogen Energy, vol 30, no 5, pp521544
Onishi, T., Nakato, I. and Tsobomura, H. (1975) The quantum yield of photolysis of water on TiO2 electrodes, Ber. Bunsen-Gess.
Phys. Chem., no 79, pp523525
Ooki, H. (1998) Global warming and issues for Japan in Global Environment Section (ed), The Environment Agency of Japan
Booklet #24, Tsuyosji Akiyama, Tokyo, Tokyo Shimbun Shuppankyoku (Newspaper Publishing), pp1312
ORegan, B. and Gratzel, M. (1991) A low-cost, high efficiency solar cell based on dye-sensitized colloidal TiO2 films, Nature,
no 353, pp737739
Prather, M. J. (2003) An environmental experiment with H2?, Science, no 302, pp581582
Rahman, M. M., Soga, T., Jimbo, T. and Umeno, M. (1996) Novel low-cost solid-state heterojunction solar cell based on TiO2
and its modification for improved efficiency, Jpn. J. Appl. Phys., no 35, pp33343342
Rahman, M. M., Krishna, K. M., Soga, T., Jimbo, T. and Umeno, M. (1999) Optical properties and X-ray photoelectron
spectroscopic study of pure and Pb-doped TiO2 thin films, J. Phys. Chem. Solids, no 60, pp201210
Rand, D. A. J. and Badwal, S. P. S. (2005) Australian Hydrogen Activity, Department of Industry, Tourism and Resources,
Australian Government, Canberra
Reisfeld, R., Eyal, M. and Brusilovsky, D. (1988) Luminescence enhancement of rhodamine 6a in sol-gel films containing silver
aggregates, Chem. Phys. Lett., no 153, pp210214
Ruiz, A. M., Sakai, G., Cornet, A., Shimanoe, K., Morante, J. R. and Yamazoe, N. (2003) Cr-doped TiO2 gas sensor for
exhaust NO2 monitoring, Sensors and Actuators B, no 93, pp509518
Ch-06
2/6/07
7:06 PM
Page 215
215
Ruiz, A. M., Dezanneau, G., Arbiol, J., Cornet, A. and Morante, J. R. (2004) Insight into the structural and chemical
modification of Nb additive on TiO2 nanopartcles, Chem. Mater., no 16, pp862871
Russ, M. (1994) Cost-Effective Strategies for an Optimised Allocation of Carbon Dioxide Emission Reduction Measures,
Umwelttechnik Series (Technique of the Environment), Verlag Shaker, Aachen, Germany, pp1180
Sawatari, H., Iguchi, E., Tiley, R. J. D. (1982) Formation energies of point defects in rutile (TiO2), J. Phys. Chem. Solids, no
43, pp11471155
Seraphin, B. O. (1979) Spectrally selective surfaces and their impact on photothermal solar energy conversion, in B. O.
Seraphin (ed) Solar Energy Conversion, Springer-Verlag, Berlin, pp556
Service, R. F. (2002) Catalyst boosts hopes for hydrogen bonanza, Science, no 297, pp21892190
Sharma, R. K., Bhaynagar, M. C. and Sharma, G. L. (1997) Mechanism of highly sensitive and fast response Cr-doped TiO2
Oxygen gas sensors, Sensors and Actuators B, no 45, pp209215
Sheppard, L. R. (2004) Defect disorder and semiconducting properties of Nn-doped titanium dioxide, Ph.D. thesis (in
progress), School of Materials Science and Engineering The University of New South Wales, Sydney, Australia
Shimbun, Y. (2001) Satellite system would generate clean fuel, The Daily Yomiuri, Tokyo, 18 August
Soliman, A. A. and Seguin, H. J. J. (1981) Reactively sputtered TiO2 electrodes from metallic targets for water electrolysis
using solar energy, Solar Energy Mater., no 5, pp95102
Son, J. and Yu, I. (1996) A study on the defect structure of TiO2 (rutile) by electrical conductivity Measurements, Korean J.
Ceram., no 2, pp131136
Tang, H., Prasad, K., Sanjines, R., Schmidt, P. E. and Levy, F. (1994) Electrical and optical properties of TiO2 anatase thin
films, J. Appl. Phys., vol 75, no 4, pp20422047
Thevuthasan, S., Shivaparan, N. R., Smith, R. J., Gao, Y. and Chambers, S. A. (1997) Rutherford backscattering and
chanelling studies of a TiO2, (100) substrate, epitaxially grown pure and Nb-doped TiO2 films, Appl. Surf. Sci., no 115,
pp381385
Thomas, C. E., James, B. D. Jr and Lomax, F. D. (1998) Market penetration scenarios for fuel cell vehicles, Int. J. Hydrogen
Energy, no 23, pp949966
Tromp, T. K., Shia, R-L., Allen, M., Eiler, J. M. and Young, Y. L. (2003) Potential environmental impact of a hydrogen economy
on the stratosphere, Science, no 300, pp17401741
United Nations Environmental Programme (2002) in Full List of Vital Climate Graphics, GRID, Arendal, Norway,
www.grida.no/db/maps/collection/climate6/austral.htm
Veziroglu, T. N. (1998) Dawn of the hydrogen age, Int. J. Hydrogen Energy, no 23, pp10771978
Veziroglu, T. N. (2000) Quarter century of hydrogen movement 19742000, Int. J. Hydrogen Energy, no 25,
pp11431150
Wagner, C. (1972) The thermoelectric power of cells with ionic compounds involving ionic and electronic conduction, Progr.
Solid State Chem., no 7, pp137
Wang, Q., Lian, G. and Dickey, E. C. (2004) Grain boundary segregation in yttrium-doped polycrystalline TiO2, Acta Materialia,
no 52, pp809820
Wilke, K. and Brauer, H. D. (1999) The influence of transition metal doping on the physical and photocatalytic properties,
J. Photochem. and Photobiol. A: Chemistry, no 121, pp4953
Yahia, J. (1963) Dependence of the electrical conductivity and thermoelectric power of pure and aluminium-doped rutile on
equilibrium oxygen pressure and temperature, Phys. Rev., no 130, pp17111719
Zhao, G., Kozuka, H. and Yoko, T. (1996) Sol-gel preparation and photo-electrochemical properties of TiO2 films containing Au
and Ag metal paricles, Thin Solid Films, no 277, pp147154
ch-07
2/7/07
2:58 PM
Page 216
Abstract
The idea of using solar heat in industry has been extensively discussed since the 1980s,
and several pilot plants have been built. In recent years the costs have been
substantially reduced and the technology much improved with highly efficient solar
collectors and improved system technology (controls, pumps and so on).
Several studies on industrial heat demand have confirmed that more than 50 per cent
of industrial heat demand is at temperatures in the low (< 60oC), medium (60oC150oC)
and medium-high (150oC250oC) temperature ranges. The potential is especially high in
the food industry, pulp and paper industry, and textile industry. The technical potential
for solar process heat in just the Iberian Peninsula is estimated to be 5804GWh (20.9PJ).
This corresponds to 3.6 per cent of the industrial heat demand and 0.7 per cent of the
total final energy demand of this region.
This paper is a summary of the results of the European research project POSHIP, a study
on the potential for solar heat in industrial processes, funded by the European Commission
within the 5th Framework Programme. Many industries in Spain and Portugal have been
analysed in this project. Case studies have been carried out for solar systems in industries
with favourable conditions resulting in possible solar industrial plants larger than 25,000m2.
Keywords solar thermal; process heat; application potential; industry; medium temperature
7.1 INTRODUCTION
The idea of using solar heat in industry was extensively discussed in the 1980s and
several pilot plants were built. Since the 1980s the costs have come down substantially
while technologies for high efficiency solar collectors and other system components such
as controls and pumps have improved.
These developments have improved the potential for solar energy for industrial
process heat even more. This paper is a summary of the results of the European research
project POSHIP, a study on the potential of solar heat for industrial processes, funded by
the European Commission within the 5th Framework Programme.
ch-07
2/7/07
2:58 PM
Page 217
217
Within this project a large number of industries in Spain and Portugal were analysed.
Case studies were carried out for solar systems in industries with favourable conditions
resulting in possible solar industrial plants larger than 25,000m2.
ch-07
2/7/07
218
2:58 PM
Page 218
H. SCHWEIGER et al
more than 50 per cent of their total process heat in the temperature range up to
200oC.
The biggest heat demand is located in the paper and food industries. A considerable
heat demand is also found in the textile and chemical industries.
Most of the process heat in the 100200oC range is used in the food, textile and
chemical industries for such diverse applications as drying, cooking, cleaning and
extraction.
ch-07
2/7/07
2:58 PM
Page 219
Total
219
100
80
60
%
40
20
0
Food
Wine and
Beverage
Paper
Textile
Malt
Automobile
Industrial sector
Note: Industries studied in POSHIP grouped by industrial sectors
100
90
80
70
60
50
40
30
20
10
0
Food
Paper
Textile
Malt
Automobile
Industrial
Note: Industries studied in POSHIP grouped by industrial sectors
Source: POSHIP (2001)
FIGURE 7.2 Technical and economical potential of solar industrial process heat (percentage of total heat
demand)
ch-07
2/7/07
220
2:58 PM
Page 220
H. SCHWEIGER et al
TABLE 7.2 Potential for solar industrial process heat in the Spanish industry (data in GWh)
Heat demand
Low and medium temp.
Total
10%
13%
5%
10%
0%
14%
41%
3%
57%
15%
18%
14%
Notes: Data for Spain, 1999 ; asterisked data estimated as no case studies were available
Source: IDAE (2001) for total heat demand in the different sectors; own data for share of low and medium temperature heat
demand
FIGURE 7.3 Distribution of the heat demand in the industry according to different industrial sectors
total technical application potential for solar industrial process heat in Spain and Portugal
can be roughly estimated to be 5.8TWh or 3.6 per cent of the total industrial heat demand.
The relative importance of each of the sectors can be seen in Figure 7.3. The category
other sectors contains all the industries that consume heat mainly at high temperatures
(for example metallurgy, ceramics and the extractive industry).
ch-07
2/7/07
2:58 PM
Page 221
221
energy to these processes. Depending on the temperature range needed, different solar
collector technologies and different concepts for their integration into the industrial
heating system can be considered. The selection of the appropriate collector type
depends mainly on the desired working temperature and the climatic conditions.
A graphical representation of the instantaneous efficiency for different collector
technologies is given in Figure 7.4.
Flat plate collectors may be used for application with temperatures up to 80C;
however, for temperatures over 80C flat plate collectors may not be appropriate because
of low efficiency. Concentrating collector technologies may be used at higher
temperatures (Rabl, 1985).
Two major types of solar collectors can be considered for industrial heat production:
1 Stationary collectors. These collectors do not use any mechanisms to track the
sun. They can produce heat at low and medium temperatures up to 150C. Flat plate
collectors, evacuated tube collectors and compound parabolic concentrator (CPC)
type concentrators belong to this group of collectors.
2 Collectors with tracking systems. Mainly one-axis tracking collectors are used
both in solar process heat plants and in large power plants for solar thermal electricity
generation. Temperatures up to 400oC can be obtained with good efficiency.
ch-07
2/7/07
222
2:58 PM
Page 222
H. SCHWEIGER et al
losses. Radiation heat losses are reduced in most standard flat plate collectors using
selective coatings (Wackelgard et al, 2001). Heat losses can be further reduced by using
a double transparent cover, which may consist of a glass sheet and a transparent film
behind it, or two glass sheets with anti-reflective coating. The use of transparent
insulation materials is another possibility to build high efficiency stationary flat plate
collectors (Carvalho et al, 1995).
Source: S. L.Viessman
ch-07
2/7/07
2:58 PM
Page 223
223
dimension, and accept varying incident angles in the other. Usually mirror and
receiver are moved as a unit. Exceptions are the fixed focus collector with moving
mirrors and a fixed receiver (Krger et al, 2000) and a roof integrated collector with a
stationary curved mirror and a moving absorber pipe, currently being demonstrated in
a project in Raleigh, US (Gee et al, 2003).
Two-axis tracking and point focusing systems are parabolic dishes, central receiver
tower with heliostats and solar furnaces.
A typical one-axis tracking collector is the so-called parabolic trough collector (PTC, see
Figure 7.7). Parabolic trough collectors are the most mature concentrating solar
ch-07
2/7/07
224
2:58 PM
Page 224
H. SCHWEIGER et al
technology to generate heat. Reflectors with a parabolic shape concentrate the direct
solar radiation onto the receiver located in the focal line of the parabola. The receiver
consists of an absorber tube of an area usually 1525 times smaller than the aperture
area. The fluid to be heated is circulated through the absorber piping. Water and thermal
oil are typically used as working fluids.
Parabolic trough collectors have a very low thermal loss coefficient and are therefore
well suited also for applications at higher temperatures. They do not use the diffuse part of
the solar radiation; however, they make better use of the direct (beam) radiation than
stationary collectors due to the sun-tracking mechanism optimizing the position towards
the sun.
ch-07
2/7/07
2:58 PM
Page 225
225
In solar air collectors pressure drop in the collector is a very important issue and because
of that sometimes the collector is designed to take into account a compromise between
thermal performance and electric energy consumption for the air flow, as can be seen
from those available on the market.
Steam generation
Direct coupling
to the process
Central steam supply
Process
Process
Process
Return water
Feed-up water
Pre--heating of
feed up water
Source: POSHIP (2001)
FIGURE 7.8 Possibilities for the coupling of the solar system with the conventional heat supply
ch-07
2/7/07
2:58 PM
226
Page 226
H. SCHWEIGER et al
Solar
irradiation
Heat to
process
Solar
collectors
Return
Heat exchanger
Source: POSHIP (2001)
In these cases, the solar system can be designed without storage. The solar heat
produced is fed directly to the process or to the heat supply system. A schematic diagram
of such a system is given in Figure 7.9.
This is an indirect system using a heat exchanger to separate the collector circuit
(primary circuit) from the load circuit, which allows the use of special fluids to protect the
collector and its materials from freezing and corrosion.
