Energy Analysis for Sustainable Mega-Cities Aumnad Phdungsilp

Energy Analysis for Sustainable Mega-Cities Aumnad Phdungsilp
Energy Analysis for Sustainable
Aumnad Phdungsilp
Licentiate Thesis
School of Industrial Engineering and Management
Department of Energy Technology
Royal Institute of Technology
Stockholm, Sweden 2006
Energy Analysis for Sustainable Mega-Cities
Aumnad Phdungsilp
Trita REFR Report No.06/52
ISSN 1102-0245
ISBN 91-7178-388-1
Licentiate Thesis by Aumnad Phdungsilp
School of Industrial Engineering and Management
Department of Energy Technology
Royal Institute of Technology
SE-100 44 Stockholm, Sweden
Printed by Universitetsservice US AB
Stockholm, 2006
© Aumnad Phdungsilp, 2006
Cities throughout Asia have experienced unprecedented economic development over the
past decades. In many cases this has contributed to their rapid and uncontrolled growth,
which has resulted in a multiplicity of problems, including rapid population increase,
enhanced environmental pollution, collapsing traffic systems, dysfunctional waste
management, and rapid increases in the consumption of energy, water and other resources.
The significant energy use in cities is not very well perceived in Asian countries. Although a
number of studies into energy consumption across various sectors have been conducted,
most are from the national point of view. Energy demand analysis is not considered
important at the level of the city. The thesis is focused on the dynamics of energy utilization
in Asian mega-cities, and ultimately aims at providing strategies for maximizing the use of
renewable energy in large urban systems.
The study aims at providing an in-depth understanding of the complex dynamics of energy
utilization in urban mega-centers. An initial general analysis is complemented by a detailed
study of the current situation and future outlook for the city of Bangkok, Thailand.
An integrated approach applied to the study includes identification of the parameters that
affect the utilization of energy in mega-cities and a detailed analysis of energy flows and their
various subsystems, including commercial, industrial, residential and that of transportation.
The study investigates and evaluates the energy models most commonly used for analyzing
and simulating energy utilization. Its purpose is to provide a user-friendly tool suitable for
decision-makers in developing an energy model for large cities. In addition, a Multi-Criteria
Decision-Making (MCDM) process has been developed to assess whether or not the energy
systems meet the sustainability criteria.
A metabolic approach has been employed to analyze the energy flow and utilization in
selected Asian mega-cities, including Bangkok, Beijing, Shanghai, and Tokyo. The approach
is applied to measure the majority of indirect energy flows or the energy embodied in the
flows of goods and services involving the residents of those cities. Since the function of cities
is to serve the lives of the residents, indirect energy consumption could be regarded as being
of equal importance as that of direct energy use. The essence of embodied energy is that an
indirect reflection upon behavior following direct energy consumption. It can illustrate how a
city relies on the outside, for example other cities, countries, etc. and provides some
interesting information that cannot be easily drawn from the direct energy demand. The study
reveals that the indirect energy demand is more significant than the direct energy demand in
Bangkok, Shanghai, and Tokyo, while direct energy demand is greater than the indirect
energy demand in Beijing. This can be explained by the fact that Bangkok, Shanghai, and
Tokyo have a greater reliance upon the outside in terms of energy demand.
The Long-range Energy Alternative Planning (LEAP) system has been selected to perform
Bangkok energy modeling. In a Bangkok case study a range of policy interventions are
selected and how these would change the energy development in Bangkok by the year 2025
is examined. Different policies can be grouped by the sectors analyzed. The only supply-side
policy considered meets an existing target of having 10% of electricity generated from
renewable sources. The study period for the model started in 2005 and ends in 2025, with
the year 2000 taken as the base year. The proposed scenarios were evaluated using the
MCDM approach to rate their sustainability. Team members found that this method provided
a methodology to help decision-makers to systematically identify management objectives
and priorities.
Keywords: Asian mega-cities; energy demand; metabolism of cities; energy modeling;
sustainable energy; energy planning; renewable energy; Multi-Criteria Decision-Making
From working on this thesis, I owe many people a great deal of gratitude for supporting my
studies financially, scientifically and also by personal friendships.
First of all, I would like to give my most sincere gratitude to my supervisor
Asst. Prof. Ivo Martinac, for giving me the opportunity to join the Sustainable Building
Systems group, sharing his knowledge, and encouraging me to think over the years even for
though it were for some time communication from afar. His words were not only thought
provoking within the academic arena but also supportive in both my public and private lives.
I wish to express my deep thanks to my principal supervisor (huvudhandledare), Prof. Björn
Palm, for his never ending support, and always being kind and helpful whenever I needed it.
I also want to thank Dhurakit Pundit University, Bangkok, Thailand for funding my
scholarship, and Swedish Energy Agency (Energimyndigheten) for complementary funds.
Experiences from the participation in various conferences financed by the Department of
Energy Technology, Royal Institute of Technology have also been very useful and have
contributed to my research.
I would like to acknowledge, Assoc. Prof. Björn Frostell from the Division of Industrial
Ecology at the Royal Institute of Technology, for his discussions and the further development
of urban metabolism. The Stockholm Environment Institute – Boston for supporting the LEAP
model. Also I gratefully acknowledge Prof. Svend Svendsen from the Department of Civil
Engineering at the Technical University of Denmark and Prof. Anne Grete Hestnes from the
Department of Architectural Design, History and Technology, The Norwegian University of
Science and Technology for introducing me to the Multi-Criteria Decision-Making process.
Prof. Raimo P. Hämäläinen from Systems Analysis Laboratory, Helsinki University of
Technology for his support with the Web-HIPRE software.
As with the Scandinavian scholars, I would also like to thank Prof. Kevin Bennett, Dr. Mark
Howells, Thomas Alfstad, and their colleagues at the Energy Research Centre at the
University of Cape Town, for providing me with research facilities and discussions
concerning the world of energy modeling, valuable comments, fun collaborations, exciting
study visits, and a rich friendly atmosphere during my time in South Africa.
Special thanks to Assoc. Prof. Shiji Kaneko from Graduate School for International
Development and Cooperation, Hiroshima University, Japan, for the calculation of the InputOutput model. Assoc. Prof. Yoshiyuki Shimoda from Division of Sustainable Energy and
Environmental Engineering, Osaka University, Japan, for his discussion on the application of
the metabolism of a city.
I would like to thank the entire staff at the Department of Energy Technology, especially Inga
Du Rietz for her help in many aspects during my studies here. My colleagues and office
workers Lic. Eng. Paulina Bohdanowicz, and Branko Simanic for providing friendly, helpful
and endless discussions. Special thanks to Wimolsiri Pridasawas and Seksan Udomsri, Thai
Ph.D. students at our department for their support and suggestions, especially to Seksan for
his never ending discussions covering scientific, economic and political perspectives.
Most importantly, I am deeply grateful to my parents, brother and lovely sister for the
continuous and unlimited love and support in helping me to achieve my goal. I would also like
to thank my former teacher in Thailand, Dr. Janthana Kunchornrat for her encouragement.
Last but not least, thanks to all the people who are not mentioned here but still are not
forgotten. You will always have a place in my heart.