Solar
irradiation
Storage
Heat to
process
Solar
collectors
Return
Heat exchanger
Heat exchanger
ch-07
2/7/07
2:58 PM
Page 227
227
220
250
2/7/07
228
2:58 PM
Page 228
H. SCHWEIGER et al
1000
FPC
Ener
ergy y
yiel
eld [[kWh
Wh/m2]
2]
900
Barcelona
elona
800
EFPC
ETC
700
CPC
600
PTC
500
400
300
200
100
0
60
80
100
120
140
160
180
200
Temperat
perature
ure [[C]
1000
Energy yield
y
[kWh/m
[k
/m2]
900
FPC
Lisbon
Lis
bon
EFPC
800
ETC
700
CPC
600
PTC
500
400
300
200
100
0
60
80
100
120
140
160
180
200
Temper
perat
atur
ure
e [C]
[
1000
FPC
900
Huel
uelva
EFPC
800
Ener
ergy y
yiel
eld [kWh
Wh/m2]
2]
ch-07
ETC
700
CPC
600
PTC
500
400
300
200
100
0
60
80
100
120
140
160
180
200
Temper
peratur
ure [[C]
FIGURE 7.11 Yearly energy yield delivered to a process for three sites, depending on the process temperature
ch-07
2/7/07
2:58 PM
Page 229
229
FIGURE 7.12 Heat costs for the 3 systems at different process temperatures
ch-07
2/7/07
230
2:58 PM
Page 230
H. SCHWEIGER et al
Figure 7.12 shows the heat costs in /kWh at process temperatures of 60C, 100C,
150C and 200C for the same locations as in Figure 7.11. The heat costs are calculated
including the investment costs for the collector field as given in Table 7.3. The annuity
(yearly amortized payment plus interest) is assumed to be 10.5 per cent based on a
15 years lifetime and 6 per cent interest. Maintenance costs are assumed to be
2.5/m2/year for stationary collectors and up to 5/m2/year for parabolic trough
collectors.
The solar heat costs for the most economic solution for each temperature range vary
from 0.04/kWh to 0.22/kWh depending mainly on the climate and the working
(process) temperature. The climatic conditions should therefore be carefully regarded
during planning.
Applications requiring temperatures below 150C can be supplied with solar heat at
significantly lower costs then those operating at higher temperatures. In the central and
northern Mediterranean climates of the Iberian Peninsula (for example Barcelona, Lisbon
or Madrid), heat costs can be below 0.08/kWh for a supply temperature of 100C. In the
southern Portuguese and Spanish climates solar heat at this temperature can even be
delivered at costs down to 0.04/m2.
Mapag, Austria
Gillich Galavanik
Bilderland GmbH
Mapag
Gillich
Hamminger
Lisec
Bilderland
Xsolar
ISL
Solution Solartechnik
GmbH, Austria
Sunwash, Austria
Graz
Hausmening
Lohnsburg
Graz
Gumpoldskirchen
Kflach
1979
1981
1994
1999
2001
2002
Actual state
Temperature
60
60
of use
40
Air 50C
Air 60C
Degreasing and
75
removal of laquer from
metal parts in baths
Cleaning of samples
for analytical lab
Process
operating
operating
operating
operating
2003
(October )
Type of
Collector
FPC
FPC
FPC
ETC
FPC
FPC
(1)
1284
916
88
1x10
1x32
45
(2)
collectors area
2:58 PM
Sunwash
Viessmann
operation
construction)
Year of
start of
(design and
Site
Companies
Company (user
of the plant)
2/7/07
AUSTRIA
Short name
ch-07
Page 231
FPC
FPC
FPC
Freudenstadt
Walz
Barnau
Clean Park
Filling Station
JVA Barnau
FREUDENSTADT
WALZ
BARNAU
FPC
FPC
PTC
ETC
170C
Mlheim
Refuse Collection
Department
Thermal
in
commissio
ning (first
steam was
produced
on 20 Oct
2003)
(1)
MLHEIM
1998
2003
Type of
Collector
89
24
24
25
151
770
1200
(2)
collectors area
Dresden
Targassonne
Cairo
Temperature
of use
Stadtreinigung
Dresden GmbH
Basic design,
specification and site
supervision by Fichtner
Solar GmbH, detailed
design, construction and
erection by Lotus
2003
Process
DRESDEN
GERMANY
TARGASSONNE
FRANCE
user of the
steam: El Nasr
Pharmaceuticals
and Chemicals
Actual state
232
(October )
2:58 PM
NREA
operation
construction)
Year of
start of
(design and
Site
Companies
Company (user
of the plant)
2/7/07
EGYPT
Short name
ch-07
Page 232
H. SCHWEIGER et al
Tannery; preheating of
water for the steam
boiler
1993
Athens
Tripou-Katsouri
S.A.
TRIPOUKATSOURI
Dairy; yoghurt
production
1993
Korinth
Mandrekas S.A.
1993
MANDREKAS
1993
Patras
Achaia Clauss,
S.A.
ACHAIA CLAUSS
Heraklion
Greenhouse; Space
heating
1994
Kozani
Kozani
Greenhouses S.A.
KOZANI
KASTRINOGIAN Kastrinogiannis
NIS
S.A.
Cosmetics stock
warehouse; adsorption
cooling
1998
Dairy; preheating of
water for steam
boiler; washing
machine
2000
Oinofita
Thessaloniki
Dairy; Preheating of
water for steam boiler
2000
Sarantis S.A.
Mevgal S.A.
MEVGAL
Thessaloniki
Pulling
40
70-85
40-45
40
40
FPC
308
170
180
308
80
2700
727
576
138
2:58 PM
SARANTIS
Alpino S.A.
Krimmer Pulling
2/7/07
ALPINO
GREECE
PULLING
ch-07
Page 233
Pretec
KNORR
operating
out of
service
1985
Vialonga
out of
service
operating
2003
1985
guas de Moura
2001
VIALONGA
TECNIVEST, SA
UCAL
Guetermann
Mexico
Pisticci
UCAL, AGUAS
DE MOURA
PORTUGAL
GUETERMANN
MEXICO
PISTICCI
Actual state
(October )
Brewery
Dairy
Adsorption cooling of
production hall
Chemical
Process
Temperature
40-45
63
188
of use
Type of
Collector
1120
*1280
438
1728
(2)
area
None
440
selective
flat plate
Non
192.8
evacuated
1.5 x CPC
PTC
PTC
PTC
92 axis)
(1)
collectors
2:58 PM
ITALY
operation
construction)
Year of
start of
(design and
Site
Companies
Company (user
234
of the plant)
2/7/07
Short name
ch-07
Page 234
H. SCHWEIGER et al
food industry:
farinaceous products
factory (Drying)
food industry
(Pasteurization)
1988
1983
Hallau
HALLAU
Desalination of water
Hattwill
shut down
80-120
1981
Alcorcn (Madrid)
Gran Canaria
207 (1)
1982
dismantled Sterilisation of meat
definitively (operating products (steam for
1985
19841988) back-up)
Merida
20-300
1995
Washing of cisterns,
cars, etc
1994
23-26
Juzbado
(Salamanca)
Huelva
Fish growing
1996
HUTTWILL
SWITZERLAND
Vinejard
Lactaria Espaola
LACTARIA
CASTELLANA
ARINAGA
Carnes y
Conservas
Espaolas S.A.
CARCESA
Autoclavados
Carte S.A.
CARTE
Huelva
Olives; preheating of
water
1997
ETC
ETC
500
400
360
600
1024
PTC
PTC
(CasaAuxini)
1080
PTC
138
1316
260
2:58 PM
Acuinova
Andalucia S.A.
ACUINOVA
Sevilla
2/7/07
ENUSA
TE-PE S.A.
TE-PE
SPAIN
ch-07
Page 235
Hotel Grand
Kaptan
ALANYA
Austin Energy
Phoenix Federal
Prison
Austin Energy
Parker Lincoln
Power Roof
PHOENIX
IST
Solargenix Energy
Ford
SOLEL
Solitem
Solitem
FORD Visitor
Center
U.S.A.
Iberotel
Sarigerme Park
operating
operating
operating
2002
1999
Phoenix, Arizona
Raleigh, NC
Austin, TX
2003
hot water
50-ton 2E absorption
chiller
30-ton 1E absorption
chiller
In
Absorption
construction
2003
operating
in
Absorption Cooling
construction and Laundry
Process
2003
2003
Alanya
Sarigerme
Actual state
Temperature
60
170
95
144
(steam)
144
(steam)
of use
Type of
Collector
240
180
180
(2)
area
PTC
1673
Tracking 530
RoofIntegrate
d Power
Roof
Fixed nonevacuated
CPC
PTC
PTC
PTC
(1)
collectors
236
(October )
2:58 PM
SARIGERME
operation
construction)
Year of
start of
(design and
Site
Companies
Company (user
of the plant)
2/7/07
TURKEY
Short name
ch-07
Page 236
H. SCHWEIGER et al
IST
ACDF
Tehachapi, California
TEHACHAPI
320-ton 1E absorption
chiller plus space
heating
Shut down
in 1987
1980
Textile
Malt
operating
operating
Druomg
Laundry
operating
1980
20-ton 2E absorption
chiller and space
heating
process-water
ICPC VTC
Collectors
Operating
1983
Notes: (1) PTC: parabolic trough collectors; ETC: vacuum tube collector; FPC: flat plate collector
(2) Useful collector area
Source: POSHIP (2001)
Texas
SAN CH07
IST
Georgia
MACON
California
Correctional
Institution
Alabama
FAIRFAX
California
PASADENA
Arizona
CHANDLER
1986
1997
105/55
110
95
85
83 to 165
1000
5620
725
1150
2676
PTC
1000
900
PTC
PTC
PTC
Sunmaste 1100
r ETC
with
external
CPC
reflectors
PTC
PTC
PTC
Non107
tracking
Integrated
Compound
Parabolic
Concentra
ting ETC
2:58 PM
Brighton, Colorado
Sacramento, CA
2/7/07
Adams County
Detention Facility
Solar Enterprises
International and BES
ch-07
Page 237
ch-07
2/7/07
2:58 PM
238
Page 238
H. SCHWEIGER et al
Top: Production of hot water with parabolic trough collectors, California, US; Bottom: Knorr factory, Portugal
Source: POSHIP (2001)
ch-07
2/7/07
2:58 PM
Page 239
239
ch-07
2/7/07
240
2:58 PM
Page 240
H. SCHWEIGER et al
(percentage of the heat demand covered by the solar system), the lower the mean
working temperature. In very low solar fractions, the mean working temperature may be
close to the fluid inlet (or return) temperature.
Ttank (C)
P(W)
DISCHARGE
CHARGE
USEFUL SOLAR
HEAT
DISCONTINUOUS
PROCESS
OF THE STORAGE
STORAGE
TEMPERATURE
HEAT
00:00 h
FIGURE 7.14 Short-term mismatch between heat demand and available solar heat
24:00 h
ch-07
2/7/07
2:58 PM
Page 241
241
Bilbao
Barcelona
Madrid
Lisboa
Huelva
1200-1300
1300-1400
1400-1500
1500-1600
1600-1700
1700-1750
Source: Meteotest
FIGURE 7.15 Global solar radiation on a horizontal surface (H) on the Iberian Peninsula
2/7/07
2:58 PM
242
Page 242
H. SCHWEIGER et al
temperatures above 100C at present are only recommended if all other conditions are
favourable (high solar radiation and continuous demand zones).
Another important factor to be considered is the systems size. System costs for small
systems (100m2) can be more than 50 per cent higher than those of large solar systems
(> 1000m2), with a corresponding increase in heat costs.
160
140
4
11
120
ch-07
100
17
3
20
80
22
23
24
60
18
19
15
10
12
1
16
40
13
20
21
2
14
0
0
20
40
60
80
100
120
140
Working temperature
Notes: Colours: very high (black), high (grey), medium (white) solar radiation.
demand;
: Systems with continuous demond during the week and weekend breaks;
: systems with
continuous demand during the week and weekend breaks, but with long-term (weeked) storage; : systems
with only seasonal operation.
Source: POSHIP (2001)
FIGURE 7.16 Costs of useful heat for the different systems as a function of the mean annual
working temperature
160
ch-07
2/7/07
2:58 PM
Page 243
243
PROCESS
Unglazed selective collectors or low cost standard flat plate collectors
Highly selective flat plate collectors or CPC collectors
CPC collectors, evac. tubes or other high efficient stationary
collectors; concentrating collectors for medium and large systems
Concentrating collectors; evac. tubes with CPC
Optimum annual heat gains per unit collector area in the Iberian Peninsula (latitude from
37 to 42) are obtained for non-tracking systems at collector inclinations of 30 and a row
separation of at least 1.5 times the gross collector (row-) height. The fraction of gross
collector area to roof area in this case is less than two thirds. If a limited roof area is available
and maximum total solar heat gains are desired, the collector inclination can be set down to
10 approximately. Nevertheless, for glazed collectors a lower limit of 20 for the collectors
inclination is recommended in order to avoid rain water infiltration and enhance self-cleaning
of the glass sheet by rain water. A deviation from south orientation of up to 45 can be
tolerated, leading to about a 10 per cent reduction in the systems output.
Parabolic trough collectors should be tracked around an axis oriented NS, to maximize
the annual heat production. The row spacing should be about three times the aperture width
to avoid excessive shadowing in the morning and evening. By an appropriate design of flow
rates, pipe diameters and pipe insulation, the electricity consumption for fluid circulation can
be below 1 per cent of the overall heat gains. Thermal losses in the piping and storage
should not be above 5 per cent of the overall heat gains for medium and large size systems.
In earlier studies, depending on the load characteristics and profile, different solutions
have been studied, like single-pass versus multi-pass strategies (Gordon and Rabl, 1982;
Collares Pereira et al, 1984) or constant versus variable flow rate (Gordon and Zarmi, 1985)
in order to assure good temperature stratification in the storage.