ABSTRACT.............................................................................................................................. iii
TABLE OF CONTENTS............................................................................................................v
LIST OF FIGURES ................................................................................................................. vii
LIST OF TABLES ....................................................................................................................ix
LIST OF PUBLICATIONS.........................................................................................................x
1. INTRODUCTION ................................................................................................................. 1
1.1. Background of the thesis ............................................................................................... 1
1.2. Sustainability and mega-cities ....................................................................................... 2
1.3. Investigation of Asian mega-cities ................................................................................. 5
1.4. The research questions and conceptual model............................................................. 7
1.5. Research approach and thesis structure....................................................................... 9
1.6. References .................................................................................................................. 11
2. UNDERSTANDING ENERGY USE IN CITIES.................................................................. 13
2.1. Energy, cities, and sustainable development .............................................................. 13
2.1.1. Energy and urbanization ....................................................................................... 14
2.1.2. Determinants of energy use in cities ..................................................................... 15
2.1.3. Urban effects on energy use ................................................................................. 18
2.1.4. International frameworks for cities and sustainable development......................... 22
2.2. Factors driving energy consumption ........................................................................... 26
2.2.1. Urban demographic changes ................................................................................ 26
2.2.2. Patterns of urbanization ........................................................................................ 27
2.2.3. Income growth and social change......................................................................... 29
2.2.4. Structure of economic activities ............................................................................ 30
2.2.5. Lifestyle and the level of consumption .................................................................. 31
2.3. The metabolism of a city ............................................................................................. 32
2.3.1. City as an ecosystem ............................................................................................ 32
2.3.2. Review of the metabolism of a city........................................................................ 34
2.3.3. Evaluating urban metabolism methodologies ....................................................... 36
2.3.4. Strengths, weakness, and extending of the metabolism modeling ....................... 40
2.4. References............................................................................................................... 42
3. ENERGY SYSTEM MODELS AND ENERGY DEMAND ANALYSIS................................ 50
3.1. Modeling the energy system ....................................................................................... 50
3.1.1. Classification of energy system models ................................................................ 51
3.1.2. The selected model in this thesis .......................................................................... 56
3.1.3. LEAP model framework ........................................................................................ 57
3.2. Estimation of urban energy use................................................................................... 63
3.2.1. Energy demand estimation ................................................................................... 63
3.2.2. Representing an urban energy system ................................................................. 66
3.3. Direct and indirect energy demand analysis................................................................ 69
3.3.1. Model for indirect energy consumption based on embodied energy analysis....... 70
3.3.2. Calculation method for embodied intensity based on Input-Output (IO) analysis . 72
3.3.3. Comparative study on indirect energy use of selected Asian mega-cities ............ 77
3.4. References .................................................................................................................. 80
ENERGY SYSTEMS ............................................................................................................. 83
4.1. Overview of multi-criteria decision-making .................................................................. 83
4.1.1. Weighted sum method (WSM) .............................................................................. 86
4.1.2. Weighted product method (WPM) ......................................................................... 86
4.1.3. Analytical hierarchy process (AHP)....................................................................... 86
4.1.4. Preference ranking organization method for enrichment evaluation (PROMETHEE)
........................................................................................................................................ 87
4.1.5. The elimination and choice translating reality (ELECTRE).................................... 88
4.1.6. The technique for order preference by similarity to ideal solutions (TOPSIS)....... 88
4.1.7. Compromise programming (CP)............................................................................ 89
4.1.8. Multi-attribute utility theory (MAUT) ....................................................................... 89
4.2. Developing of multi-criteria decision-making in energy systems assessment ............. 90
4.2.1. A framework of the multi-criteria decision-making process ................................... 90
4.2.2. Selected main criteria and sub-criteria .................................................................. 91
4.2.3. Generating energy scenarios ................................................................................ 93
4.2.4. Performance prediction ......................................................................................... 94
4.2.5. Weighting of criteria and aggregating scores ........................................................ 95
4.3. The Web-HIPRE decision-support software ................................................................ 96
4.3.1. The analytic hierarchy process method................................................................. 97
4.3.2. The multi-attribute value theory method ................................................................ 98
4.3.3. Combined use of the methods .............................................................................. 99
4.3.4. Group decision-making.......................................................................................... 99
4.3.5. Analyzing the results ........................................................................................... 100
4.4. References ................................................................................................................ 102
5. CASE STUDY: CITY OF BANGKOK, THAILAND ........................................................... 106
5.1. Analysis of the energy situation ................................................................................. 106
5.1.1. Energy organization in Thailand .......................................................................... 106
5.1.2. Energy supply and demand analysis in Thailand ................................................ 107
5.1.3. Energy use pattern in Bangkok ........................................................................... 109
5.2. Direct and indirect energy consumption .................................................................... 110
5.2.1. Data specification ................................................................................................ 110
5.2.2. Direct and indirect energy consumption in Bangkok ........................................... 111
5.3. Bangkok energy modelling ........................................................................................ 113
5.3.1. Model of energy demand and supply .................................................................. 113
5.3.2. Energy policies and scenarios............................................................................. 124
5.3.3. Results and discussion of the scenarios ............................................................. 126
5.4. Energy decision with multi-criteria decision-making .................................................. 134
5.4.1. Weighting of criteria............................................................................................. 134
5.4.2. Lessons learned .................................................................................................. 139
5.5. References ............................................................................................................. 141
6. CONCLUSIONS AND FUTURE WORK .......................................................................... 143
6.1. Conclusions ............................................................................................................... 143
6.2. Future work................................................................................................................ 144
Appendix 1. Estimated Bangkok energy demand by economic sectors under BAU scenario
............................................................................................................................................. 146
Figure 1.1. Research conceptual model. ................................................................................. 8
Figure 2.1. Growth of urban agglomerations in selected mega-cities in 1950-2015.............. 14
Figure 2.2. Trend in GDP per capita in selected Asian countries .......................................... 29
Figure 2.3. The movement of chemicals and materials through the natural ecosystem........ 33
Figure 2.4. The movement of chemicals and materials through a system resulting from
human activity...................................................................................................................... 33
Figure 2.5. Comparison of urban metabolisms...................................................................... 35
Figure 2.6. The extended metabolism model of the city ....................................................... 41
Figure 3.1. LEAP calculation flows ........................................................................................ 58
Figure 3.2. Example of the tree structure in the energy demand module.............................. 59
Figure 3.3. The general structure of transformation modules................................................ 60
Figure 3.4. Example of a wood resources structure. ............................................................. 61
Figure 3.5. An example of a costing boundary in cost-benefit analysis................................. 62
Figure 3.6. A physical representation of the RES.................................................................. 66
Figure 3.7. Example of part of the RES ................................................................................. 68
Figure 3.8. Energy balance for sector j on a city scale .......................................................... 70
Figure 3.9. Direct and indirect energy consumption of selected Asian mega-cities .............. 78
Figure 4.1. Multi-criteria decision process ............................................................................. 85
Figure 4.2. The overall structure of criteria and sub-criteria use. .......................................... 93
Figure 4.3. An example of a star diagram with indicated individual criteria scores and
reference scores .................................................................................................................. 95
Figure 4.4. An example of value tree window of the residential case.................................... 97
Figure 4.5. An example of the AHP weighting window .......................................................... 98
Figure 4.6. An example of the overall values of the alternatives ......................................... 100
Figure 5.1. Comparison of total primary energy supply and domestic production of primary
energy.................................................................................................................................. 107
Figure 5.2. Energy intensity in Thai economy during 2000-2004 ........................................ 108
Figure 5.3. Final energy consumption by sectors in Thailand. ............................................ 109
Figure 5.4. Estimation of energy consumption by sectors in Bangkok ................................ 109
Figure 5.5. The estimated total direct and indirect energy demand in Bangkok for 1998.... 112
Figure 5.6. Total commercial energy used by building types in 2000.................................. 114
Figure 5.7. Energy consumption of industrial sub-sectors in 2000. ..................................... 116
Figure 5.8. Energy consumption of residential sector in 2000............................................. 118
Figure 5.9. Energy consumption of transport sector by vehicle types in 2000. ................... 120
Figure 5.10. Energy savings in commercial buildings.......................................................... 127
Figure 5.11. Avoided CO2 emissions from commercial buildings. ....................................... 127
Figure 5.12. Final energy demand of the industrial sector under different scenarios. ......... 128
Figure 5.13. CO2 reductions in the industrial under industrial policies................................. 129
Figure 5.14. Trend of energy savings in residential sector. ................................................. 130
Figure 5.15. Avoided CO2 emissions from residential policies. ........................................... 130
Figure 5.16. Trend of energy savings under the transport modal shift scenario.................. 131
Figure 5.17. The transport CO2 reductions under the modal shift scenario......................... 131
Figure 5.18. Local pollutant reduction in the transport sector.............................................. 132
Figure 5.19. Avoided CO2 emissions with renewable energy target.................................... 133
Figure 5.20. Results of main criteria weighting.................................................................... 134
Figure 5.21. Total weighted scores for the alternatives in the residential sector. ................ 135
Figure 5.22. Star diagrams for residential energy policies................................................... 135
Figure 5.23. Total weighted scores for the alternatives in the commercial sector. .............. 136
Figure 5.24. Star diagrams for commercial energy policies................................................. 136
Figure 5.25. Total weighted scores for the alternatives in the industrial sector. .................. 137
Figure 5.26. Star diagrams for the industrial energy policies............................................... 137
Figure 5.27. Total weighted scores for the alternatives in the transport sector.................... 138
Figure 5.28. Star diagrams for the transportation energy policies. ...................................... 138
Figure 5.29. A star diagram for the alternative on the supply-side. ..................................... 139
Table 1.1. Ten largest cities in the world in 1950, 2000, and 2015 ......................................... 3
Table 1.2. Population in Asian urban agglomerations of more than 3 and 5 million inhabitants
in 1990 ..................................................................................................................................... 6
Table 1.3. Summary of selected Asian mega-cities investigation............................................ 6
Table 3.1. Energy model characteristics................................................................................ 54
Table 3.2. A summary of models used in energy-environmental planning ............................ 55
Table 3.3. Energy demand disaggregation and the estimation methods............................... 65
Table 4.1. The design main criteria and sub-criteria.............................................................. 92
Table 4.2. Example of measurement scale for annual electricity use.................................... 94
Table 5.1. Comparison estimated breakdown direct and indirect energy demand in
Bangkok in 1998. ................................................................................................................. 112
Table 5.2. The number of projected designated buildings in Bangkok ................................ 115
Table 5.3. Average breakdown of annual energy consumption and percentage share of
energy services per building ................................................................................................ 115
Table 5.4. Proportion of energy utilization in industrial sub-sectors .................................... 116
Table 5.5. Average useful energy intensities in industrial sub-sectors ................................ 117
Table 5.6. The number of projected households in Bangkok Metropolitan.......................... 119
Table 5.7. Energy intensity of the appliances used in the residential sector ....................... 119
Table 5.8. Lower limits of persons per vehicle..................................................................... 122
Table 5.9. Comparison of model-based forecast and actual number of registered vehicle
in 2000 ................................................................................................................................. 122
Table 5.10. Estimated travel demand for road transportation in Bangkok........................... 123
Table 5.11. Fuel economy of the automotive technologies by fuel types in Bangkok
Metropolitan Area ................................................................................................................ 123
Table 5.12. Description of energy policies and interventions for sustainable energy
development in Bangkok. .................................................................................................... 125
Table 5.13. Final energy demand by sectors in BAU scenario............................................ 126
Table 5.14. The energy demand by fuel types in BAU scenario.......................................... 126
Conference papers (Peer Reviewed):
Phdungsilp A., Martinac I. (2004), A Multi-Criteria Decision-Making Method for the
Retrofitting of Designated Building in Thailand. In Proceedings of the 21st Conference
on Passive and Low Energy Architecture, September 19-22, 2004, Eindhoven, The
Phdungsilp A., Martinac I. (2004), Sustainable Cities Cooling: A Case of District
Cooling System in Thailand. In Proceedings of the 1st International Solar Cities
Congress, November 14-18, 2004, Daegu, Republic of Korea.
Phdungsilp A. (2005), Towards sustainable urban energy use in cities: a metabolism
approach. In Proceedings of the 2005 World Sustainable Building Conference,
September 27-29, 2005, Tokyo, Japan.
Journal papers:
Phdungsilp A., Bangkok energy modelling for the promotion of sustainable city energy
development, Manuscript submitted to Energy Policy.