Solar gains are highest if the load and solar supply flow rates are matched with the
help of flow regulators. A bypass for the preheating of the primary circuit during the
morning hours and the extraction of the residual heat after sundown by appropriate
regulation can improve solar gains by up to 20 per cent.
7.4.2.2 Storage
An introduction to solar systems with heat storage has been given in section 7.3.3.
Short-term heat storage is recommended whenever a mismatch between available
solar radiation and heat demand occurs. For short-term storage (several hours), volumes
of about 25 litres/m2 are recommended. Short-term storage may even be recommended
for continuously operating processes, in order to lower the mean working temperature of
the solar system and thereby improving its efficiency, especially if low cost solar
collectors with high thermal loss coefficients are used.
The larger the systems size, the more effective the heat storage over longer periods
(for example weekends). Weekend storage becomes economic from about 500m2 and
above. Weekend storage volumes should be about 250 litres/m2. In this case, storage
2/7/07
244
2:58 PM
Page 244
H. SCHWEIGER et al
costs can range from 10 to 20 per cent of the total system costs. Storage for longer
periods (seasonal storage) can only be considered for very large systems (> 5000m2).
Hot water
to process
Solar
collectors
Waste
water
Heat
exchanger
Heat
recovery
FIGURE 7.17 Combination of solar thermal system and waste heat recovery
3000
2500
2000
MWh
ch-07
1500
1000
500
0
solar heat
cogeneration
cogeneration
electricity
high eff.
low eff.
Note: Conversion efficiency (electrical) for: (a) highly efficient (combined cycle) cogeneration plants (50%); (b) low efficiency
cogeneration plants (25%)
Source: POSHIP (2001)
FIGURE 7.18 Primary energy saving with respect to a conventional steam boiler, for 1000MWh of
industrial process heat produced either by solar thermal systems or conventional cogeneration system
ch-07
2/7/07
2:58 PM
Page 245
245
TABLE 7.6 Evaluation criteria for solar industrial process heat systems for the Iberian Penninsula
CRITERION
Working temperature
Climate
System size
Source: POSHIP(2001)
be introduced only after first preheating by waste heat recovery systems (Figure 7.17), and
not as an alternative to these systems. Even if the waste heat recovery raises the working
temperature in the solar system, the combination of both systems yields better results
than a solar system at lower temperature without heat recovery.
If solar heat is being fed to several processes, a control strategy should be chosen
in order to reach an optimum overall energy saving. In most cases feeding solar energy
ch-07
2/7/07
246
2:58 PM
Page 246
H. SCHWEIGER et al
to the process with the lowest temperature might be the best choice, but in some cases
giving preference to heat production at higher temperatures during the hours around
noon with high insolation results in a better overall performance. The control strategy
should be optimized for the specific cases using dynamic system simulation
techniques.
7.4.2.5 Summary
The most important evaluation criteria for solar process heat plants are summarized in
Table 7.6.
ch-07
2/7/07
2:58 PM
Page 247
247
TABLE 7.7 List of the most relevant parameters of the demonstration projects proposals
COMPANY
SECTOR
LOCATION
PROCESS
WORKING
TEMPERATURE
RANGE
COLLECTOR
TYPE**
PROJECTED
COLLECTOR
AREA (m2)
ESTIMATED
COST ()
Malte Ibrica
Malting factory
Poceiro, Portugal
Low temperature
(2060C)
FPC or CPC
3500
624,000
Air preheating
for malt drying
Low temperature
(1080C)
SFPC
3046
930,000
Beirala Lanificios
Textile factory
Seia, Portugal
FPC or CPC
1750
340,000
CPC
60
55,000*
**FPC: flat plate collector; CPC: compound parabolic concentrator; SFPC: selective flat plate collector.
Source: POSHIP (2001)
ch-07
2/7/07
248
2:58 PM
Page 248
H. SCHWEIGER et al
Cold generation
19%
Malt drying
81%
Source: POSHIP (2001)
air at about 60C). Most of the remaining 20 per cent is required for air cooling in the
germination process (Figure 7.20).
The profile of the heat demand in this factory is ideal for the implementation of a solar
system:
The heat demand is high and continuous during 7 days a week and 24 hours a day,
so nearly all the collected solar energy can be used directly in the process without
need of large storage volumes.
The required temperatures are low. Air is preheated by the solar system only a few
degrees above ambient temperature.
The factory is located in a region with a high solar radiation (1770kWh/m2a).
Energy costs are a significant cost factor in the malt production process, so a
reduction of energy consumption can help improve competitiveness.
A solar system is proposed for the preheating of air for the malt drying process. An
additional hot water to air heat exchanger will be installed in series before the existing
heat exchanger using steam from the conventional heat supply system. The solar system
consists of a solar collector field of a total surface of 5000m2 and a storage tank of 1000m3
(Figure 7.21 and Table 7.8). The solar collectors will be integrated in the saw-tooth shaped
roof structure forming a rain-tight solar roof.
ch-07
2/7/07
2:58 PM
Page 249
249
FIGURE 7.21 Scheme of the proposed solar system for the malting factory in Seville
TABLE 7.8 Characteristic data of the proposed system for solar air heating in a maltting factory
in Seville, Spain
Company
Location
Ind. sector
Process
Working temperature
Solar collector area
Solar collector type
Solar storage
Annual solar gains (useful heat)
Solar fraction
Total investment
Total own contribution
Yearly net savings
Payback
Source: POSHIP (2001)
The factory produces malt, which undergoes three steps during the process:
1 wetting of barley;
2 germination of green malt (cold); and
3 drying of malt (heat).
The largest part of the energy consumption is for the production of superheated water
(180C, 14 bars), which is used with the waterair heat exchangers for drying the malt
ch-07
2/7/07
250
2:58 PM
Page 250
H. SCHWEIGER et al
(6080C). A small part is consumed for keeping thick fuel at its appropriate storage
temperature. The factory is quite new (less than 10 years old) and its design incorporates
heat recovery measures that improve the potential use of solar energy when considering
new energy-saving actions.
preheated air
2550C
Collector field
FIGURE 7.23 Layout of the proposed solar system for the malting factory at Poceiro, Portugal
ch-07
2/7/07
2:58 PM
Page 251
251
There are several favourable factors for the implementation of a solar system in the
factory:
Grupo Unicer
Poceiro (Setbal)
Malting factory
Air preheating for malt drying
2080C
3500m2
Selective flat plate or CPC collectors
3762MWh/year
20%
624,121
405,679
118,320
4 years
ch-07
2/7/07
2:58 PM
252
Page 252
H. SCHWEIGER et al
FIGURE 7.24 Solar radiation in Portugal and location of the Beiral plant
access to the national natural gas network is not expected in the short term;
due to the high transportation costs, it is very expensive to change from heavy fuel
costs and is ideal to meet the strategic ISO 14000 certification target;
a large area, with good orientation, is available on the roof of the factory for placing
solar collectors;
the solar system can be designed in a very common way, storing the collected
solar tank after increasing its temperature by the back-up conventional energy
system as necessary.
Back-up
FIGURE 7.25 Scheme of the proposed solar system for the Beiral plant
2/7/07
2:58 PM
Page 253
253
The scheme of the proposed solar system is given in Figure 7.25. The most relevant data
for the proposed plant are summarized in Table 7.10.
TABLE 7.10 Characteristic data of the proposed system for a textile drying process
Company
Location
Ind. sector
Process
Working temperature
Solar collector area
Collector type
Storage volume
Annual solar gains (useful heat)
Solar fraction
Total investment
Total own contribution
Yearly net savings
Payback
Beiral
Seia
Textile
Preheating of water for dyeing process
2080C
1750m2
Stationary collectors, selective flat plate or CPC type
75m3
1331MWh/year
33%
340,430
221,280
84,496
3 years
Heating
DHW
10000
8000
6000
[kWh]
4000
2000
be
r
be
r
De
ce
m
No
ve
m
O
ct
ob
er
be
r
Se
pt
em
Au
gu
s
Ju
ly
Ju
ne
ay
M
Ap
ril
ar
ch
M
Fe
br
ua
ry
ry
Ja
nu
a
ch-07
FIGURE 7.26 Heat, cold and domestic hot water demand of the installations of Bodegas Mas Martinet
2/7/07
2:58 PM
254
Page 254
H. SCHWEIGER et al
Solar contribution
10000
8000
6000
4000
2000
be
r
be
r
De
ce
m
No
ve
m
O
ct
ob
er
be
r
Se
pt
em
Au
gu
s
Ju
ly
Ju
ne
ay
M
Ap
ril
ar
ch
M
Fe
br
ua
ry
ry
Ja
nu
a
ch-07
FIGURE 7.27 Total heat demand of Bodegas Mas Martinet and solar contribution throughout the year
favourable conditions are encountered for an alternative thermal cooling system that
substitutes the electricity consumption of the compression chillers.
A solar system is proposed for the air conditioning of the wine cellars and the office
building using a lithium bromide absorption cooling machine (Yazaki Company). The
excess heat from the solar collectors during the winter months is used for space heating
and domestic hot water production. In the proposed system, fan-coils are used for the
distribution of heat and cold to the wine cellar and the connected office building, and a
floor heating system for a nearby residential building.
The solar system consists of a solar collector field of 60m2 of CPC collectors
producing hot water at 8090C and 1500L of hot water storage. The absorption chiller
has a nominal cooling power of 35kW and is buffered by a 3000L cold storage. The total
heat demand of the factory (including the heat demand of the absorption chiller) and the
solar contribution is given in Figure 7.27. Figure 7.28 shows the hydraulic scheme of the
plant and Table 7.11 gives the characteristics of the solar system.
ch-07
2/7/07
2:58 PM
Page 255
Zona casa
floor
heating
system
cooling
from pool
Parte ya instalada
ACS casa
Bacs-casa
Suelo radiante
Ttemp-SR
Vcasa Vtemp-SR
255
Retorno a balsa
Agua de
la balsa
T
BSR
T
VSR
Caldera
calefaccin
Bcasa
Tbalsa
Filtro
TSR-ret
Bbalsa
Campo de colectores
Acumulador de calor
Thcol-o
Vdep1500
Ftorre
Mquina de absorcin
Caldera auxiliar
Ths-top
Thx-hi
Tcaldera
Thot-ma
Vhot-ma
Bhot-ma
Vbalsa
1 500 l
B1ario
Thx-ho
B2ario
Ths-bot
Tret-2ario
Benfriamiento
Agua de
renovacin
Tevap-o
ACS oficina
Vinter-dep
Tacs-of-top
solar
thermal
system
Sala de calderas
Bacs-of
Vcal-ida
T cal-ida
Vcr-of-ret
Tcs-top
TFC-ret
Vfro-aux2
absorption
cooling
system
VD-evap2
Vcr-of
3 000 l
Bof
VD-evap1
Tcs-bot
Acumulador de
fro / calor
Vfro-ida
Vfro-aux1
Balm
Tfro-ida
Bbod
F.C.- B2
F.C.- B1
F.C.- B3
F.C.- B4
F.C.- Alm.
F.C-Of.1
F.C-Of.2
F.C-Of.3
FIGURE 7.28 Scheme of the proposed solar system for Bodegas Mas Martinet
TABLE 7.11 Characteristic data of the proposed system for space heating and cooling of a wine cellar
Company
Location
Ind. sector
Process
Working temperature
Solar collector area
Solar collector type
Solar storage
Annual solar gains (useful heat)
Solar fraction
Total investment
Total own contribution
Yearly net savings
Payback
11 years
Reference data for this study were obtained from a typical factory in Hamm, Germany.
The main steam-consuming process is the curing of the material under precisely
controlled conditions in large autoclaves (Figure 7.29). Several autoclaves are operated in
a staggered mode, and Ruth-type (saturated steam) storages are commonly used to
ch-07
2/7/07
256
2:58 PM
Page 256
H. SCHWEIGER et al
pressure [bar]
steam flow [t/h]
15
10
5
0
-5
time [hours]
10
10
15
Source: Hennecke et al (2002)
manage the steam flows efficiently during charging and discharging. It should be pointed
out that the flexibility of the system, designed to cope with the severe load changes
caused by the batch processes, also provides distinct advantages for the integration of
solar process steam. A solar field was designed to fit the roof of the main factory hall,
consisting of small collectors like the parabolic trough collector from Industrial Solar
Technology (IST), with a total aperture area of about 1400m2 (Figure 7.30). The nominal
power output (600kWth) represents about 5 per cent of the total fossil-fired boiler capacity
installed. The area of the surrounding storage yards is large enough to provide up to 100
FIGURE 7.30 Solar field integration via Ruth-type (saturated steam) storage
ch-07
2/7/07
2:58 PM
Page 257
257
per cent of the factorys total steam demand with collector systems mounted on elevated
structures.
Care must be taken in the long, horizontal and asymmetrically heated boiler tubes to
avoid potentially dangerous flow patterns of the two-phase flow. Theoretical
investigations indicate that the problem of flow stratification will be reduced at lower
pressures. However, below about 5 bars the low steam density leads to high velocities
and pressure drops, which might prevent the achievement of high steam qualities at the
outlet of the solar steam generator. The recirculation process provides a convenient
solution, particularly if the existing ruths steam storages can be used as the steam drum
to separate steam and water. During times when no steam is extracted for the production,
the solar energy input will lead to rising pressure in the system, increasing the exergetic
value of the stored energy and consequently reducing live steam consumption
requirements by the same amount. At the same time, any direct interference between the
solar steam generator and the delicate production process can be avoided.
The annual system performance has been simulated for different sites as a basis for
cost estimations, indicating that in the Mediterranean area the solar steam may be
competitive with conventional boilers (Table 7.12). At favourable sites, 10 per cent of the
annual steam demand can be satisfied from the solar field sized to provide 5 per cent of
the fossil boiler capacity.
TABLE 7.12 Annual system performance for different sites
LOCATION
Wrzburg, Germany
Portoroz, Slovenia
Mugla, Spain
DNI
kWh/m2a
894
1330
1900
7.6 CONCLUSIONS
Industrial solar thermal systems as a renewable energy source can cover a significant fraction
of the industrial heat and electricity demand. This industrial heat demand constitutes about a
third of the total final energy demand in southern European countries. About 7 per cent of the
total final energy is consumed in the form of industrial process heat at temperatures below
250C. Fulfilling the goal of installing 2,000,000m2 of solar thermal collectors for industrial
process heat and solar cooling, as outlined in the Campaign for Take-Off of the European
Commission would represent a primary energy saving of about 2,000,000MWh/year.
Therefore, solar thermal energy in industry can be an important contribution to a reliable,
clean, safe and cost-effective energy supply based on renewable energy sources.
Solar collectors for hot water production at low temperatures are a well-known and
well-extended standard technology. With recently developed high performance solar
collectors, heat at temperatures up to 250oC can be produced with excellent efficiency.
Heat at these temperatures is required in many industrial processes, such as steam
generation, washing, drying, distillation and pasteurization.
ch-07
2/7/07
2:58 PM
258
Page 258
H. SCHWEIGER et al
At present, industrial solar thermal systems are operating in Europe with a total
installed collector surface of about 10,000m2. Solar collector arrays can be integrated
either into industrial roofs or installed on an available ground area.
Large-scale industrial installations may lead to very low cost systems, meaning solar
systems for industrial process heat production can become economically competitive
with fossil fuels on a very short term. Present investment costs for solar thermal systems
range from 250 to 500/m2 (corresponding to 2501000/kW of thermal power), leading
to average energy costs in southern Europe from 2 to 5 cents/kWh for low temperature
applications and from 5 to 15 cents/kWh for medium temperature systems.
The present cost per kWh of saved primary energy in industrial applications is less
than that of small- to medium - scale solar domestic hot water applications and can be
further reduced. Cost reductions can be obtained by large-scale production, reducing
operation and maintenance costs and improving collector efficiency and system design,
especially for medium temperature solar collectors. Cost reductions of up to 50 per cent
on a medium term are expected by 2010.
REFERENCES
Carvalho, M. J., Collares Pereira, M., Oliveira, J. C., Mendes, J. F., Haberle, A. and Wittwer, V. (1995) Optical and thermal
testing of a new 1.12X CPC solar collector, Solar Energy Materials and Solar Cells, no 37, pp175190
Collares Pereira, M., Gordon, J. M., Rabl, A. and Zarmi, Y. (1984) Design and optimisation of solar industrial hot water
systems with storage, Solar Energy, no 32, p121
Garg, H. P. (1987) Advances in Solar Energy Technology, vol 2, Reidel Publishing Company, Dordrecht, The Netherlands
Gee, R., Cohen, G., Winston, R., Greenwood, K. and McGuffey, B. (2003) Design, installation and early operation of a roofintegrated solar cooling and heating system, proceedings of the ASES Annual Conference, Dordrecht, The Netherlands,
June 2126
Gordon, J. M. and Rabl, A. (1982) Design, analysis and optimisation of solar industrial process heat plants without storage,
Solar Energy, vol 28, no 6, pp519530
Gordon, J. M. and Zarmi, Y. (1985) Single-pass open-loop solar thermal energy systems: On maximum energy delivery by
variable flow rate, Journal of Solar Energy Engineering, no 107, p273
Hennecke, K., Meinecke, W. and Krger, D. (2000) Integration of solar energy into industrial process heat and cogeneration
systems, proceedings of 9th International Symposium on Solar Thermal Concentrating Technologies, Font-Romeu, France, June
ch-07
2/7/07
2:58 PM
Page 259
259
Hennecke, K., Ktter, J., Michel, O. and Peric, D. (2002) Solar process steam generation for the production of porous
concrete, in A. Steinfeld (ed) 11th SolarPACES International Symposium on Concentrated Solar Power and Chemical
Energy Technologies, Zurich, Switzerland, 24 September, Paul Scherrer Institut, Villigen, Switzerland
Huerdes, J. V. and Lachal, B. (1986) Calor Solar Industrial, Spanish edition, Delta Energy Ltd, Rafz, Switzerland
IDAE Instituto para la Diversificacin y el Ahorro de la Energa (2001) Private communication: official data from the Spanish
Ministry for Industry and Energy (MINER) for 1999
Kreider, J. F. (1979) Medium and High Temperature Solar Processes, Academic Press, New York
Krger, D., Hoffschmidt, B., Hennecke, K., Pitz-Paal, R., Rietbrock, P. and Fend, T. (2000) Results from parabolic trough collectors for
process heat at the DLR Cologne, in H. Kreetz, K. Lovegrove and W. Meike (eds) Proceedings of the 10th International
Symposium on Solar Thermal Concentrating Technologies, Sydney, 810 March, Australia National University, Canberra, Australia
Krger, D., Pitz-Paal, R., Lokurlu, A. and Richarts, F. (2002) Solar cooling and heating with parabolic trough collectors in the
Mediterranean, in A. Steinfeld (ed) Proceedings of the 11th International Symposium on Concentrated Solar Power and
Chemical Energy Technologies, Zurich, Switzerland, 24 September, Paul Scherrer Institut, Villigen, Switzerland
Laue, H. J. and Reichert, J. (1994) Potential for medium and large sized industrial heat pumps in Europe, Directorate General
for Energy (DGXII), European Commission, contract no XVII/7001/90-8, final report
Lewis, C. W. (1980) The prospects for solar energy use in industry within the United Kingdom, Solar Energy, no 24, pp4753
Nitsch, J. and Luther, J. (1990) Energieversorgung der Zukunft, Springer-Verlag, Berlin
POSHIP (2001) The potential of solar heat for industrial processes, Project NNE5-1999-0308, final report, funded by the
European Commission
Rabl, A. (1985) Active Solar Collectors and Their Applications, Oxford University Press, Oxford, UK
Rabl, A. (1976) Comparison of solar concentrators, Solar Energy, no 18, p93
Schreitmller, K. (1986) Erfahrungen mit solaren Prozewrmeanlagen Solar Thermische Prozewrme, Klaus Scharmer,
Gesellschaft fr Entwicklungstechnologie mbH, Aldenhoven
Schreitmller, K. (1987) Stand und Perspektiven der solaren Prozewrmebereitung unter besondrere Bercksichtigung der
Aspekte von Entwicklungslndern, DFVLR, Institut fr technische Thermodynamik
TNO Bouw (1995) Potentieelstudie Hoogrendement-zonnecollectoren en Nederland, Report 94-BBI-R 1368, TNO, Delft, The
Netherlands
TRNSYS (2000) Software for Dynamical System Simulation, Solar Energy Laboratory, Madison, Wisconsin, US, Version 14.2
(1997) and 15.0
Wackelgard, E., Niklasson, G. A. and Granqvist, C. G. (2001) Selectively solar-absorbing coatings, J. Gordon (ed) Solar
Energy, the State of the Art, James & James, London
Welford, W. T. and Winston, R. (1978) The Optics of Non-Imaging Concentrators Light and Solar Energy, Academic Press
Winter, C.-J. (1997) Sonnenenergie nutzen. Technik, Wirtschaft, Umwelt, Klima, VDE-Verlag GmbH, Berlin
ch-07
2/7/07
2:58 PM
Page 260
ch-08
2/7/07
2:38 PM
Page 261
Abstract
Our study reveals that concentrating solar power (CSP) can be used to fulfil the total
electricity demand in Middle East (ME) and North African (NA) countries of 1700TWh/y
in 2025, rising to 3600TWh/y in 2050, especially as an expanding technology with a
growth of 25 to 35 per cent a year. A strong competition of several CSP technologies
could lead to higher overall CSP growth rates than those assumed in the present study
(around 30 per cent per year). In 2025 14 per cent of the electricity demand of MENA
may be covered by CSP, which will become the dominating power source with a share
of 57 per cent by 2050.
According to our scenarios the mix of renewable energies will cost around 5
cents/kWh in 2050. The cost of CSP will be reduced from the current 8 cents/kWh to
6 cents/kWh in 2010 to 5 cents/kWh in 2020 (other scenarios quote 15 cents/kWh). From
2020 until 2050, the MENA region will save US$250 billion compared to the business as
usual policy scenario, where it was assumed that the fuel prices start at US$25/bbl for
oil and US$49/ton for coal and escalate by only 1 per cent per year (currently the prices
are at a level of US$55/bbl and US$65/ton respectively and the escalation rates have
amounted to 40 per cent per year since 2003).
The study shows that solar thermal power plants are suitable for seawater
desalination. A concentrating solar thermal collector array required for desalinating
1 billion m3/y would cover a total land area of approximately 10 x 10km, corresponding
to about 10m3 of desalinated water per m2 of collector area. About 10 per cent of the
desalinated water would suffice to irrigate the desert land beneath the collectors with
a water column of 1m/y.
The study shows that the total carbon emissions of electricity generation of all
MENA countries can be reduced from 770 million tons per year to 475 million tons in
2050, instead of increasing to 2000 million tons.
Furthermore, the greenhouse emissions from constructing the plants for most
renewable energy technologies are between 10 and 25g/kWh of useful energy. In
the case of photovoltaic systems, reductions are possible in the medium term to
ch-08
2/7/07
2:38 PM
262
Page 262
about 50g/kWh. The scenario reaches a per capita emission of 0.58 tons/cap/y in the
power sector in 2050, which is acceptable in terms of the recommended total emission
of 11.5 tons/cap.
Keywords installed power capacity; global solar radiation; water deficiency; MENA; wave and tidal
power
8.1 INTRODUCTION
The MED-CSP study Concentrating solar power for the Mediterranean region (DLR,
2005), which aims to catalyse the cooperation between Europe and Africa in solar energy
application, focused on the electricity and water supply of the regions and countries
illustrated in Figure 8.1, including southern Europe (Portugal, Spain, Italy, Greece, Cyprus
and Malta), North Africa (Morocco, Algeria, Tunisia, Libya and Egypt), western Asia (Turkey,
Iran, Iraq, Jordan, Palestine & Israel, Lebanon and Syria) and the Arabian Peninsula (Saudi
Arabia, Yemen, Oman, United Arab Emirates, Kuwait, Qatar and Bahrain).
The Arab countries in North Africa (NA) and in West Asia (WA), in general, lack energy
and water. The average annual rainfall in Arab countries is from 5 to 45mm/y, while in
European countries (mainly those in the Mediterranean EU) it ranges from 200 to
400mm/y. The total internal water reserve is only 100km3 for Arab countries, while in
Europe it may exceed 400km3 (bear in mind that the world internal water reserve is
43,764km3). Nearly 50 per cent of the Arab countries have water availability per capita less
than the absolute water scarcity level (200m3/capita/year) while the rest, except Iraq, are
in water scarcity threshold level (1000m3/capita/year) (ESCWA, 2003). In Europe, the per
Southern Europe
Western Asia
Arabian Peninsula
North Africa
ch-08
2/7/07
2:38 PM
Page 263
263
capita water resource ranges as high as 85,478m3/year (Norway). In fact among the
countries with least water resource (UNDP, 2002) we have 14 Arab countries (Kuwait, UAE,
Qatar, Libya, Saudi Arabia, Jordan, Bahrain, Yemen, Oman, Algeria, Tunisia, Egypt,
Morocco and Palestine) where the per capita water supply ranges from 10m3/year (Kuwait)
to 971m3/year (Egypt).
In 1999, electricity consumption per capita in Western Europe was never less than
4000kWh, while it was as low as 46kWh in Sudan. In the majority of the Arab countries,
except Arabian Gulf countries (GCC), the average person consumes electricity at the rate
of 1200kWh annually. Probably more than 1 million Arab citizens (of some 300 million)
have no access to electricity (UNDP, 2002).
Furthermore, the European Organisation for Economic Co-operation and Development
(OECD) countries emitted 3800 million tons of CO2, which is 15.2 per cent of the global
emission (25,000 million tons), in 2001. The CO2 emissions from the Middle East are only
4.8 per cent compared to North America (27.7 per cent), Eastern Europe (12.6 per cent),
Western Europe (15.6 per cent), Africa (8.8 per cent), Central and South America (4.1 per
cent), and the Far East and Oceania (31.5 per cent) (World Almanac, 2004). According to
the latest reports, Germany had managed to reduce emission of CO2 by only 20 per cent
(compared to the base year 1990), while others had actually increased their emissions or
stayed the same for example the UK 12.5 per cent, Italy 6.5 per cent, France 0 per cent,
Russia 6 per cent, the US 7 per cent and Japan 6 per cent (Global Emission Newsletter,
2004). These industrialized nations are required to reduce their CO2 emissions according
to the Kyoto Protocol, which took effect on 16 February 2006 by 5.2 per cent of their
emissions in 1990. The increased temperature of the globe due to global warming due
to extensive use of fossil fuels has led to more extreme weather and hence
catastrophes. According to Swiss Re-insurance Company (World Almanac, 2004) in its
report on natural catastrophes and man-made disasters in 2003, nearly 36 natural
accidents had occurred in Europe with 424 victims wasting US$2173 million (11.8 per
cent of the total wasting cost budget) while in the Middle East and Asia 178 accidents
had occurred with 51,894 victims wasting US$1447 million (7.8 per cent). Therefore,
mutual cooperation between Arab and European countries in the field of energy
production using solar radiation is strongly advisable.
The Arab countries are characterized and blessed with abundant direct solar radiation,
ranging from 4.1kWh/m2/day in Mosul, Iraq to 6.7kWh/m2/day in Nouakchott, Mauritania
(Alnaser et al, 2004; ESCWA, 2001). Of course, the total solar radiation will be larger than
these figures by nearly 30 per cent. Moreover, the maximum recorded annual mean
sunshine duration ranges from 7.5 hrs in Tunis to 10.7 hrs in Egypt. These figures are
much larger, at least three times, compared to European countries (Alnaser et al, 2004).