Phdungsilp A., Energy decision with a multi-criteria decision-making for a sustainable
energy planning in cities, Manuscript submitted to Energy Policy.
Other publication:
Phdungsilp A. (2006), Integrated Energy and Environmental Modelling. Lecture notes
in Energy and Environment Course, International Master Programme in Sustainable
Energy Engineering, Department of Energy Technology, Royal Institute of Technology,
Stockholm, Sweden.
1.1. Background of the thesis
Cities have been considered as the centers of cultural, social and technological innovation,
economic development, inventions and their applications, and also the political power of a
particular country. After the industrial revolution, urban areas in different parts of the world
grew rapidly due to their increased populations. The most important growth of urban areas is
the formation of so-called mega-cities or the city of human settlements of about or above 10
million inhabitants. These cities are the magnets for people, functions, organizations, and the
structuring of the country and the world around their social and economic dynamics. They
always interact on local and global levels. Therefore, the impact of mega-cities on the natural
environment is being felt far beyond their geographic boundaries.
Mega-cities are not considered to be just cities of a large size. They have a strong internal
coherence. They are not a juxtaposition of different areas. Most often they are fundamentally
one single area. They consist of a complex unit of production for consumption, a large labor
market, a specific system of power and extreme social and cultural differentiation. Functions
and activities performed within their territories are spatially interconnected. Their territories
are used daily by millions of people within its boundaries. As for the rest of the country, it
increasingly becomes the hinterland for the functions and power that emerge from megacities (Castells 1993).
The previous decade has seen Asia become the key region in need of environmental
protection. The region has become increasingly vulnerable to environmental damage and
ecological disasters caused by rapid urbanization, population growth, and the expansion of
industrial activities. As a result of the growing urban population and its increased levels of
consumption (i.e. energy, water, and other resources), negative effects have been brought
upon the environment.
A significant increase in the consumption of energy and other resources as well as the
contribution of Greenhouse Gas (GHG) emissions is expected to take place in cities within
Asia, especially in mega-cities. These cities that have rapidly expanding populations are
enjoying higher living standards and more material affluence than people in rural areas and
smaller cities. The problems mega-cities face today will be similar to those that smaller cities
face in the future. Therefore their actions may provide models for other cities (IGES 2004).
Hence, a study of mega-cites may provide a good basis for governments to create
comprehensive action strategies to promote sustainable development. However, the
strategies will differ from city to city and no single universal strategy will work in all situations.
Cities in rapidly industrialized regions of Asia face many tasks related to economic and
environmental issues. So far, the energy use and emissions are not well understood. Urban
authorities are largely not aware of the multiple benefits of energy management and carbon
dioxide (CO2) reduction. The literature has shown that there has been limited research on the
management of energy within the cities. The studies were not considered important at the
city level until recently, and most energy related decisions were made at the national level
(IGES 2004). Due to growing concerns about climate change, efforts are being made to
understand the use of energy at the city level in greater detail. It is important to integrate
energy analysis with urban development. The efforts should be directed towards supporting
cities in understanding the problem and identifying possible measures for implementation.
The priority for a systematic approach is an analysis of the energy flow into the city, the
driving forces behind energy use, and associated policy measures. In this regard, it is
imperative to establish a comprehensive study at the city level and to prepare appropriate
measures for responding to specific local needs and conditions.
An appropriate response to a city’s energy and environmental problems requires an analysis
of past and present situations, formulation and implementation of remedial measures,
development and transfer of technologies and enhancement of human capacity. Although, at
the national as well as at the city level, authorities wish to learn from a sensible practical
model in order to improve their local development, there has been little progress due to the
limited exchange of well-analyzed information. There is a need for the relevant analysis of
experiences, successes, and failures in urban energy use from different cities, providing a
systematically arranged policy that is truly useful for national and city authorities in
developing their policies, plans, and programs (IGES 2004). In order to provide relevant
information for policy-makers at different levels of government and agency, information
gained from the studies should be properly analyzed and should provide examples of policy
alternatives or demonstrate policy guidelines for effective measures that are generally
1.2. Sustainability and mega-cities
The term mega-city was created by the United Nations in mid-1980 in a study addressing
issues generated by rapid urbanization and a growing population as experienced by cities.
The common issues relating to urban energy use include transportation, building and
housing, public health and safety, and an increase in the standard of living. Therefore, the
development of mega-cities and their effects upon future human society cannot be
In 1800 only 3% of the world’s population lived in cities later in 2000 more than half of the
world’s population lived in urban areas. Big cities are growing at unprecedented rates and
sizes. UN projections show that by the year 2015 there will be 36 mega-cities and 23 of them
in Asia alone (UNPF 1997). UN data shows that mega-cities are growing faster than any
other size of city (Salas 1986; UN 1995). In the past 30 years the urban population in
developing countries has tripled, and projections suggest that the urban growth will continue
into the next century. Urban areas in Asia, Africa, and Latin America are growing at an
average rate of one million people per week (UNDP 1995; 1996).
In the process of urbanization, the urban concentration has been shifting from industrialised
nations to developing ones. Table 1.1 shows the top ten largest cities in 1950, 2000, and
2015. Today the world has 22 mega-cities, and 18 are in developing countries, mostly in Asia
and Latin America. Estimation was predicted that from 1950 to 2050 the urban population in
the Third World countries will increase almost 16 times from under 200 million to 3,150
million people (UN 1985; 1995; 2002). So far, most of the discussion regarding mega-cities
has focused on their problems such as air pollution, modes of transport, etc. rather than on
seeing the city as a large consumer of energy, natural resources, goods and services, for
example Girardet 1996; Rogers 1997; Moavenzadeh et al. 2002.
The concept of sustainable development has been applied to the so-called sustainable city.
However, this concept is difficult to define precisely because it refers to process rather than
to end-point. Cities can vary in their definitions and classifications depending on the issues
being considered. In fact, there are several terms that apply to a city, most often they refer to
conceptual, physical and administrative boundaries, including urban area, rural area,
municipality, metropolitan, city, etc. In some cities the buffer zone, such as sub-urban/peerurban, can be included within the definition.
Furthermore, to analyze the city it cannot be looked at as a single and self-contained. A city
requires a flow of natural resources, services, people, information, etc. for its different stages
of development. These flows are often uncontrollable and in many cases have benefits but in
others problems such as local pollution, traffic congestion, and waste water (SEI 2003).
Table 1.1. Ten largest cities in the world in 1950, 2000, and 2015 (UN 2002).
Population (millions)
New York, USA
London, England
Tokyo, Japan
Paris, France
Moscow, Russia
Shanghai, China
Essen, Germany
Buenos Aires, Argentina
Chicago, USA
Calcutta, India
Tokyo, Japan
Mexico City, Mexico
Bombay, India
Sao Paulo, Brazil
New York, USA
Lagos, Nigeria
Los Angeles, USA
Calcutta, India
Shanghai, China
Buenos Aires, Argentina
Tokyo, Japan
Bombay, India
Lagos, Nigeria
Dhaka, Bangladesh
Sao Paulo, Brazil
Karachi, Pakistan
Mexico City, Mexico
New York, USA
Jakarta, Indonesia
Calcutta, India
In general, sustainable development is defined as a constraint upon present consumption in
order to ensure that future generations will inherit a resource base (or a set of opportunities)
that is no less than the inheritance of the previous generation. Moavenzadeh et al. (2002)
argued that sustainability concerns issues of both technology and equity. However, there are
many more definitions. The following section gives an overview of how sustainable cities are
defined by different organizations in different geographical regions.
In Sweden, the Swedish Research Council for Environment, Agricultural Science and Spatial
Planning (Formas) defines the sustainable city as a society in which economic development,
social welfare and cohesion are united with a healthy environment. The society satisfies its
current needs without jeopardizing the ability of coming generations to satisfy theirs (Formas
2004). This vision is formulated in a governmental paper, which is based on the Brundtland
The Regional Environmental Center for Central and Eastern Europe (REC) refers to the
sustainable city as one that aims to ensure adequate availability of resources, (re-) utilization,
social comfort, equity, economic development and prosperity for future generations (REC
2003). They have also provided a list of some important issues that must be taken into
consideration, including minimizing urban sprawl, use of renewable or recyclable resources,
minimizing material output and its ecological effects, democracy, etc.
The URBAN 21 Conference in Berlin developed the definition of a sustainable city. They
defined sustainable urban development as improving the quality of life in a city, including
ecological, cultural, political, institutional, social and economic components without leaving a
burden on future generations (URBAN 21).
The concept of sustainable cities was defined by Development Alternatives in India in
response to the conditions in India. They said that a sustainable city means one that is able
to provide basic needs for the populace along with the necessary infrastructures of civic
amenities, health and medical care, housing, education, transportation, employment, good
governance, etc. It should take care of the population’s needs, all sections of society without
discrimination (Development Alternatives, India 2003).
In South America Fundacion Ambiente y Recursos Naturales (FARN) which is a non-profit
organization in Argentina whose mission is to promote sustainable development. FARN
defines a sustainable city as a city that integrates environmental dimensions into the social
and economic sectors in order to meet the needs of present generations without
compromising those of the future. Thus, a sustainable city requires institutions and systems
that can facilitate public participation in decision-making regarding the use and management
of the environment (FARN 2003).
The Sustainable Cities Programme which is a joint facility of the United Nations Centre for
Human Settlements (UN-Habitat) and the United Nations Environment Programme (UNEP),
has defined a sustainable city as a city where achievements in social, economic, and
physical development are made to last. A sustainable city has a lasting supply of the natural
resources on which its development depends, (the use of a sustainable yield) and a lasting
security from environmental hazards which may threaten development (although allowing for
acceptable risk). The planning and management for a sustainable city development requires
agreements and coordinated action by a variety of public, private, and popular sector
involvement at the individual, community, city, and national level (UNDP 2003).