The electricity demand and temporal behaviour among MENA and European countries
is found to complement each other. Therefore cooperation will enhance renewable energy
utilization worldwide and it will increase to more than the current level of 13.8 per cent of
the total primary energy supply, where 2.3 per cent is from hydro, 11.0 per cent from
combustible renewable and waste and only 0.5 per cent covers solar (0.039 per cent), wind
(0.026 per cent), tidal (0.004 per cent) and geothermal (0.440 per cent) power. The solar
power market is already growing it was 1000MW in 2000 and is expected to be
ch-08
2/7/07
264
2:38 PM
Page 264
14,000MW by 2010 and 70,000MW in 2020 (UNDP, 2002). This, of course, will minimize the
cost per watt for each renewable energy source, especially solar thermal and photovoltaic.
The present study is the result of an analysis of the Middle East and North Africa
(MENA) countries with respect to their electricity demand and the possibility to satisfy this
demand by solar thermal power plants in the coming decades. Most of the findings of this
paper are derived from the final report Concentrating solar power for the Mediterranean
Region (DLR, 2005).
This paper describes the importance of clean and sustainable energy exports among
MENA countries and Europe in the form of high voltage potential produced in Arab
countries, using solar thermal (the most suitable technology for the region) or other
renewable energy resources; this is, of course, after using part of this energy for water
desalination in Arab countries, where 65 per cent of the water resources are politically in
debate with non-Arab countries, which may ignite conflicts. It will also show how the
environment will be protected against CO2 emissions if solar thermal technology is utilized.
8.2.1.1 Population
All scenarios share the same population prospect the World Population Prospect by the
UN, middle path, and revision 2002, including the age structure of the population. The last
is not taken into account explicitly but will be used to discuss the feasibility of economic
growth and what might happen if the work force and not the population would be
considered as a driving force.
In addition, it should be noted that according to the scenario in the next 50 years most
of the states under investigation are in a demographic transformation due to decreasing
birth rates, which will raise the share of the workforce in the population to a high level. At
the end of the period the share of old people will increase significantly, however. Even with
given wage levels, such a demographic transformation offers the opportunity to reap a socalled demographic dividend within the next 30 years when the share of working people
will increase. Therefore rising income per capita and rising savings can be expected.
ch-08
2/7/07
2:38 PM
Page 265
265
8.2.2 THE LINK BETWEEN GDP PER CAPITA AND ELECTRICITY CONSUMPTION
PER CAPITA
The essential part for deriving the electricity demand is a link between GDP per capita and
electricity consumption per capita. With the population growth, the overall electricity
demand can obviously be calculated.
The link between the per capita values used the following steps:
1. For every year in the period 19602001, regressions were calculated between GDP
per capita and total final consumption (TFC) for electricity per capita (IEA, 2002;
Heston et al, 2002).
2 For the two parameters of the regression equations, time trends were estimated
(using power and alternatively linear functions). A power function gave a
significantly better fit for the absolute term. For the power it was difficult to
distinguish a linear trend from a power trend. So both were used, resulting in two
alternative links between GDP per capita and electricity consumption per capita.
The linear trend gives a scenario with high efficiency increase while the power
trend results generally in increasing electricity intensities. It should be noted that
with these parameter trends, the link between GDP per capita and electricity per
capita no longer depends solely on growth rates. Also the absolute value shows an
influence.
3 From the TFC for electricity, the domestic demand was derived using International
Energy Agency (IEA) data for 2001 (IEA, 2002) on distribution losses, consumption in
the energy sector and the so-called own use. These consumptions were split into a
proportional and a fixed term. The fixed term is meant to accommodate the use for
oil production. The proportional term was linearly reduced to a level which is now
common in industrial countries (8 per cent). It should be noted that the data on these
terms are not of a high quality and are sometimes missing. Fortunately, the impact of
these terms is generally small.
4 The resulting general functions were calibrated to individual countries
assuming a linear mix of the current values and the estimated value. The
weight of the estimated value is assumed to increase linearly from a current
0 to 1 in 2050.
5 The two scenarios are obtained by combining high economic growth with high
efficiency growth and low economic growth with low efficiency growth, as the
increase of efficiency is coupled to investment and the higher growth rates result in
higher investment rates and a higher share of new machineries.
ch-08
2/7/07
266
2:38 PM
Page 266
Both the technical and economic potentials were defined for each renewable energy
resource and for each country. For each resource and for each country, a performance
indicator was defined that represents the average renewable energy yield with which the
national potential could be exploited (Table 8.1). The economic potentials are those with a
sufficiently high performance indicator that will allow new plants in the medium and long
term to become competitive with other renewable and conventional power sources,
considering their potential technical development and economies of scale (Table 8.2).
The renewable energy potentials for power generation differ widely in the countries
analysed within this study. Altogether they can cope with the growing demand of the
developing economies in MENA. The economic wind, biomass, geothermal and hydropower
resources combined amount to about 400TWh/y. Those resources are more or less locally
concentrated and not available everywhere, but can be distributed through the electricity grid,
which will be reinforced in the future in line with the growing electricity demand of this region.
The greatest resource in MENA is solar irradiance, with a potential that is by several orders
of magnitude larger than the total world electricity demand. The solar energy irradiated on the
ground equals one to two barrels of fuel oil per square metre per year of primary energy. This
magnificent resource can be used both in distributed photovoltaic systems and in large
central solar thermal power stations. Thus, both distributed rural and centralized urban
demand can be covered by renewable energy technologies.
The accuracy of a global resource assessment of this kind cannot be better than 30
per cent for individual sites as it depends on many assumptions and simplifications.
However, it gives a first estimate of the order of magnitude of the renewable energy
treasures available in Europe and MENA.
ch-08
2/7/07
2:38 PM
Page 267
267
direct normal irradiation can now be generated, as can an hourly time series for every
single site.
The mean bias error of the annual sum of direct normal irradiation, which is decisive
for economic assessment, is usually in the order of 5 per cent. More information
TABLE 8.1 Renewable electricity performance indicators representing the average renewable electricity
yield of a typical facility in each country
5000m
TABLE 8.2 Economic potentials of renewable energy sources in the southern EU and MENA region in
TWh/y
ch-08
2/7/07
268
2:38 PM
Page 268
ch-08
2/7/07
2:38 PM
Page 269
269
GWh/y on 30 x 50 km pixel
serious concern that should be taken into account in energy planning. Efficiency of
hydropower use should be enhanced systematically in order to, at least, partially
counteract such effects.
(1)
The data of the reference plant were taken from Chopra (2003).
ch-08
2/7/07
270
2:38 PM
Page 270
FIGURE 8.4 Temperature at 5000m depth for hot dry rock geothermal power technology
The annual amount of electricity that can be generated from hot dry rock depends on
the heat in place and the time of extraction. That time was assumed to be 1000 years in
order to ensure that the geothermal potentials can be renewed within this time span. At
such a slow rate, the geothermal power potentials can be considered as renewable
energies that could be used continuously without limitations in time like other renewable
energy sources.
In 2000 approximately 600MW of conventional geothermal power capacity was
installed in the analysed countries producing 4.6TWh/y of electricity. The total economic
potential was estimated to be around 400TWh/y; this is, however, a quite rough and
conservative estimate.
ch-08
2/7/07
2:38 PM
Page 271
271
steadily with each year. We have assumed a constant municipal waste productivity of
0.35tons/cap/year and a waste-to-electricity conversion factor of 0.5MWh/ton. It was
estimated that 80 per cent of this potential will be used until 2050.
Solid biomass (mainly wood) potentials were assessed from a global map of biomass
productivity in tons/ha/year and from the existing forest areas of each country (Figure 8.5).
A possible change in productivity or forest areas in the future has been neglected. Results
were cross-checked for plausibility with historical data from European countries (WEC,
2004). There will also be competition with traditional fuel wood use in most MENA
countries which must be taken into consideration. Therefore the rate of use of the fuel
wood potential was assumed to be 40 per cent until 2050. Annual full load hours are used
as performance indicator.
The total installed capacity of biomass power plants in the analysed countries in 2000
amounted to 1.8GW; plants were generating a total of 6.4TWh/year of electricity. For the
total region, a biomass electricity potential of 400TWh/y was identified, of which about 50
per cent might be used until 2050. Potential from residues dominate in MENA, while
power from solid and other biomass sources is also very important in Europe.
ch-08
2/7/07
272
2:38 PM
Page 272
FIGURE 8.6 Annual average wind speed at 80m above ground level in m/s
For the other countries, electricity potentials were estimated from the wind map in
Figure 8.6 taking into account wind speed and area restrictions and site exclusion similar to
that used for CSP but adapted to wind power. The original wind speed was taken from
ECMWF (2002) for 33 and 144m height and was interpolated to 80m height. This map gives
a very rough estimate of the distribution of wind speed as an average for an area of 50 x
50km. The original data has a geographic resolution of 1.12 degrees.
Wind electricity potentials were calculated as functions of the average wind speed
according to well-known equations. We have assumed a maximum installed capacity of
10MW per square kilometre of land area. Areas with annual full load hours over 1400h/y
equivalent to a capacity factor of 16 per cent were considered as long-term economic
potential. Results were cross-checked and eventually corrected for those countries that
have made a national resource assessment (WEC, 2004; OME, 2002; GTZ, 2004; Alnaser
and Al-Karaghouli, 2000; Gral et al, 2003a). Annual full load hours (capacity factor) define
the performance indicator. These have been derived from the literature, the World Wind
Atlas (WWA, 2004) for a selection of sites in each country, and from the wind speed map.
Potentials include both onshore and offshore.
In 2000 a total of 3.3GW of wind capacity was installed in the analysed region producing
7.2TWh/y of wind electricity (ENERDATA, 2004). The total economic wind power potential in
the region amounts to 440TWh/y, of which 285TWh/y could be exploited by 2050.
ch-08
2/7/07
2:38 PM
Page 273
273
expansion is still limited by their high investment cost. Using present growth rates and
scenarios for very large PV systems and distributed applications, PV potentials were
assessed in a relatively intuitive way. For EU states, the literature gives mid-term
potentials for PV (EU, 2004). The global irradiance on a surface tilted according to the
latitude was used as a performance indicator (Figure 8.7) (Meteonorm, 2004). Although
we have not introduced any economic threshold, the learning curves of PV suggest that
this technology will become competitive by the middle of this century under the irradiance
conditions of the MENA region.
FIGURE 8.7 Annual global irradiation on surfaces tilted south with latitude angle in kWh/m /year
ch-08
2/7/07
274
2:38 PM
Page 274
More information on the equations used to estimate the potential of each type of
renewable energy resource can be obtained from Alnaser (1995).
Export
Desalination
MENA
FIGURE 8.8 Domestic electricity demand in MENA and electricity supplied by existing power stations, by
new power plants and by CSP for domestic consumption, export and seawater desalination for the scenario
Closing the Gap
2/7/07
2:38 PM
Page 275
275
100
90
80
70
60
50
40
30
20
10
20
50
20
45
20
40
20
35
20
30
20
25
20
20
20
15
20
10
20
05
20
00
ch-08
Year
FIGURE 8.9 Growth rate of CSP production in MENA in the scenario Closing the Gap including domestic
power supply, export electricity and seawater desalination
expansion. A long-term growth rate of 30 per cent, as achieved in the scenario between
2015 and 2025, is quite realistic and has already been observed in other technologies.
Nevertheless, until 2025 considerable production of conventional and other renewable
electricity of 1440TWh/year will have to be built in the form of new conventional power
plants or other renewable power technologies such as wind, hydro or geothermal power.
Once installed, this new power plant capacity will be there for a time span of about 3040
years, until it will again be replaced.
The scenario assumes that once CSP has reached appropriate production volumes,
the expansion of new fossil-fuel-fired power plants will stagnate. After 2025 most new
plants will be hybrid, using CSP and fossil fuels as input energies (Table 8.3). Part of the
existing fossil-fuel-powered capacity could then be replaced subsequently with
concentrating solar power, if the land area required for the collector fields is considered
in the planning of those power plants. This additional CSP potential has been neglected
except for the islands of Cyprus and Malta.
In our scenario the CSP capacity of each country is proportional to each countrys
share of the total MENA power deficit. Countries showing larger deficits will also require
larger CSP capacities. The total supply of electricity by CSP in MENA will reach 235TWh/y
by 2025 and 2045TWh/y by 2050. It is assumed that the installed CSP capacities would be
distributed evenly among the MENA countries.
In the real world, there might be competition among the MENA countries to achieve
higher growth rates within each individual country. A strong competition among several
CSP technologies could lead to higher overall CSP growth rates than those assumed here
(around 30 per cent per year). In 2025 14 per cent of the domestic electricity demand of
ch-08
2/7/07
276
2:38 PM
Page 276
TABLE 8.3 Electricity demand in MENA and generation by conventional old and new
plants and by CSP technologies in MENA until 2050 (all numbers rounded to 5)
YEAR
Old Plants
New Plants
CSP Supply
MENA Electricity Demand
CSP Desalination
CSP Export
Total CSP in MENA
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
2000
785
785
2025
40
1440
235
1710
115
45
395
2050
0
1560
2045
3605
525
645
3215
MENA will be covered by CSP, which will become the dominating power source with a
share of 57 per cent by 2050.
The necessary solar capacities for seawater desalination are quantified in section 8.7.
A CSP production of 115TWh/y in 2025 and 525TWh/y in 2050 would be used for
desalination purposes. After 2030, the CSP desalination capacities would be large enough
to cope with the freshwater demand and desalination will grow much more slowly. While
in 2025 about 29 per cent of the total CSP production would be used for desalination, in
2050 only 16 per cent would be used for that purpose.