In general, sustainability is a much more subjective concept. It refers to a set of social goals
like improving the welfare system and the quality of the environment, increasing individual
freedoms, having a more even distribution of wealth, and improving the health and education
systems. Sustainability is not only concerned with the continuing ability to increase income
through economic growth, but also the ability to achieve other more qualitative goals. Cities
can be seen as focal points for achieving sustainability. They offer many opportunities. From
an energy point of view, the energy efficiency of cities both in transport and building sectors
has great potential. The provision of many services is easier and more efficient in an urban
area. Theoretically and technically it is possible to have very effective transport system
without cars (Register 1987; Richards 1990; Girardet 1996; Roelefs 1996; Rogers 1997;
Roseland 1997).
The challenge of achieving the goal of a sustainable city is not a technological problem, but it
will need a change in politics, economics, and social relations, as well as an integration of
these aspects into the system (Hopwood 2001). The sustainable city includes an integrated
system of co-ordination and cross-sector work between agencies, and developmental
activities related to a holistic approach in its design and management. So far, policies
attempting to adopt the concepts of sustainability within a city have not gone far enough into
the root of the issue. To achieve this it would require dramatic changes in the governmental
process of decision-making concerning both social and economic policy.
Mega-cities already represent extreme or more intense manifestations of the urbanization
profile. However, mega-cities will continue to grow, both in their size and in their
attractiveness as a location for the lives and functions of a populace. The goal of achieving
sustainability is worth striving for and will require a continuation of the present small steps, as
well as the incorporation of large strides.
1.3. Investigation of Asian mega-cities
If the 20th century was the century of urban sprawl, then the 21st century will be the century of
mega-cities. Asian mega-cities are at the center of the region’s economic growth and
development. They are growing rapidly, not only in terms of population but also in terms of
economy and industry. Throughout Asia there is an increasing concentration of urban growth
in large cities. The reasons for this concentration include: the effects of large sale economies
and forces of agglomeration, wider opportunities for employment, the effect of the global
economy, and the relatively greater availability of social, health and cultural services. Much
of this growth is taking place in areas around the built-up city as a mixture of residential,
industrial, and commercial developments spread over the surrounding areas. Bangkok is a
good example of this.
In the year 2000, about 30% of the population in the Asian region lived in cities. There are
contradictions in the rate of growing urbanization. Japan was the first rapidly urbanize and
was followed by South Korea. During the 1960s over 79% of the population in these
countries lived in urban areas. In Southeast Asian countries urbanization started slowly but
has grown faster since the 1970s. The South Asian region, including Bangladesh, India,
Nepal, and Pakistan, has experienced a gradual rise in urbanization. The growth rate of the
population has been intensive in some major cities such as Mumbai, Calcutta, Dhaka, and
Karachi. Since 1960 the urban populations have been rising at a rate of approximately 3 to
5% annually (World Bank 2003).
Asia now has more major cities than any other region in the world. Table 1.2 shows the
distribution of larger cities in Asia compared with the world’s total. China and India, the most
highly populated countries, have the largest number of major cities.
Table 1.2. Population in Asian urban agglomerations of more than 3 and 5 million inhabitants
in 1990 (World Bank 2003; UN-HABITAT 2001).
Sub Total
World Total
More than 5 million
Number of cities
Population (Mill.)
More than 3 million
Number of cities
Population (Mill.)
Due to limited research and data on Asian mega-cities, this thesis considers a few selected
cities i.e., Bangkok, Beijing, Shanghai, and Tokyo. Analysis of these cities can be used to
generate wider perspectives on other cities in Asia. Table 1.3 summarizes the investigation
of Bangkok, Beijing, Shanghai, and Tokyo, respectively. These cities share the
characteristics of having a high population density and of being the most important cities in
their respective countries. They differ in terms of a number of factors such as their levels of
income and development, forms of governance and institutional capacities.
Table 1.3. Summary of selected Asian mega-cities investigation.
A combination of a
modern and a
historical city.
Suffers from
systems and an
undeveloped mass
transport network.
Large consumer of
transforming with
a growing
population, many
new buildings,
and increasing
automobile traffic.
Tokyo’s model for
transport but with
a phase-lag.
The richest megacity in China.
developing. New
business facilities,
automobile traffic.
Most developed
mega-city in Asia.
A modern urban
infrastructure, a
mass transport
system, and a
number of new
Modal split of
Majority of travel
by private cars
and buses. Since
1998, the mass
transit system i.e.
Skytrain, subway,
was developed.
Oil, electricity, and
gas dominated.
Rapidly rising
number of cars’
shared since
1990. Increasing
in the number of
Majority of
transportation by
bus. Highest level
of non-motorized
Majority of travel by
rail and subway.
Bus is insignificant.
Heavily, coal
Heavily, coal
Oil, electricity, and
gas dominated.
Moderate to high
Fuel structure
efficiency of
1.4. The research questions and conceptual model
Cities throughout Asia have experienced an unprecedented economic development over the
past decades. In many cases, this has contributed to their rapid and uncontrolled growth, and
has resulted in multiple problems, which include rapid population increase, enhanced
environmental pollution, collapsing traffic systems, dysfunctional waste management, as well
as a rapid increase in the consumption of energy, water, and other resources.
Considering sustainable development in relation to sustainable cities, as mentioned earlier,
different interpretations and various aspects have been defined. However, they do not give a
clear definition that can be easily implemented in practice. This thesis focuses on energy use
in Asian mega-cities, and sustainability is interpreted as a future where the energy system is
based upon renewable resources and energy efficiency. The energy system is characterized
by both sides those of supply and demand.
The study focuses on the dynamics of energy utilization in Asian mega-cities, and ultimately
aims at providing strategies for maximizing the use of renewable energy in large urban
systems. The study aims at providing an in-depth understanding of the complex dynamics of
energy utilization in urban mega-centers. An initial general analysis will be complemented by
a detailed study of the current situation and a future outlook for the city of Bangkok, Thailand.
At present there is a lot of research and development regarding energy systems and
sustainable cities. The problem is, however, that much of the research is not done in a
holistic manner. Research institutions, development agencies, and organizations address
each issue individually in a singular way. For instance, agencies involved with the building
sector focus on certain technologies, while others deal with pollution, transportation,
environment, or social issues. An integrated approach seeks to incorporate all the important
aspects into a holistic synthesis. In this approach, all of the different criteria will be focused
on simultaneously and tradedoff against each other in order to optimize the overall efficiency
of the design.
Figure 1.1 shows the research concept model. The study begins with understanding energy
flow in Asian mega-cities, and reviews studies from international research teams to quantify
the energy demand and supply in selected Asian mega-cities, including Bangkok, Tokyo,
Shanghai, and Beijing. At this stage, the major drivers affecting the energy demand in the
process of dynamic structural changes are also identified. A city’s metabolism has been
studied to distinguish the direct and indirect energy consumption by using Input-Output (IO)
analysis. The research into metabolism aims to understand the physical flows into, within,
and out of the cities. Direct and indirect energy demand can be a proxy to emphasize how
cities rely upon the outside in terms of energy demand.
Energy Flows in
Empirical Study (I)
Direct/indirect energy
Tool: I-O table
Estimation of
Energy Demand
Numerical Study (II)
Tool: energy modeling
Bangkok Energy
Futures: policies
& scenarios
Empirical Study (III)
Tool: MCDM
Decision Support
Figure 1.1. Research conceptual model.
The second stage is to estimate the energy demand in various sub-sectors (Bangkok case
study); the Long-range Energy Alternatives Planning (LEAP) system has been used to
estimate energy demand, as well as to analyze energy scenarios in Bangkok. The
backcasting approach is applied to create the city’s energy scenarios to give an image of
future energy systems. According to Dreborg (1996), backcasting means that the
development of scenarios is dictated by starting in the future. Backcasting is used to develop
images of the future where the result describes a possible and desirable future for
sustainable energy systems in city. This method has been applied in a number of energy
future studies, particularly in the study of complex long-term sustainability problems.
The last stage is the Multi-Criteria Decision-Making (MCDM) method, which has been
developed in order to evaluate the city’s energy scenarios. The intent of MCDM is to improve
the quality of decisions that involve multiple criteria by making the options more explicit,
rational, and efficient.
1.5. Research approach and thesis structure
The following study items are included within the thesis:
Identification of the parameters that affect the utilization of energy in mega-cities.
Detailed analysis of the energy flows in mega-cities and their various sub-systems,
including the transportation, industrial, residential, and commercial sectors.
Evaluation of the energy models most commonly used for analyzing and simulating
energy utilization.
Analysis of the potential for renewable energy utilization.
Development of strategies for renewable energy use.
Development of a tool for decision-making.
Case study: city of Bangkok, Thailand.
In regards to the study items, the following methodologies are employed:
A review of the studies from international research groups and clarification of the work
in each group, in order to find what kind of aspects within sustainable cities they are
working on.
Cooperation with other organizations, including the Stockholm Environmental Institute
(SEI), Sweden, the Institute for Global Environmental Strategies (IGES), Japan, for
data sharing and methodology development; and the System Analysis Laboratory,
Helsinki University of Technology, Finland, for the application of MCDM.
A comprehensive analysis of data and methodology development from literature, as
well as research institutes to quantify the energy supply and demand, as well as the
drivers affecting energy consumption.
The collection of energy and economic data at national and city levels. Also a
collection of complementary data from various research reports. This is crucial for
energy modeling because in order to construct a model it is a question of what kind of
data is available.
An investigation into the city’s metabolism and a survey of the methods used for
calculation to analyze the relationship between the activity levels of major urban
sectors and energy demand.
Development of energy modeling for Bangkok is done in cooperation with the Energy
Research Centre at the University of Cape Town, South Africa.
The surveying and development of a method for multi-criteria evaluation used within
energy planning.
The application of all methods to the case study.