Export of solar electricity will be limited to minor volumes before the scheduled
finalization of the Mediterranean Ring Interconnection in 2015. Also after that, additional
transmission capacities will be required to export considerable quantities of electricity
from MENA to Europe, as the Mediterranean Ring transmission capacity will be limited to
about 500 to 1000MW. The scenario assumes that export through additional high voltage
direct current (HVDC) interconnection of North Africa and Europe will start in 2015 with
2TWh/y. An extended HVDC grid including other MENA regions will transmit 45TWh/y by
2025 and 645TWh/y by 2050, which would be equivalent to about 15 per cent of the
European electricity demand of that time. In the scenario, export capacities are distributed
to all MENA countries in proportion to their share of the total MENA domestic power
demand. The idea behind this is that if a country installs high capacities for domestic
supply, it will also have a good industrial base for export. However, there are different
priorities for electricity export and different starting times for interconnections, depending
on the distance to Europe, the solar energy resources and the priority of the domestic
demand for power and water. By 2025, 11 per cent of the total CSP production of MENA
will be exported, increasing to about 20 per cent by 2050. The islands are not considered
as potential CSP exporters in this scenario. On the other hand, the electricity
interconnection between the Gulf Cooperation Council countries (southwest Asia Saudi
Arabia (SA), Kuwait (KU), Qatar (QR), United Arab Emirates (UAE), Bahrain (BN) and Oman
(ON)) has already been agreed on, with a budget of US$1.2 billion; it started in early 2005
and will be complete in 2008.
The scenario is a rough estimate of the CSP potential in MENA. There will be three
types of plants distributed among all three CSP categories (domestic, export and
desalination) used in different combinations:
ch-08
2/7/07
2:38 PM
Page 277
277
1 CSP plants for cogeneration with coupled seawater desalination must be at the
coast, as the cogenerated heat cannot be transferred over long distances. Their
electricity can be used for additional reverse osmosis (RO) desalination, domestic
use or export. As the coastal regions in MENA are strongly used by other human
activities, this plant type will be limited to regions with appropriate site conditions
and available land area.
2 CSP plants used exclusively for power generation can be anywhere on the grid. Their
electricity can be transmitted to any other place and used for domestic supply, export
or RO desalination. This type of plant will be placed where good irradiation coincides
with good infrastructure conditions.
3 CSP plants for industrial cogeneration will be limited to appropriate industrial sites.
While their heat will be used on site, their electricity might be used on site too or be
sold to the grid for domestic use, export or RO desalination. Cogeneration plants are
considered as part of the domestic CSP production potential.
In the real world, there will be a mix of these three plant types, which will vary according
to the regional demand of each country and the local supply-side conditions. The scenario
can only give a rough estimate of the overall potentials of the region, showing the
amounts of energy potentially used for domestic supply, export or desalination. However,
it cannot distinguish and quantify the different plant types that will be erected in each
country, which will be subject to national strategic power expansion planning.
ch-08
2/7/07
278
2:38 PM
Page 278
TABLE 8.4 Domestic electricity demand in North Africa and generation of domestic
power, export electricity and power for desalination by CSP technologies until 2050
YEAR
Old Plants
New Plants
CSP Supply
North Africa Demand
CSP Desalination
CSP Export
Total CSP in North Africa
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
2000
160
160
2025
10
415
70
490
55
20
145
2050
0
430
800
1230
285
385
1470
FIGURE 8.10 Domestic electricity demand in North Africa and electricity supplied by old power stations,
by new power plants and by CSP for domestic consumption, export and seawater desalination for the scenario
Closing the Gap
ch-08
2/7/07
2:38 PM
Page 279
279
FIGURE 8.11 Electricity demand in western Asia and electricity supplied by old power stations, by new
power plants and by CSP for domestic consumption, export and seawater desalination for the scenario Closing
the Gap
TABLE 8.5 Domestic electricity demand in western Asia and generation of domestic
power, export electricity and power for desalination by CSP technologies until 2050
YEAR
Old Plants
New Plants
CSP Supply
Western Asia Demand
CSP Desalination
CSP Export
Total CSP in Western Asia
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
2000
375
375
2025
20
680
110
810
10
10
130
2050
0
685
905
1590
60
145
1110
ch-08
2/7/07
280
2:38 PM
Page 280
the site conditions in western Asia are not as conducive as North Africa and the domestic
demand in western Asia is very large and will probably be binding the capacities.
On the basis of country statistics from the FAO (2004), a considerable demand for solar
powered desalination can only be detected for Jordan, Israel and Syria. The potential for
desalination will achieve 8 per cent of the total CSP production in 2025 and 5 per cent in 2050.
Due to the considerable growth of demand, domestic electricity production will be the
dominant potential for CSP in western Asia, with comparatively small potentials for
desalination and for export electricity. In 2025 14 per cent of the domestic electricity
demand will be covered by CSP, which will grow to become the dominate electricity
source with a share of 57 per cent in 2050.
FIGURE 8.12 Electricity demand of the Arabian Peninsula and electricity supplied by old power stations, by
new power plants and by CSP for domestic consumption, export and seawater desalination for the scenario
Closing the Gap
ch-08
2/7/07
2:38 PM
Page 281
281
TABLE 8.6 Domestic electricity demand of the Arabian Peninsula and generation of
domestic power, export electricity and power for desalination by CSP technologies until
2050 (rounded to 5)
YEAR
Old Plants
New Plants
CSP Supply
Demand Arabian Peninsula
CSP Desalination
CSP Export
Total CSP Arabian Peninsula
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
TWh/y
2000
235
235
2025
10
345
55
410
50
10
115
2050
0
445
345
790
180
120
645
Due to the considerable growth of demand, especially in Saudi Arabia and Yemen,
domestic electricity production will be the dominant potential for CSP on the Arabian
Peninsula, followed by a large potential for desalination and comparatively small
potentials for export. In 2025 14 per cent of the domestic electricity demand will be
covered by CSP, which will grow to become a very important electricity source with a
share of 44 per cent in 2050.
8.4.4 SUMMARY
From this analysis we can conclude that there are several consequences related to the
scenario Closing the Gap:
Until 2025, the CSP market in MENA will be limited neither by the available solar
energy resources nor by the expected energy demand, but only by the intrinsic
achievable speed of CSP capacity expansion.
Under the assumptions of the scenario Closing the Gap, CSP would cover about 60
per cent of the expected electricity demand for power and water in MENA in 2050,
the rest would have to be covered by conventional power technologies using fossil
fuels. In spite of the considerable expansion of CSP and other renewable energies,
CO2 emissions in MENA would still double by 2050. However, without CSP CO2
emissions in 2050 would become roughly five times those of 2000.
Economic growth and a reasonable growth of the population do not necessarily lead
to an excessive consumption of natural resources if sustainable pathways are taken
immediately. This includes the intensification of energy and water efficiency and the
use of all available renewable energy technologies.
Due to the abandoning of CSP in the beginning of the 1990s in California, about 15
years of development, learning and expansion have been lost. These missing 15
years will lead to a considerable gap in the energy and water supply in MENA during
the 2020s, which must be bridged by non-sustainable resources like fossil fuels and
groundwater reserves, hoping that those resources will be able to carry on long
enough. An expansion of CSP in time could have avoided those gaps. Now, every
year of delay of the introduction and expansion of CSP will lead to more critical
ch-08
2/7/07
282
2:38 PM
Page 282
situations in the MENA countries. The cost of CSP varies widely. It depends on the
producer and consumer ability. Fully exporting this technology, relying fully on
expatriates and foreign manufacturing and services of CSP will make the cost twice
that found in our scenario
Export of solar electricity and of CSP technology can become an important economic
factor in those MENA countries that will be able and willing to use this new, clean
and everlasting natural resource to increase their national income, while their income
from fossil reserves will inevitably decrease.
TABLE 8.7 Annual per capita water consumption in 2000 and projected demand in 2010 for the GCC
countries The poorest in natural water resources
COUNTRY
Bahrain
Kuwait
Oman
Qatar
Saudi Arabia
United Arab Emirates
Total
DOMESTIC
DEMAND (m3)
2000
237
308
54
258
140
340
1337
2010
279
302
76
296
137
352
1442
AGRICULTURAL
DEMAND (m3)
TOTAL
DEMAND (m3)
2000
197
64
467
341
978
573
2622
2000
436
372
521
599
1118
913
3959
2010
215
60
417
349
880
638
2559
2010
494
362
493
645
1017
990
4001
ch-08
2/7/07
2:38 PM
Page 283
283
With relatively large land area available in these countries, it is possible to utilize the solar
power to produce enough water in the GCC (and other Arab states), while the electricity can
be transmitted to Europe, which suffers from heavy air pollution due to burning conventional
fuel, especially in those countries with heavy industry and advanced technology.
Cooperation between MENA and Europe is indeed needed. The investment in solar
water desalination may be very successful. According to certain newspaper reports, the
GCC is planning to spend more than US$40 billion.
Table 8.9 shows the useful electrical equivalent per m3 of distillate (UEPCMD) in kWh,
and the specific electrical energy input per m3 of distillate (SEEPCMD) in kWh/m3 for typical
multi-stage flash (MSF), multiple effect distillation (MED), thermal vapour compression
(TVC), mechanical vapour compression (MVC), reverse osmosis (RO) and electrolysis (ED).
TABLE 8.9 Comparison of energy consumption for six types of water desalination
DESALINATION TYPE
UEPCMD(kWh)
SEEPCMD(kWh/m3)
MSF
16
4
MED
11.5
2.0
TVC
14.56
1.7
MVC
11.5
RO
1.5
ED
accounting for 80 per cent of the world total of 9.8 million m3 per day;
ch-08
2/7/07
2:38 PM
284
Page 284
TABLE 8.10 The available desalination capacity in the GCC countries, in 2000, using different desalination
technologies, in cubic metres per day
DESALINATION
TYPE
FEED
WATER
RO
MSF
VC
MED
ED
Brackish
Sea
Waste
and others
Total
Brackish
Sea
Total
Brackish
Sea
Waste
Total
Brackish
Brine
Sea
Total
Brackish
Total
Total
BAHRAIN KUWAIT
GCC COUNTRIES
OMAN QATAR
SAUDI
ARABIA
65,504
48,217
44,103
3000
5151
13,840
4154
5000
974,201
444,363
1130
70,133
113,721
47,103
297,202 1,468,036
297,202 1,468,036
1635
1635
1135
1272
1200
1135
2470
3350
2418
3350
2418
416,861 1,520,029
UNITED
ARAB
EMIRATES
73,969
58,344
TOTAL BY
TECHNOLOGY
(m3/day)
1,167,082
572,764
71,263
1300
6362
1600 19,590
59,012 103,735
1242
3115 19,590
66,616 103,735
500
3000
2542
6900
4655
3000
2542
6900
5155
56,156
2500
56,156
2500
186,121 532,370 5,111,054 2,184,299
1,811,109
1300
7,858,006
7,859,306
7877
185,572
1242
194,691
1635
1262
18,297
21,204
64,424
64,424
the GCC countries have 9.35 million m3 per day or some 19.4 per cent of the total RO
(280,319m3 per day for GCC countries, representing only 9.5 per cent of the world
total) (ESCWA, 2003).
ch-08
2/7/07
2:38 PM
Page 285
285
This means that the demand of the agricultural sector will be growing proportionally
to population.
The demand of the domestic and industrial sector will grow proportionally to the
GDP, which is calculated for every country adding the population growth rate to the
per capita GDP growth rate.
The efficiency of water use in the agricultural and municipal sector will be increased
from todays country-specific values to a maximum value which depends on the
selected scenario, the water demand growth rate thus becoming lower than the
population or GDP growth rates. Enhanced technologies will additionally de-couple
water demand from economic growth as experienced in Australia in recent decades
(AQUASTAT, 2004; Gleick, 1998).
In the analysis of the power sector, two different economic scenarios have been used to
determine baselines for water demand predictions:
1 The scenario Following Up assumes an average per capita GDP growth rate of only 1.2
per cent for every country from today until 2050. This implies that the relative distance
between the actual GDP/capita (US$-PPP) of the respective country and the US will
remain constant because the GDP of the US at the same time will also be growing by
1.2 per cent. Efficiencies of the agricultural and the municipal water supply system and
the reuse of wastewater increase gradually from the present national performance
values to a future better value of an enhanced system. However, the efficiency
enhancements are limited by the slow economic development. Population growth and
the agricultural sector dominate the water demand growth rates in this case. Decoupling of the water demand from the economic growth by using enhanced water
supply technologies is also limited in this scenario (Gleick, 1998; Heston et al, 2002).
2 The scenario Closing the Gap assumes that the relative distance between the actual
GDP/capita (US$-PPP) in the US and the respective country is reduced to 50 per cent
by 2050 while the GDP of the US at the same time is growing by 1.2 per cent. This
scenario assumes that the MENA countries will by 2050 achieve GDP per capita
values close to that of the European countries. In this case, the industrial and
domestic sectors will dominate the water demand growth. However, efficiencies will
also be increased and a significant de-coupling of water demand and economic
growth as experienced in Australia in the past decades will take place.
The water demand in the MENA region today is made up of 85 per cent agricultural use,
9 per cent domestic use and 6 per cent industrial use. The future demand is calculated
individually for each country and aggregated to the regions of North Africa, western Asia
and the Arabian Peninsula.
Under the assumptions of the scenario Following Up, the share of agricultural water
use will fall to about 80 per cent and the domestic and industrial share will increase to 12
ADVANCES IN SOLAR ENERGY 2007 VOLUME 17 PAGES 261304
2/7/07
286
2:38 PM
Page 286
per cent and 8 per cent respectively. The total water demand will increase from 300 billion
m3/y today to about 510 billion m3/y in 2050. The scenario reflects the influence of
enhanced water management, policies and efficiencies that are of highest priority for a
sustainable water future in MENA, but that are limited by the slow economic growth
within this scenario.
Under the assumptions of the scenario Closing the Gap, the share of agricultural
water use will fall to about 66 per cent and the domestic and industrial share will increase
to 18 per cent and 16 per cent respectively, more and more dominating the water
demand. The total water demand will increase from 300 billion m3/y today to about 540
billion m3/y in 2050 (Figure 8.13). The scenario also reflects the pronounced influence of
enhanced water management, policies and efficiencies, giving them highest priority for a
sustainable water future in MENA, especially since this scenario is oriented towards a
high economic growth.
In terms of water demand, both scenarios are rather optimistic compared to other
scenarios that predict a doubling of demand for 2025. This is determined by extrapolating
600
500
Freshwater Demand
[billion m/y]
ch-08
Industrial
Domestic
400
Sus tainable
Water
300
200
Agricultural
100
0
00
20
05
20
10
20
15
20
20
20
30
25
20
20
Year
35
20
40
20
45
20
50
20
Note: The white line indicates the sustainable renewable freshwater resources.
FIGURE 8.13 Water demand structure in MENA and its evolution until 2050; scenario Closing the Gap
the water demand growth rates as experienced in the last decades. However, we believe
that a reduction of the agricultural sector and the successive de-coupling of economic
growth and water demand are more realistic approaches. At first glance, it is surprising
that both scenarios culminate in a rather similar water demand of 510 to 540 billion m3/y
by 2050, which obviously will be achieved with or without economic growth. It reflects the
positive impact of economic stability and development on the water supply. In the
scenario Following Up, consumption is limited by availability, while in the scenario
Closing the Gap, it is rather limited by the enhanced efficiency of the supply system.
ch-08
2/7/07
2:38 PM
Page 287
287
As the future deficits and the additional demand for non-conventional resources will
not change considerably assuming one scenario or the other, the scenario Closing the
Gap (which is more desirable from the point of view of the MENA countries) will be used
as reference in further analyses.
Western Asia still has large sustainable water resources that will be increasingly
exploited in the future. However, even in this region, non-sustainable use as from fossil
fuelled desalination and from unsustainable groundwater withdrawal is already
experienced on a local level and shows an increasing trend in the future. Unsustainable
water supply from fossil-fuelled desalination and from excessive groundwater withdrawal
is considered as potential future deficits (Figure 8.14).
The sustainable freshwater resources of North Africa are being used today almost to
their limits, and therefore no considerable increase of their exploitation can be expected
for the future. Unsustainable use from fossil desalination and from excessive
groundwater withdrawal is already taking place to a considerable extent, with a dramatic
increase of this situation ahead. On the Arabian Peninsula, the relation of sustainable and
unsustainable use of water is even more dramatic.
The total annual water deficits in MENA will increase from 35 billion m3/y today, at
present supplied by excessive groundwater withdrawals and fossil-fuelled desalination, to
about 155 billion m3/y by 2050. There is no sustainable resource in sight to supply such
deficits except renewable energies. The cost of fossil fuels is already today too high (US$50
in August 2004) for intensive seawater desalination and its volatility and the fact that fossil
fuels are limited in time eliminate fossil fuels as a resource for sustainable water security in
MENA. Nuclear power is a very limited and costly resource, and in addition faces unsolved
problems such as nuclear waste disposal, proliferation and other serious security issues.
FIGURE 8.14 Water supply from sustainable sources and deficits in MENA (Closing the Gap)
2/7/07
2:38 PM
288
Page 288
(a)
The water demand growth rates will decline in all three MENA regions from about 1.5
to less than 1 per cent per year. The per capita water demand and its future trend is
different in the three regions (Figures 8.15a and 8.15b). The MENA average per capita
demand is expected to stay almost constant at about 800m3/capita/y. Western Asia will
reduce its per capita demand from 1000 to about 900m3/capita/y, while the demand in
North Africa will grow from 700 to about 800m3/capita/y, due to a relatively moderate
growth of the population and an increasing importance of the domestic and industrial
sector, mainly in Egypt. The specific consumption on the Arabian Peninsula will fall from
600m3/capita/y (currently) to about 400m3/capita/y in 2050 due to a strong growth of the
3.0
2.5
2.0
North Africa
Western Asia
Arabian Peninsula
Total MENA
1.5
1.0
0.5
0.0
00 05 10 15 20 25 30 35 40 45 50
20 20 20 20 20 20 20 20 20 20 20
Year
1400
(b)
Consumption [m/capita/y]
ch-08
1200
1000
North Africa
Western Asia
Arabian Peninsula
Total MENA
800
600
400
200
0
00 05 10 15 20 25 30 35 40 45 50
20 20 20 20 20 20 20 20 20 20 20
Year
FIGURE 8.15 a) Water consumption growth rates in MENA (Closing the Gap); b) Water consumption per
capita in MENA (Closing the Gap)
2/7/07
2:38 PM
Page 289
289
population and the persisting importance of the agricultural sector, coupled with very
limited natural water resources.
ch-08
250
200
Industrial
Domestic
Agricultural
150
100
50
0
0
20
05
20
10
20
FIGURE 8.16 Water demand structure in North Africa and its evolution until 2050
1000m3/cap/y), and lowest in Algeria and Malta (200m3/cap/y), with a slightly increasing
trend in all countries.
The strong economic growth of the scenario Closing the Gap reveals the challenge
of this path, as the water demand of the industrial and domestic sector will grow very
quickly and overcompensate for possible reductions in the agricultural sector.
2/7/07
290
2:38 PM
Page 290
ch-08
300
250
200
Industrial
Domestic
Agricultural
150
100
50
0
20
00
20
05
20
10
20
15
20
20
20
25
20
30
20
35
20
40
20
45
20
50
Year
FIGURE 8.17 Water demand structure in western Asia and its evolution until 2050
The demand growth rates are high in Jordan but at a very low level of per capita
demand, as can be appreciated from Figure 8.20. Strong consumers are Iraq, Turkey and
Syria, with only Syria facing a short-term deficit. The average per capita demand of the
western Asian region will be slightly reduced from 950 to 850m3/cap/y, while in all
countries the consumption growth rates will be reduced.
2/7/07
2:38 PM
Page 291
ch-08
291
70
60
50
Industrial
Domestic
Agricultural
40
30
20
10
0
20
00
20
05
20
10
20
15
20
20
20
25
20
30
20
35
20
40
20
45
20
50
Year
FIGURE 8.18 Water demand structure for Arabian Peninsula and its evolution until 2050
ch-08
2/7/07
292
2:38 PM
Page 292
Saudi Arabia
1456%
Kuwait
1275%
Qatar
Gaza
Bahrain
Jordan
Yemen
Syria
UAE
Oman
Algeria
Tunisia
Morocco
Cyprus
0%
50%
100%
150%
200%
250%
300%
350%
400%
FIGURE 8.19 Groundwater withdrawals as percentage of safe yield for selected countries
a renewable, sustainable freshwater source with much lower and more stable costs than
fossil fuels is required.
In the present study, we have assumed that the unsustainable water supplied by
groundwater depletion and fossil-fuelled desalination represents a potential future deficit
that could be covered by concentrating solar power (CSP) plants in cogeneration with
multi-effect desalination (MED) and additionally using the remaining electricity for reverse
osmosis (RO) desalination.
To estimate the minimum CSP capacity potential in the water supply sector, we have
assumed that all plants would be coupled to MED desalination plants, while the electricity
generated is completely used for RO desalination in order to produce larger amounts of
desalinated water. In view of the quick increase of future deficits of water in MENA, this
could become necessary in order to substitute for the unsustainable water supply more
quickly. This approach leads to the minimum installed capacity of CSP that is necessary
to cover the future water deficits in MENA.
The capacity potential for CSP would, in reality, be higher, as part of the plants would
be only used for cogeneration of city power and MED desalination, but without RO
desalination. The installation of such plants would be limited to the coast. Another part
would only be used for power generation for RO, but without making use of cogeneration
2/7/07
2:38 PM
Page 293
293
with MED plants. Those CSP plants could be anywhere on the grid, while only the RO
desalination plant must be located at the coast.
To this capacity, the potential of CSP electricity generating and cogeneration plants
outside the water sector may be added. This is described in other chapters of this study.
600
Water Demand and Supply
billion m/y
500
400
CSP-Desalination
Non-Sustainable Supply
Sustainable Supply
300
200
100
45
40
35
30
25
20
15
10
05
50
20
20
20
20
20
20
20
20
20
20
00
0
20
ch-08
Year
FIGURE 8.20 Water demand, sustainable freshwater resources, non-sustainable supply and potential
future supply by CSP via cogeneration by MED plus direct generation via solar electricity in RO plants
ch-08
2/7/07
294
2:38 PM
Page 294
FIGURE 8.21 Covering the future freshwater deficits in MENA by transitory non-sustainable sources and
by CSP plants using MED in cogeneration plus solar electricity for RO
desalination, could amount to a total of 67GW. North Africa (35GW) has the largest
potential for CSP desalination plants, followed by the Arabian Peninsula (24GW) and the
western Asian countries (8GW) (Figure 8.22).
FIGURE 8.22 Capacity potential for CSP-Desalination plants with MED and RO in the three main MENA
regions
ch-08
2/7/07
2:38 PM
Page 295
295
the basis of country statistics, no potential can be detected for Morocco, Malta and
Tunisia, although there may be deficits at the local level.
ch-08
2/7/07
296
2:38 PM
Page 296
TABLE 8.11 Global solar radiation, monthly and annual means (in kWh/m2/day)
CITY
Amman
Aquaba
Abu Dhabi
Mohraaq
Alain
Tunis
Cabes
Alger
Oran
Tamanrset
Riyadh
El-Medina
Taif
Khartoum
Juba
Damas
Almsleamea
Bagdad
Mosul
Mascat
Oman
Jerusalem
Doha
Kuwait
Tripoli
Sirte
Cairo
Aswan
Casablanca
Marrakech
Nouak Chatt
Sanaa
Mocha
JAN
2.7
3.0
4.3
3.6
3.9
2.4
2.9
2.2
2.8
5.2
3.5
4.5
4.4
5.5
5.5
3.1
2.0
3.0
2.0
4.0
4.6
2.7
3.7
3.1
2.9
3.4
5.7
4.6
2.7
3.4
5.7
4.0
4.9
FEB
3.7
4.6
5.0
4.8
4.3
3.1
3.7
3.0
3.7
6.1
4.6
5.4
5.2
6.4
5.5
3.5
3.0
3.8
2.8
4.7
4.7
3.4
4.4
4.1
4.0
4.2
4.0
5.6
3.3
4.2
6.2
4.4
5.4
MAR
5.0
5.9
5.7
4.6
5.3
4.4
4.9
4.1
4.9
6.9
5.1
6.2
5.6
6.8
5.4
4.6
4.1
4.8
3.6
5.5
5.5
5.0
4.9
5.5
5.0
5.0
5.2
6.5
4.5
5.2
7.2
4.8
6.1
APR
6.8
6.7
6.7
6.1
6.6
5.7
5.9
4.9
6.2
7.7
5.5
6.6
5.8
7.2
5.6
5.8
5.3
5.7
4.5
6.3
6.3
6.0
5.7
6.2
6.0
5.6
6.3
7.3
5.4
6.0
7.8
4.4
6.3
MAY
7.8
7.6
7.6
6.9
7.5
6.5
6.8
6.0
6.9
7.8
5.6
7.4
5.8
6.6
5.6
7.4
6.2
6.5
5.5
6.9
6.9
6.9
6.2
7.1
6.5
6.2
7.1
7.4
6.3
6.7
7.7
5.6
6.2
JUN
8.4
8.3
7.6
7.4
7.5
7.3
7.2
6.2
7.4
7.7
6.1
8.2
6.4
6.3
5.2
8.0
7.9
7.3
6.3
6.7
6.7
7.6
6.5
7.9
7.0
6.9
7.6
8.0
6.5
7.3
7.7
5.3
5.8
JULY
8.2
8.0
7.0
7.0
6.8
7.4
7.3
7.0
7.5
7.6
6.1
8.3
6.3
6.1
5.0
7.9
7.7
7.2
6.2
6.1
6.1
7.6
6.0
7.5
7.1
6.9
7.4
7.8
6.4
7.6
7.2
4.7
4.3
AUG
7.5
7.5
6.7
6.3
6.6
6.6
6.5
6.4
6.8
7.4
5.9
8.0
5.9
6.0
5.4
7.2
7.1
6.6
5.6
6.0
6.0
7.2
5.8
7.1
6.5
6.3
6.8
7.4
5.9
7.0
7.1
4.6
4.8
SEP
6.4
6.5
6.5
5.8
6.4
5.3
5.4
5.1
5.6
6.7
5.7
6.8
5.5
6.0
5.9
6.1
6.1
5.7
4.8
5.8
5.8
6.1
5.5
6.2
5.5
5.6
5.9
6.6
5.0
5.9
6.7
5.2
5.3
OCT
4.8
5.1
5.7
4.9
5.7
4.0
4.2
3.3
4.0
6.0
5.3
5.8
5.2
5.9
5.8
4.5
4.1
4.4
3.5
5.2
5.2
4.8
4.8
4.8
4.0
4.1
4.7
5.7
3.8
4.6
6.3
4.0
6.1
NOV
3.6
4.0
4.8
3.9
4.7
2.9
3.2
2.7
2.9
5.2
4.5
4.8
4.6
5.7
5.5
3.0
2.8
3.3
2.3
4.4
4.4
3.5
4.1
3.4
3.1
3.2
3.5
4.8
2.7
3.6
5.5
4.0
5.4
DEC
2.7
3.4
4.0
3.0
4.6
2.3
2.7
2.0
2.4
4.8
3.6
4.3
4.4
5.5
5.3
2.3
2.0
2.7
1.8
3.8
3.8
2.6
3.5
2.9
2.3
3.0
5.9
4.3
2.4
3.2
5.1
3.9
5.1
ANNUAL
5.6
5.9
6.0
5.3
5.8
4.8
5.1
4.4
5.1
6.6
5.1
6.4
5.4
6.2
5.5
5.2
4.9
5.1
4.1
5.4
5.4
5.3
5.1
5.5
5.0
5.0
5.4
6.3
4.6
5.4
6.7
4.6
5.5
ch-08
2/7/07
2:38 PM
Page 297
297
TABLE 8.12 Areas required for renewable electricity generation in 2050 for the scenario CG/HE
9,251
1,648,000
438,317
13,295
21,946
97,740
17,818
10,452
212,457
11,437
2,240,000
185,180
77,700
536,869
2,381,741
18,584 1,002,000
1,775,500
458,730
163,610
131,957
301,302
11,828
13,696
46,798
12,028
13,109
92,389
504,782
779,452
74,018 13,099,653
plants are applied. This would mean winning additional land rather than land
consumption. Photovoltaic energy has no additional land use if installed on roofs, and a
slightly higher land use than CSP if installed in large installations. An average land use of
7km2/TWh/y was assumed. It may seem paradoxical that solar and geothermal power
generation has the best land use efficiency among all power technologies, even when not
considering the potential land gain effect.