This thesis is organized into two parts, a study of the literature, followed by a development of
the methodologies and the application of the methodologies to the case study. The aim of the
first part is to present an overview of a selection of Asian mega-cities and the relationships
between sustainable development and mega-cities focusing on energy utilization. The
second part discusses the development of methodologies, including the metabolism of cities,
energy modeling and MCDM. The objective is to find the tools for a holistic approach and
then to use them to analyze and build a city’s energy scenarios and policies. The overall goal
of this study is to identify the characteristics of Asian mega-cities within the perspective of
energy utilization and to develop a user-friendly tool of an integrated approach for the cities'
Chapter 1 discusses the background to the thesis and the relationship between sustainable
development and mega-cities. The research approach and conceptual model is explained.
Selected Asian mega-cities are investigated.
Chapter 2 examines the understanding of energy flows in a large urban system, of the
surveying of international frameworks of cities and of sustainability. Factors driving energy
consumption and the metabolism of a city are described. A review of cities’ metabolisms is
identified and described.
In Chapter 3 the estimation of energy demand, both direct and indirect, is calculated and
described. A comparative study on the analysis of the embodied energy of selected Asian
mega-cities is discussed, as well as a survey of the energy modeling that is widely used in
energy planning.
Chapter 4 deals with the multi-criteria decision-making (MCDM) in the assessment of the
sustainability of energy systems within the city. It starts with a review of MCDM in energy
planning and the development of a method for multi-criteria evaluation to be used with
energy systems.
Chapter 5 shows the testing of methods, including embodied energy analysis, energy
modeling and MCDM in the city of Bangkok as a case study.
Chapter 6 includes conclusions and future work based upon the case study.
1.6. References
Castells M. (1993), Why the Mega-Cites Focus: Mega-cites in the New World Disorder.
Mega-Cities 7th Annual Coordinators Meeting, August 1 - 7, 1993, Jakarta, Indonesia.
Development Alternatives, India (2003), What is a Sustainable City? Available at, as accessed 15.05.2003.
Dreborg K.H. (1996), Essence of backcasting. Future 28 (9), pp.813-828.
FARN (2003), What is a Sustainable City? Available at, as accessed
Formas (2004), State of the Art Sustainable Urban Development in Sweden. The Swedish
Research Council for Environment, Agricultural Science and Spatial Planning, Stockholm,
Girardet H. (1996), The Gaia Atlas of Cities: New directions for sustainable urban living.
Gaia Books Ltd., London.
Hopwood B. (2001), Sustainable Cities: Phrases or Practice? The 31st International Making
Cities Liveable Conference, October 2001, San Francisco, USA.
IGES (2004), Urban Energy Use and Greenhouse Gas Emissions in Asian-Cities: Policies
for a Sustainable Future. Institute for Global Environmental Strategies, Japan.
Moavenzadeh F., Hanaki K., Baccini P. (2002), Future Cities: Dynamics and Sustainability.
Kluwer Academic Publishers.
REC (2003), What is a Sustainable City? The Regional Environmental Center for Central
and Eastern Europe. Available at, as accessed 14.05.2003.
Register R. (1987), Ecocity: Berkeley. Berkeley, North Atlantic.
Richards B. (1990), Transport in Cities. Architecture Design and Technology Press,
Roelefs J. (1996), Greening Cities: Building Just and Sustainable Communities. Bootstrap
Press, New York.
Rogers R. (1997), Cities for a Small Planet. Faber & Faber, London.
Roseland M. (1997), Eco-City Dimensions. New Society Publishers, Gabriola Island, BC
Salas (1986), Super Cities of the Future. New Scientist, May 1986, pp.27.
SEI (2003), Characteristics of a Sustainable City. Stockholm Environment Institute,
Sweden. Available at, as accessed 14.05.2003.
Steen P., Dreborg K-H., Henriksson G., Hunhammar S., Höjer M., Rigner J., Åkerman J.
Kommunicationsforskningberedningen, KFB-rapport 1997:7.
UN (1985), Estimates and Projections of Urban, Rural and City Populations 1950 – 2025.
United Nations, New York.
UN (1995), World Urbanization Prospects. United Nations, New York.
UN (2002), World Urbanization Prospects: The 2001 Revision. Population Division,
Department of Economic and Social Affairs, United Nations Secretariat, New York.
UNDP (1995), Choices. Vol. 5, No. 1, United Nations Development Programme, New York,
UNDP (1996), Human Development Report 1996. United Nations Development
Programme, New York, pp.25.
UNDP (2003), The Sustainable Cities Programme. United Nations Development
Programme. Available at, as accessed
UN-HABITAT (2001), The State of the World's Cities 2001. United Nations Centre for
Human Settlements (UN-HABITAT), Nairobi, Kenya.
UNPF (1997), The state of world population 1996: changing places: population,
development and the urban future. United Nations Population Fund, New York.
URBAN21 (2003). Available at, as accessed 14.05.2003.
WECD (1987), Our Common Future. World Commission on Environment and
Development, Oxford University Press, Oxford.
World Bank (2003), World Development Indicators 2002. World Bank, Washington DC.
2.1. Energy, cities, and sustainable development
Sustainable urban development is closely linked to patterns of energy use and urbanization.
Today, almost 75% of the population in industrialized countries lives in urbanized areas. The
number of people living in urban areas is rapidly increasing worldwide. In 1950, about 30% of
the world’s population lived in urban areas. The figure has increased to 47% in 2000 and is
expected to reach 60% by 2030 (IGES 2004). Cities contribute towards promoting both local
and global sustainability. They are centers of high living standards and they consume large
amounts of energy, materials, and other resources, which lead to the over utilization of
limited natural resources and large increasing volumes of GHG.
However, urban situations differ between continents, regions, countries, and even within the
same country. Sustainable energy systems are poised to meet growing energy demands,
while simultaneously addressing energy poverty, energy security, and environmental
improvement. An estimate shows that trillions of dollars will be invested worldwide within the
next two decades in order to develop and upgrade global energy systems, with nearly 1
trillion dollars per year required in developing countries. But investments in sustainable
energy systems are still a small fraction of the total investment in the energy sector that
continues to rely on fossil fuels (CEG 2004). The creation of sustainable energy systems is
intimately linked to environmental issues. In addition to the misery and suffering associated
with the affects of ill health, fossil fuels damage the economy, leading to direct costs and
losses in productivity.
As one of the key drivers in economic development and a main source of environmental
problems, energy is an essential component of a society. It transforms materials and labor
into useful goods and services. There is a strong relationship between the quantity of energy
used by a country and the size of its economy. However, this relationship is not necessarily
linear in form, so with the right policies may be able to decrease the amount of energy
needed in order to produce the Gross Domestic Product (GDP). For example, China has
shown that it is possible to achieve economic growth without a corresponding jump in GHG
emissions. Its CO2 emissions fell by 6-14% between 1996 and 1999, while its economy grew
by 22-27%. In contrast with the U.S. situation, the CO2 emissions during the same period
increased by approximately 5% (CESA 2004). Energy is a complex element of sustainable
development policies and the connection between energy and other issues, such as the
social aspects, becomes even more complex.
From the thermodynamics point of view, the availability of energy resources can easily
determine, for instance, how much food is grown and cooked or how a living space is heated
or cooled. Moreover, the quality of energy is as crucial as its quantity. Solid, liquid, and
gaseous fuels vary in their ability to provide energy services depending on their applications
and end-use devices. Coal can be a higher quality energy resource than wood, while oil has
a higher quality than coal, and electricity has a higher quality than solid, liquid, or gaseous
Within an urban settlement there is an increasing acceptance amongst engineers and
specialists from other disciplines that current modes of human existence are unsustainable in
terms of environmental, social, and economic perspectives. There are many indicators to
support this view such as global climate change, resource depletion, droughts, floods, local
pollution, deterioration in the quality of life, etc. Therefore, it is a challenge to build a
sustainable society by focusing on urban settlements. To study sustainable urban
development, we may consider it as inputs and outputs of than urban system, including its
energy use, the recycling, treatment and disposal of its waste as well as its manpower and
knowledge. More specifically, a model can consider the energy and environmental impacts of
the commercial, transportation, industrial and residential sectors. Details of an inputs and
outputs analysis will be described within the study of a city’s metabolism.
2.1.1. Energy and urbanization
It is well recognized that the bulk of the world’s energy consumption is within cities, and
much of the rest is used for producing and transporting goods and people to and from cities.
It is thus crucial to develop strategies for the use of sustainable energy. The major factors
influencing energy use in cities include patterns of urban settlement, transportation systems,
incomes and lifestyles, the energy efficiency of technologies, industrial processes, building
technologies, climate and methods of waste disposal. Some factors, such as income and
lifestyle changes, are particularly significant in Asian mega-cities e.g., Bangkok, Beijing, and
Shanghai. Significant changes in energy use can be achieved by a combination of energy
efficiency and more efficient energy supply systems. There are many good examples,
particularly from Scandinavian countries, of how energy efficiency combined with efficient
supply systems can dramatically reduce urban energy use.
From 2000 to 2030, virtually all population growth is expected to occur in urban areas and
mostly in less developed regions of the world (UN 2002). Figure 2.1 shows the growth of
urban agglomerations in selected mega-cities. New York and London are typical mega-cities
in more developed countries that developed in the 1800s and early 1900s, reached their
current size in the middle of the last century, and have since experienced slow growth or
decline. Mega-cities in some developing countries, such as Mexico City, grew rapidly
between 1950 and 1980, and are growing more slowly now. Many Asian and African megacities, such as Lagos and Bombay, are experiencing very rapid growth and are projected to
continue at this pace.
Population (millions)
ic o
Figure 2.1. Growth of urban agglomerations in selected mega-cities in 1950-2015
(UN 2002).