The total mix of renewable energies in 2050 within the scenario Closing the Gap/High
Efficiency (CG/HE) has an average land use of 22.5km2/TWh/y, which is in the same order
as the average value of natural-gas-fired combined cycle power stations, which represent
the best available fossil-fuelled power technology. Disposal of sequestrated CO2 is not
considered within this figure. The land use of oil- or coal-fired steam cycles is between 50
and 100km2/TWh/y. Considering the long time during which areas are affected by nuclear
waste disposal and uranium mining, nuclear plants also have a high land consumption in
the order of 100km2/TWh/y. This figure does not account for nuclear accidents like the one
at Chernobyl. The change to renewable energies will therefore lead to more efficient land
use for power generation.
Solar thermal power plants will also be used for seawater desalination. A concentrating
solar thermal collector array required for desalinating 1 billion m3/y would cover a total land
area of approximately 10 x 10km, corresponding to about 10m3 desalinated water per m2
of collector area. In case of linear Fresnel or multi-tower technology, the collectors could
act like blinds, blocking the intense direct solar radiation and creating a cool space
underneath with sufficient light for horticulture or other purposes.
ch-08
2/7/07
298
2:38 PM
Page 298
About 10 per cent of the desalinated water would be sufficient for irrigating the desert
land beneath the collectors with a water column of 1m/y. In 2050, our scenario arrives at
2900TWh/y of electricity (including solar power generation and desalination) and 160
billion m3/y of desalted water. For this a collector field of 120 x 120km2 would be
necessary, which is equivalent to not more than 0.15 per cent of the Sahara Desert.
2/7/07
2:38 PM
Page 299
299
CO2
CO2
CO2
FIGURE 8.23 CO2 emissions of electricity generation in million tons per year for all countries for the
scenario CG/HE and emissions that would occur in a business as usual case (BAU)
CO2
ch-08
FIGURE 8.24 Annual per capita CO2 emissions of power generation (Scenario CG/HE)
For the security of power generation and our planet, the diversity of energy
production, using different sources of energy, should transform from the existing pattern
(Figure 8.27) to that proposed by our scenarios (Figure 8.28).
ch-08
2/7/07
300
2:38 PM
Page 300
FIGURE 8.25 Annual electricity demand and generation within the analysed countries in the MED-CSP scenario
FIGURE 8.26 Installed power capacity and peak load within the analysed countries in the scenario CG/HE
the energy sector (Figure 8.29). Renewable energies in the future will be the least cost
option and may be the only option to obtain sustainable energy and water security in the
MENA region. In most countries, the dependency on energy imports is reduced, opening
new business opportunities for industrial development. In the total EU-MENA region there
may be two million direct and indirect jobs in the renewable energy sector by 2050.
Renewable energies are characterized by their diversity of resources and technologies and
their enormous capacity range from a few watts to hundreds of MW. They can be adapted to
any kind of energy service and closely interlocked with conventional modern energy
technologies in order to provide full power availability and security of supply at any time and
place. Renewable energy technologies fit very well into modern supply systems that are
increasingly relying on distributed generation and network integration, like in virtual power
plants. Furthermore, intercontinental grid connections can effectively combine the different
regional resources to yield the necessary redundancy of supply and address the sustainability
goal of international cooperation (Figure 8.30). Large centres of supply will evolve at sites with
ch-08
2/7/07
2:38 PM
Page 301
301
FIGURE 8.28 Total electricity consumption and share of different technologies for electricity generation in
the analysed countries in 2050 according to the MED-CSP scenario
very abundant, and thus cost-effective, renewable energy resources, providing electricity and
renewable hydrogen to the regions of demand, in other words large urban areas in
industrialized and developing countries, by means of high voltage direct current (HVDC)
transmission and by pipelines respectively (ABB, 2004). At the same time, such centres will
become a regional nucleus of economic development and wealth and will help to stabilize
socio-economic structures. Many of those centres will be established in developing countries,
contributing considerably to the positive progress of our developing world (Knies, 2004).
Using solar energy means manufacturing machines that use renewable energies. It
means replacing minerals from the subsoil by capital goods. Renewable energies require
a large amount of labour on all industrial levels from base materials like steel, glass and
concrete to civil engineering and advanced technology applications. Increased industrial
ch-08
2/7/07
2:38 PM
302
Page 302
FIGURE 8.29 Example of electricity costs and learning in the MED-CSP scenario
activities will create job opportunities and reduce the brain-drain from MENA to the
industrial countries. Considerable shares of the equipment and construction materials of
the solar field and the power block can be produced domestically in many countries with
potential CSP deployment. For parabolic trough systems, an evaluation of the supply
capability of selected countries like Morocco, Spain and Brazil indicates domestic shares
ranging between 40 and 60 per cent for the first plants. Local supply shares can be
increased for subsequent projects if domestic industries adopt increased production of
solar field and power block components.
ACKNOWLEDGEMENTS
The authors wish to thank Prof. Yogi Goswami, University of South Florida, Co-director of
the Clean Energy Research Center, Dr. Russell L. Shoemaker, University of Bahrain, and
Barbara J. Graham, Clean Energy Research Center at the University of South Florida, for
their assistance in editing this paper.
REFERENCES
ABB (2004) HVDC references ABB HVDC classic and light projects, ABB power transmission, www.abb.com/global/
abbzh/abbzh251.nsf!OpenDatabase&db=/global/ gad/gad02181.nsf&v=17EA&e=us&c
=C1256D71001E0037C1256C330030750D
Alnaser, W. E. (1995) Renewable energy resources in the State of Bahrain, Applied Energy, vol 5, pp2330
Alnaser, W. E. et al (2004) First solar radiation atlas for the Arab world, Renewable Energy, vol 29, pp10851107
ch-08
2/7/07
2:38 PM
Page 303
303
FIGURE 8.30 Projection of a future trans-Mediterranean grid interconnecting the best sites for renewable
energy use in EU-MENA
Alnaser, W. E. and Al-Karaghouli, A. (2000) Wind availability and its power utility for electricity production in Bahrain,
Renewable Energy, vol 21, pp247254
Al-Zubari, W. K. (2002) Alternative water resource management policies in west Asia, Industry and Environmental Journal,
vol 25, no 1, pp4347
AQUASTAT (2004) Aquastat, FAOs information system on water and agriculture, online database,
www.fao.org/ag/agl/aglw/aquastat/dbase/index.stm
Bazilevich, N. (1994) Global Primary Productivity Database, unpublished work, Clark University, Worcester, MA, US
Bennouna, A. (2004) Perspektiven der Netzgebundenen Umwandlung Regenerativer Energien in Marokko, Rabat
Bruchmann, U. (2004) Umweltpolitik Geothermie Energie der Zukunft, Bundesministerium fr Umwelt, Naturschutz und
Reaktorsicherheit, Berlin
Chopra, P. (2003) The economics (hot rock energy power plant), Australian National University, Canberra, 30 July 2003,
http://hotrock.anu.edu.au/economics.htm
Dabbagh, T. A. (1995) The role of desalination and water management in sustaining economic growth in the Gulf, International
Desalination Association CIDA, World Congress in Desalination and Water Sciences, Abu Dhabi, 1824 November
DLR (2005) Concentrating solar power for the Mediterranean region, German Aerospace Center, Institute of Technical
Thermodynamics, Federal Ministry for the Environment, Nature Conservation and Nuclear Safety, Germany,
www.dlr.de/tt/med-csp
ECMWF (2002) Re-analysis generated for 15 years (ERA-15) from December 1987 to February 1994, European Center for
Medium-Range Weather Forecasts, London, http://wms.ecmwf.int/research/era/Era-15.html
ENERDATA (2004) GasElecStat 2004 databases, ENERDATA S.A.
ESCWA (2001) Energy Options for Water Desalination in Selected ESCWA Member Countries, United Nations, New York
ESCWA (2003) Review of Sustainable Development and Productivities, no 1, United Nations, New York
EU Commission (1997) Energy for the future: Renewable sources of energy, white paper, COM (97) 599 final
EU (2004) The Share of Renewable Energy in the EU, Commission of the European Communities, Brussels
FAO (2003) Review of World Water Resources by Country, FAO, Rome
FAO (2004) FAO database collection FAOSTAT 2004, http://faostat.fao.org/faostat/collections?version=ext&hasbulk=0
Gardner, P. et al (2002) Wind energy The facts, European Wind Energy Association, Berlin,
www.ewea.org/06projects_events/proj_WEfacts.htm
Gawell, K., Reed, M. and Wright, M. (2004) Geothermal energy: The potential for clean power from the Earth, Geothermal
Energy Association, Washington, www.geo-energy.org/PotentialReport.htm
ch-08
2/7/07
2:38 PM
304
Page 304
Index
2/7/07
2:41 PM
Page 305
Index
electrons, 75
1.5 kW Society, 37
Abdul-Baqi, 13, 38
absorber tube, 1323
accumulated energy, 1523
active wells, 17
adsorption, 139, 145, 147, 14951, 164
advanced oxidation processes, 130, 163
Alaska, 10, 14
Al-GaAs/GaAs, 45, 49, 52, 54, 66, 71
alternative fuels, 37
aluminum reflector, 137
ammonia, 16, 18
Angola, 910
annual discoveries, 14
annual per capita CO2 emissions of power
generation, 299
anti-bonding orbital, 75
application potential, 21718, 220
Arabian Gulf countries, 282
areas required for renewable electricity
generation, 297
artificial photo-charge synthesizers/converters, 95
Association for the Study of Peak Oil, 2, 4, 14,
22, 38
Auger recombination, 49, 73
Azerbaijan, 1112
Azeri-Chirag-Guneshli, 11
background doping, 50, 60
Bakhtiari, 13, 38
Barent Sea, 21
batch mode, 1534
bi-continuous double diamond phase (OBDD), 88
biodegradability, 162
biomass, 36, 305, 37, 39, 414
body-centered cubic lattice (BCC), 88
bonding orbital, 75
bound states, 489
BP, 2, 5, 7, 10, 12, 1920, 22, 25, 39
Bragg reflector, 60
Brazil, 910
bulk heterojunction, 812, 96
calcium carbonate, 1001,11819, 1213, 1259
Canada, 2, 1011, 1718, 39
Cantarell, 9
capacity of CSP plants, 293
capacity potential for CSP, 292, 294
carbon dioxide emissions, 25, 278, 36
carbon emissions, 261, 295, 298
carrier escape, 501, 60, 701, 73
Index
2/7/07
306
2:41 PM
Page 306
Japan, 21, 39
lamellae (LAM), 88
largest oil fields, 910
lattice mismatched, 45, 55, 61
L-H isotherm, 138
light intensity, 142, 147, 164
light scattering, 148, 151
light-trapping, 60, 63
lightweight, 75
limited resources, 6
LNG, 2, 1820
LNG terminals, 1819
logistic function, 29
logistic growth, 1, 29
Kamchagarak, 12
Kara Sea, 21
Kashagan, 12
Kazakhstan, 1112
kinetic constants, 151
kinetic studies, 138, 164
Korea, 21
Index
2/7/07
2:41 PM
Page 307
INDEX
307
PCE or PCE, 87
peak of oil production, 34, 7
peroxydisulfate, 1456
pesticides, 130, 1401, 1501, 158, 160, 1646
phenol, 134, 147, 1656
photo doping, 93
photocatalysis, 136, 138, 142, 1467, 14950,
1545, 158, 161, 1637
photochemical, 132, 137, 146, 166
photo-efficiency, 160
photo-electrochemistry, 16970, 211
photo-electrode, 16970, 1745, 17781, 183,
190, 1945, 198, 2014, 206, 211
photo-Fenton, 1302, 134, 14950, 159, 161,
1636
photon flux, 133, 1467
photon loss, 74, 801, 83, 87, 901, 94
photo-reactor, 1489
photosynthesis, 76
photovoltaic systems, 261, 266
pilot plant, 130, 142, 147, 151, 1536, 1634,
166
p-i-n, 4955, 58, 60, 63, 65, 71
pipeline, 1617, 1920
plant size, 142
pollutant, 1301, 1345, 1457, 14950, 152,
155, 162, 1657
polyacetylene, 77
polyamide, 6, 112, 118, 1201, 127
polybutylene (PB), 112, 125
polymer, 99, 101, 11214, 11729
polyolefin, 11213, 115, 118, 12527
poly-p-phenylenevinylenes (PPVs), 77, 90
polysulfone (PSU), 112, 125
polythiophenes, 77, 90
primary structure, 89
process heat, 21621, 2267, 2312, 234, 236,
2389, 241, 2467, 25860
production profile, 89, 1415, 19, 25, 28, 39
PSS-PEDOT, 80, 846
quantum dot, 66, 68, 70
quantum efficiency, 1367
quantum well, 456, 489, 679, 703
quantum yield, 1446, 161
quasi-Fermi level, 55, 656, 68, 70
R/P-Ratio, 20, 257
radiation resistance, 601
Rayleigh number, 100, 1035, 108, 124
reaction rate, 130, 13949, 1512, 161
reaction rate constant, 13941
reactors, 224, 26
recombination, 45, 4950, 535, 58, 646, 68,
71, 73
recombination quenching parameter (RQP),
924
reduction of emission of greenhouse gases, 298
reflectivity, 133
regio-regular, 90
relative photonic efficiency, 147
renewable electricity performance
indicators, 267
Index
2/7/07
308
2:41 PM
Page 308