The potential for urban growth in Asia is tremendous. It is estimated that the population
increase of cities in developing countries will be from 37% to over 54% by 2030 (UN 2002).
This means that 2.6 billion people will live in Asian cities. The number will exceed twice the
current population of China and represent 53% of the world urban population by 2030
(ECOASIA 2001). Predictions show that in 2015 there will be 358 cities worldwide with a
population of over a million people, of them 153 are expected to be in Asia (UN-HABITAT
2001). The sustainability of these cities will have enormous implications.
Cities are one of the key element contributions towards global sustainability. A high
population density and massive consumption open up several options to effectively utilize
natural resources and promote efficient urban infra-structural development. For example,
compact settlement and a high population density in a city may reduce per capita the
infrastructure and distribution costs and open up opportunities for economies of a large
scale. This can be seen in Hong Kong, which is a very densely populated city. Thus, cities
can greatly facilitate the implementation of measures, which reduce energy use and stress
upon the environment. Cities and their energy consumption bring two major environmental
issues to the forefront for policy-makers. The first is the massive consumption of energy and
materials that affects natural systems and, ultimately, areas and peoples outside the
boundaries of cities and even the next generation of residents. The second is the exposure
of a large and concentrated urban population to worsening air, water, and waste pollution.
2.1.2. Determinants of energy use in cities
Energy has long played a central role in the development and function of the economy. It is
an essential input for economic activities. Reliance on energy will obviously continue to grow
with increase in populations and improvements in the standards of living, especially in cities.
History has shown that increased mechanization and energy use brings its own burdens
upon the environment.
As regards urbanization, it is no doubt that energy has no conflict with it in regards to the
population pattern and type of infrastructure. Cities open up many opportunities for the
efficient utilization of energy. European and North American cities stand in contrast to Asian
cities, which tend to be more dense and expansive. European cities are likely to be less
dense and typically have more efficient transportation and energy distribution systems. North
American cities are of a low density and sprawl over large areas, thus requiring large
amounts of energy to run the transportation and distribution systems; their public
transportation is cost ineffective.
Energy use concerns are not only related to availability or to use but also to implications,
which follow the energy use. Fossil fuel produces local air pollutants, GHG, and waste. If
pollutant concentrations reach an unacceptable level then there are serious problems
concerning human health and the environment. The efficient utilization of energy and the use
of renewable resources are the keys to conserving conventional energy resources, as well as
to reducing the levels of air pollutants such as nitrogen oxide (NOx), sulfur dioxide (SOx),
carbon monoxide (CO), carbon dioxide (CO2) and particulate matter (PMs). The following
sub-sections describe several key factors that determine the nature of energy use in cities. Pattern of urban settlements
The compactness of urban settlements influences the energy demand in regards to
transportation and for other energy utilization areas such as district heating and cooling for
space conditioning. Another factor related to urban planning is the urban spatial structure
and functions. They all affect energy use as they influence the demand for the mobility of the
urban dwellers. Mixed land use, for example residential and industrial or residential and
commercial, results in a different energy use. Urban zoning policy and the relocation of
industry from city centers to peri-urban areas in some Asian cities significantly influence
travel demand and, thus, energy use. Similarly, the patterns of energy use in commercial or
service cities are different from those in industrial cities. Income level and lifestyle
As mentioned earlier, mega-cities are the engines for the economic growth of a region or
even country. Castells (1993) argued that mega-cities will continue to grow both in their size
and attractiveness for the location of the functions and lifestyles of people because they are
the centers of economic, technological, and social dynamism. They actually are the
development engines of their own countries. Mega-cities are the connecting points of a
network on which everybody relies on in the new economy.
It is often difficult for policy-makers to constrain increasing energy consumption, especially in
rapidly industrializing cities such as Bangkok, Kuala Lumpur, and Shanghai. Studies on the
relationship between income and energy use at the national level have clearly demonstrated
that there is a strong correlation between per capita energy consumption and GDP. It is
generally accepted that energy use per capita increases with income, but at some point it
may decrease (IGES 2004). A higher income is associated with a better lifestyle and more
consumption, which affects energy use within and beyond the borders of cities.
Consumption-oriented lifestyles are associated with the greater utilization of natural
resources, as well as the greater energy use needed to produce goods and materials;
consequently, they create more GHG emissions. This relationship is important in the
sustainability debate as it seeks to provide welfare and equity along all scales. Energy efficiency technologies
Energy efficiency is defined as the energy needed to produce a unit of energy service. Since
energy demand is for service purposes, energy efficiency can directly improve or worsen
energy consumption. Energy technologies dominate consumption within a city, including the
automobile along with household and commercial appliances used for lighting, heating,
cooling, and cooking. Improving energy efficiency is a function both of the technology itself
and of the patterns of energy utilization.
There is significant untapped potential for improving energy efficiency in almost all energy
applications. Improving energy efficiency is often highly profitable, but there are still some
barriers that remain. Technological improvements will only be effective if there is a
commitment by society to limit the total throughput of matters and energy at a sustainable
level (Huesemann 2003). The very first step towards sustainability would be a public
discussion of complicated and controversial value-laden issues. However, citizens are most
likely led to believe that technological solutions alone will be sufficient to guarantee
sustainability. The use of new technologies in the city can make everyday life easier and
deliver better social and cultural services than there ever was before. Industrial processes
These issues are related to energy efficiency technologies; however, separating the
discussion aims to give an overview of disaggregated sectors, including the transportation
and building sectors. More details about energy demand in different sectors will be discussed
in Chapter 3 (energy system models and energy demand analysis).
The energy efficiency of all industrial production processes affects the energy use in the city
too. Asian mega-cities are rapidly relocating their primary industries either to peri-urban
areas or to areas outside the cities' borders, leaving cities increasingly dominated by the
tertiary sector. Small and medium-sized enterprises (SMEs), however, remain prevalent in
many Asian cities and pose a big challenge for energy planners. The role of local policymakers in this area is limited in South and Southeast Asia (IGES 2004). Another relevant
issue is the type of industry there is within the considered city. For example, heavy industries
and car manufacturing industries consume more energy than electronic and agricultural
industries. Transportation systems
The energy consumed by the transport sector is influenced by its systems. Transportation
systems are very important, for mobility is a key aspect of urban life. Historically, cities have
moved from non-motorized to rail-based and then gradually to automobile-dominated
Energy utilization within the transport sector depends upon a number of factors. For
example, the infrastructure for rail and road networks, the mass transportation systems, the
share of public and private travel demands, the role of alternative fuel vehicles, and the share
of small-occupancy vehicles such as two-wheelers as used in Bangkok, Jakarta, Delhi, etc.
(IGES 2004). Increasingly, cars dominate transportation due to rising incomes, an increased
social status through car ownership, and inefficient public transportation systems. These
trends seem to be typical problems in Asian cities.
Improper transportation systems threaten energy security and result in the increase of local
air pollution and GHG emissions. Traffic congestion reduces the quality of life in the city and
wastes time as well as energy. This problem is often found in big cities in South and
Southeast Asia. The design, placement, and density of buildings in an urban environment
also have a great influence upon the consequent transportation patterns. Building technologies and climate
The technologies in the building sector, for example air-conditioning, district heating and
cooling systems, insulation systems, and other energy management building systems, have
a significant effect on the use of energy. Energy services, including space conditioning and
lighting, are the most dominant form of energy use in the building sector and depend directly
upon floor spaces, whose use depends on a number of factors such as real estate market
prices, business culture and socio-cultural factors.
The climate characteristics of the cities (cool, temperate, hot-arid and hot-humid) have
greatly influenced energy use in building as well. Hot and cold climate conditions directly
affect energy use due to the greater demand for space conditioning. North Asian cities, such
as Beijing, require more energy than cities of more temperate climates. Buildings in
Southeast Asian cities require space cooling all year round. Cities in East Asia, such as
Tokyo and Seoul, suffer from urban heat island and require both heating and cooling. The
literature has shown that the heat island effect may be beneficial to relatively cooler cities
because of the reduction of their heating demand in winter. However, the penalty in summer
has far exceeded the winter gains in cities, which experience hot and humid summers (EC
2.1.3. Urban effects on energy use
This section reviews some of the central aspects of major urban environment in relation to
their energy use. A city changes its landscape from the ecosystem and therefore is
consequent to the change in the relationship between biological and physical aspects of the
environment. City dwellers are exposed to more kinds of toxic chemicals in higher
concentrations and to more human-produced noise, heat, and particles than are their rural
neighbors. According to Akbari et al. (1992), there are four basic environmental goals for a
city, namely, reducing energy use, reducing and removing pollutants, helping to create a
pleasing environment and aiding biological conservation. The following sub-sections describe
all the common urban influences the use of energy has in Asian cities. Urban heat island
An urban heat island is a phenomenon in which the core urban temperature is higher than
the land surrounding it. Studies show a direct correlation between the density and population
of a city and the intensity of the heat island effect. Higher urban temperatures increase the
demand for electricity, which is needed for cooling, which in turn leads to an increase in the
production of CO2 and other pollutants if the electricity is fossil fuel based. The urban heat
island is not a problem in cities situated at high altitudes, since the effect enables a reduction
in the demand for heating. This could be true in winter, but studies have shown that these
benefits are far out-weighed by costs incurred during summer (Landberg 1981; EC 2000).
Heat island effects have been reported in dense and highly urbanized cities all around the
world. The urban heat environment is worsening mostly in Asian mega-cities regardless of
the stages of development or income levels. For example, Bangkok, Shanghai, and Tokyo
are becoming warmer and warmer every day.
Man-made changes to the urban environment are the source of the urban heat island
phenomenon. The radiation balance within the urban system is disrupted as surfaces absorb
long-wave radiation and are unable to radiate it back out. An increase in anthropogenic heat
discharge, a decrease in surface evaporation, changes in the thermal characteristics of
urban surfaces, an increase in traffic and air pollution, together with the reduction in airflow
and humidity caused by the sheltering effect of buildings, are the major factors behind these
changes (EC 2000). Mega-cities are characterized by having a population of a high density
and a high level of energy consumption per capita. The energy demand is fulfilled through
electricity and the combustion of fossil fuels, which ultimately discharge heat into the urban
Direct heat discharge sources are usually categorized as stationary or mobile. The single
greatest source of stationary heat discharge from buildings comes from air-conditioning units.
These units are very densely concentrated throughout cities, for example in the inner
Bangkok area. Some waste incineration plants and industries located in cities, such as in
Tokyo, release heat directly into the urban environment.
The most important mobile source of heat is automobiles. In city centers and high traffic
zones, the concentration of this discharged heat is increased by congestion and fuel
inefficient vehicles, of which Bangkok is a good example. A closer look at a mid-size urban
city reveals that only 13% of the total energy input into transportation is converted into useful
work. The rest is dissipated as heat into the environment of the city (Moavenzadeh et al.
Rapid urbanization and population growth in mega-cities has resulted in the build-up of
massive infrastructures and dense settlements. With the development of the urbanized world
vegetated land surfaces are converted into concrete and asphalt. A change in the nature of
surfaces has primarily affected solar reflectivity, so-called albedo. This parameter is different
from normal reflectivity in the sense that reflectivity might only account for visual bands,
whereas albedo accounts for all the incoming radiation to a surface. Asphalt roads, concrete
pavements and corrugated roofs, which form the major part of dense cities, have a low value
of albedo. For instance, asphalt has 0.05-0.20, and concrete has 0.10-0.35 (Bretz et al.
1998). Low albedo surfaces absorb significant proportions of solar radiation and contribute to
the worsening of the urban heat environment.
In addition, loss of vegetation inhibits the evapotranspiration process, in which plants use
heat from the air to evaporate water from their leaves. This process enables vegetation to act
as a heat sink. Changes in wind patterns have also exacerbated the urban heat problem.
The formation of an urban canopy changes the wind pattern, preventing wind from entering
and blowing heat away from ground surfaces, and traps heat inside the canopy. The
cumulative effects of the above factors make the urban environment hotter than surrounding
areas. Air pollution
Air pollution is a persistent and pervasive urban environmental problem that imposes
significant health and economic costs. The impacts of the urban system on the air can be
added to the problem of the urban heat island, as mentioned earlier. Asian mega-cities, such
as Bangkok and Beijing, experience similar and serious air pollution problems. In addition,
the number of cities with very poor air quality continues to grow, including many cities in
developing countries.
The main pollutants reducing air quality are SOx, PM, NOx, CO2, O3, lead, heavy metals, and
volatile organic compounds (VOCs). The sources of pollution are urban transport, the
generation of electricity and heat from non-renewable energy resources, and industrial
activities. Air pollution may be worsened by the local topography, wind direction and weather
conditions. Many cities around the world have undergone and continue to undergo rapid
growth. Yet, in the absence of adequate environmental management measures in many
cities, this growth is happening at a considerable and often increasing cost. More people,
more industry, and more motor vehicles produce ever-worsening air pollution, an already
serious environmental threat in many cities.
The grave consequences of air pollution for public health are measured not only in terms of
human sickness and death, but also in terms of a loss in productivity and missed educational
and other human developmental opportunities. It is because of these health costs,
productivity losses and degradation of air quality that the city's economic growth and
contribution to national development is hindered. In addition, air pollution imposes significant
additional operating costs on businesses, industries, households and public services.
Likewise, it accelerates the deterioration of buildings and historic monuments, and a
reputation for bad air pollution certainly deters investments from outside the city. Apart from
its local urban effects, air pollution has a profound regional and global impact as well. Urban
air pollution is a major contributor to the problems of ozone layer depletion, global warming
and climate change. To meet these challenges at a global level it is required that urban air
pollution be controlled and reduced.
The technical aspects of urban air pollution are reasonably well understood and the
necessary technologies are generally available. Urban populations are aware of the nature of
the air pollution from which they suffer and are increasingly unwilling to let it worsen (WHO
1996). In many cities, both the attitudes of the populace and of the official have changed and
there is a growing political commitment to the need for change. To convert these more
favorable conditions into actual improvements in air quality there is a need to create a
management system, an urban air quality system, which would be capable of addressing the
complex issues. In addition, the system should ensure the formulation and implementation of
realistic and effective strategies, as well as action plans that will systematically address the
short and long-term causes of urban air pollution and help the city achieve a sustainable
growth path. Microclimate
Cities convert land from its natural state in order to accommodate buildings, houses, roads
and other infrastructures. This tends to modify the albedo, hydrologic and wind patterns,
which in turn affects the microclimate. These affect the microclimate of the building in relation
to comfort as well as that of the energy use in buildings for heating, cooling and lighting.
Wind velocities in the city are generally lower than in the surrounding countryside, which is
due to buildings obstructing the flow of air. The wind affects temperature, rate of evaporative
cooling and plant transpiration, and is thus an important factor at a microclimatic level. Builtup areas with tall buildings may lead to complex air movements through a combination of
wind channeling and resistance. This often results in wind turbulence in some areas and
concentrated pollution where there are wind shadows.
As regards thermal comfort, moderate air movement has a cooling effect and is usually
desirable because it helps disperse local atmospheric pollutants. Air movement also helps to
remove heat from the human body’s surface, although increasing wind velocities can
significantly reduce thermal comfort. Therefore, the planner should consider how wind
direction and wind velocity will affect existing and proposed buildings, especially when
designing high-rise buildings that are popular in East and Southeast Asian mega-cities.
Vegetation is the most obvious and easiest way to improve the urban environment with
respect to air movement. Trees help in a number of ways: they reduce wind speed, which is
undesirable, but the net effect depends upon the location and the type of tree. However, the
effect of vegetation in the urban environment depends upon its proportion compared to the
whole area of the city. In a dense and highly built-up city, the area available for vegetation is
limited and its effect on the heat environment could be minimal. Also, its effectiveness
depends on the conditions of the local climate, whether it is hot and dry or hot and humid. Transport and road traffic
Transport and traffic congestion has a significant effect upon the quality of the urban
environment, the quality of life, time and energy consumption. Private cars, trucks, and other
motor vehicles dominate most Asian cities. It is estimated that a private vehicle typically
consumes more than twice of the energy per passenger-kilometer than a train, and almost
four times that of a bus (Whitelegg 1993). The energy, as well as pollution, implications of an
urban layout that does not maximize the potential of public transport are likely to be very
A study has shown that cities of a high density, for example Hong Kong, have far lower
transport energy demands per capita than low density cities. On average, when comparing
10 major cities in the U.S. with 12 European cities, the latter are five times as dense but the
U.S. cities consume 3.6 times as much transport energy per capita (Newman & Kenworthy
1989). The conclusion often drawn from such data is that dense cities are low energy cities in
terms of energy demand for the transport sector. In the absence of an effective integrated
public transport system, increasing the density will inevitably increase traffic and pollution.
The challenge is to anticipate and implement public transport systems before the increases
in population density to ensure that the increasing existing and future inhabitants use public
transport instead of private cars and thus shift the balance away from the car (Steemers
In Asia, many cities have implemented measures at different stages, ranging from end-ofpipe interventions to more upstream measures. The measures vary from command and
control to those that are market based. Cities often focus on how to organize transport
demands into a better modal structure, a step requiring the integration of urban planning and
land-use policies together with transportation and environmental planning. Policy-makers are
concerned about the share of private transportation both in meeting demand and in its
contribution to pollutant concentrations. End-of-pipe approaches, such as setting emission
standards, improving fuel quality, and implementing vehicle technology interventions, are
limited to vehicles and their increased use. While such measures are necessary they are not
sufficient. A long-term solution to the environmental and congestion problems of urban
transportation introduced in Asian cities requires additional steps.
Solutions to urban transportation issues are to a great extent linked with the city and
transportation planning. Asian cities lack proper city planning and the growth of urban
centers has been haphazard. Factors influencing urban transportation vary from country to
country depending on its economic situation. Population growth and urbanization, as well as
inadequate infrastructures are considered to be the major determinants of transportation
problems in Asian urban areas. There has been a considerable increase of motorization in
almost all of the countries in Asia with the few exceptions being in Central Asian cities (World
Bank 2004). For example, in Bangkok the number of vehicles grew more than sevenfold
between 1970 and 1990, whereas in Beijing it was a threefold increase between 1991 and
2000. Similar trends in growth occurred in Jakarta and Kuala Lumpur (Ramanathan 2000).
Congestion is the common mark of motorization in most growing Asian cities such as
Bangkok, Delhi, Jakarta, Manila, and Seoul. They are particularly congested with weekday
peak-hour traffic speeds reported to be on average 10 km per hour or less. One estimate
puts the average travel time for commuters in Asia at 42 minutes. This number can be much
higher as in the case of Bangkok, where the average is estimated to be about 60 minutes
(UN-HABITAT 2001; World Bank 2004). A study on Bangkok estimated that the direct
economic costs of congestion could be as high as US$3.71 billion annually. A recent World
Bank study estimated that a 10% reduction in peak-hour trips in Bangkok would provide
benefits of about US$400 million annually (UN 2001). Food supply
Food supply is an important issue regarding urban management and sustainable
development. It may not be directly effected by energy use but it can be considered as being
influenced by air pollution and GHG emissions from the energy sector. A sustainable city
should be self sufficient both in terms of energy and food supplies. Also taking food
production into consideration is a crucial factor for life in the city.
Shanghai can be cited as an example of how mega-cities can be organized in terms of food
production. In the Shanghai region two factors have contributed to its recent increased food
production. One is the transfer of the control of the production, distribution and marketing of
food from many diverse operating units to the Shanghai municipal government. The other is
the modernization of farming achieved by the means of mechanization, electrification, waterconservancy work, and the large-scale provision of modern inputs. Spatially, two zones are
distinguished as being for the purpose of urban agriculture. The suburbs near to the inner
zone immediately surrounding the built-up area are devoted to year-round vegetable
production and the outlying suburbs are where coarse and hardy crops are grown. Thus,
Shanghai is totally self-sufficient in regards to vegetables, most grains, and significant
proportions of pork, poultry, and other foods (Yeung 2000).
In other cities, the trend of increased food production is offset by the accelerating loss of
suburban farmland. In Beijing, the vegetable land of 607,000 hectares in 1949 dwindled to
427,000 hectares in 1980, a 30% loss, which resulted from the state taking land from nearby
suburbs for new housing and factories (Smil 1984). Similarly, Bangkok has lost suburban
farmland due to real estate development and increase of industrial areas.
2.1.4. International frameworks for cities and sustainable development
Recent global summits, including the United Nations Conference on Environment and
Development (1992), the United Nations Conference on Population and Development
(1994), the World Summit for Social Development (1995), and the Second United Nations
Conference on Human Settlements (1996), have placed health and human development on
the political agenda. These initiatives have achieved a global consensus on issues of
development that emphasize policy integration, and participation and action at all levels,
especially at the local level. This section reviews some key international initiatives that are
relevant to cities and sustainable development.
The review of different frameworks can provide an insight into the path to sustainability
particularly at the level of the city. At present Asian mega-cities are reliant upon fossil fuels
mainly for transportation and the generation of electricity. As a consequence the local as well
as global environments are unsustainable and new strategies need to be considered.
Different initiatives have shown their attempts to improve such sustainability. However, there
is no single solution for every city. This is due to the fact that different cities have different
conditions. From the energy point of view, more sustainable energy systems in the city
should inevitably include the increased use of renewable energy, energy efficiency in all
economic sectors, efficient public transportation, and ultimately a steady reduction of the
dependence on fossil fuels as well as the introduction of cleaner fuels. In addition, social
sustainability requires that the residents have access to appropriate, affordable, healthy and
safe energy services. Agenda 21
The United Nations program of action on sustainable development is a detailed plan of
initiatives that has greatly influenced thinking and action on all levels. It emphasizes the
importance of local action. Agenda 21 (UN 1993) addressed all major areas of development,
especially emphasizing the environment but also considering international cooperation,
poverty, human health and population. It indicates several key aspects of social
development, including equal rights, empowerment and education. Agenda 21 devotes one
chapter to the role of local authorities and identifies local governments as one of the main
partners in its implementation. About two thirds of the recommendations have implications for
cities. Planning, local capacity building, community and inter-sector involvement and
information are highlighted as processes for implementation. Several guides to Agenda 21
have been produced, some specifically for local governments (UN 1993; WHO 1997).
However, Agenda 21 does not contain a specific chapter on energy issues but it has some
aspects to do with energy in relation to environment and development and has been
addressed in Chapter 9 of Agenda 21. This chapter is emphasis on the protection of the
atmosphere, which includes four program areas, improving the scientific basis for decisionmaking, promoting sustainable development, preventing stratospheric ozone and preventing
trans-boundary atmospheric pollution (UN 1993).
Agenda 21 forms the basis of action for many sectors and levels of government, and the key
role of local governments is clearly highlighted. While the role of energy in sustainable
development is addressed in several chapters dealing with human health, promoting
sustainable human settlements, the protection of the atmosphere, transportation, industry,
agriculture and technology. Energy issues and strategies were not comprehensively dealt
with within Agenda 21. Second United Nations Conference on Human Settlements
The Second United Nations Conference on Human Settlements (UN-HABITAT 1996) was
held in Istanbul on 3-14 June 1996. The main objective was to make the world's cities, towns,
and villages healthy, safe, equitable and sustainable. Other objectives included: placing
urbanization and urban-rural relationships at the top of national and international
development agendas; promoting new policies and strategies for managing urbanization and
housing development; and helping to improve community environments and highlight needs
and opportunities for investment into infrastructures and services. The Habitat Agenda
comprises a Statement of Principles and Commitments and a global plan of action adopted
at the conference. The Agenda addresses key challenges facing the world's towns and cities
over the next two decades, with a special focus on remedial action from 1996 to 2000. The
growth of urban agglomerations also provides a challenge for the future, especially when
such urbanization is accompanied by increased social polarization, poverty and
unemployment. The European Sustainable Cities & Towns Campaign
The European Sustainable Cities & Towns Campaign (EC 1994) has arisen from Agenda 21
and the European Commission’s fifth action program on the environment. The Campaign
forms part of the Sustainable Cities Project, the overall cooperation between the Campaign,
the European Commission’s Expert Group on the Urban Environment, and the Commission’s
Directorate-General for Environment, Nuclear Safety and Civil Protection. The objective of
the Campaign is to promote development towards sustainability through Local Agenda 21
processes by strengthening partnerships among all actors within the local community, as well
as cooperation between local authorities.
The Aalborg Charter was adopted at the first European Conference on Sustainable Cities &
Towns in Aalborg, Denmark in 1994. The charter initiated the campaign and proposed a
number of stages in preparing a local action plan. More than 200 local authorities have
signed the charter, committing them to producing a local action plan for sustainable
development. The campaign has initiated a number of activities through the collaboration of
networks to support and encourage these cities. The activities included the development and
publication of guidance for local planning, training courses, seminars, networking and
databases on good practices.
The second phase of the campaign (1996-1998) was the implementation of the Lisbon Action
Plan adopted at the Second European Conference on Sustainable Cities & Towns in Lisbon,
Portugal in 1996. The action plan focuses on how to implement the principles laid down in
the Aalborg Charter and how to start and conduct a Local Agenda 21 process, and provides
practical steps on how to plan the implementation of local plans for sustainability.
23 The WHO Healthy Cities Project
The Healthy Cities Project (WHO 1996) started in 1987 and designated 35 cities as
European project cities. The first phase of the project, from 1987 to 1992, emphasized
advocacy, and by tackling the political and institutional barriers to change, laid the foundation
for successful work towards health for all. The strategic objectives of the second phase
(1993-1998) included speeding up the adoption of policy at a city level, strengthening
national and sub-national support systems, and building strategic links with other sectors and
organizations that influence urban development. Comprehensive city health plans were
setting explicit targets and tackling challenges such as equity and sustainable development
while establishing mechanisms to promote accountability for health. The Sustainable Cities Programme
The Sustainable Cities Programme (SCP 1999) is a global program of the United Nations
Centre for Human Settlements (UN-HABITAT) and the United Nations Environment
Programme (UNEP). It is the leading technical cooperation program in the field of urban
environmental planning and management, and is the principal activity of the United Nations
system for operational sustainable urban development, and thus contributing to the
implementation of the globally-agreed Agenda 21 and Habitat Agenda. The SCP is currently
active in the following places:
Africa: Accra (Ghana), Dakar (Senegal), Dar es Salaam (Tanzania), Ibadan (Nigeria),
Lusaka (Zambia), Maputo (Mozambique), Moshi (Tanzania), Nampulo (Mozambique),
Zanzibar (Tanzania);
Asia: Colombo (Sri Lanka), Madras (India), Cagayan de Oro, Tagbilaran, and Lipa
(Philippines), Shenyang (China), Wuhan (China);
Middle East: Ismailia (Egypt), Tunis (Tunisia);
Latin America: Concepcion (Chile); and
Central & Eastern Europe: Katowice (Poland), Moscow (Russia), St. Petersburg
(Russia). Kitakyushu Initiative
The Kitakyushu Initiative Network (IGES 2004; 2005) was formed (as of April 2005) from
members from 61 cities in 18 countries in the Asia-Pacific region. It is forecasted that by
2015 two-thirds of the mega-cities in the world will be in the Asia-Pacific region. It is thus
essential for cities to join together to help each other and tackle environmental issues. The
Network is playing an important role in fostering the capacity building of local environmental
staff through conducting pilot projects and sharing lessons learned among the member cities
in order to improve the quality of the environment. This initiative was adopted in 2000 at the
Ministerial Conference on Environment and Development in Asia and the Pacific (MCED),
which is one of the most important official meetings in the Asia-Pacific region for environment
and development.
In the year 2004 the countries of the Asia-Pacific reviewed their activities towards
sustainable development. During its network meeting, the Kitakyushu Initiative adopted the
Kitakyushu Initiative Action Plan for future activities (2005-2010). The action plan included
concrete activities to help member cities reduce and manage solid waste, improve air quality,
promote urban environmental planning, conserve and improve water quality, build
management capacities particularly through the use of Information Communication
Technology (ICT), and increase energy conservation. Member cities’ lack of capacity and
appropriate technology as well as difficulties of procuring financial resources, were noted.
Cooperation for capacity-building, financing, technology transfer, and institutional
strengthening were identified as possible mechanisms to address these challenges. The
Network has also proposed the outline of eight functions as follows:
Assistance in the preparing and implementing of integrated and sustainable urban
development plans and strategies with quantitative indicators.
Periodical monitoring of implementation and progress in terms of quantitative
Promotion of information exchange and sharing of experiences among participating
local governments.
Provision of a platform for the transfer of technology and know-how packages, good
practices and a successful municipal/regional model for sustainable development.
Links, catalysation, and facilitation of internal and external financial support for
international cooperation initiatives to network local authorities.
Facilitation of capacity-building activities for environmental administration staff in
participating local governments.
Promotion of environmental education program in inter-city cooperation such as
student exchanges.
Encouragement of private enterprises to participate in infrastructural development
and environmental quality enhancement program. The International Solar Cities Initiative
The International Solar Cities Initiative (ISCI 2004) is a new organization for sustainable
action in urban energy management worldwide. It is a partnership between concerned cities
and researchers involved in climate research, renewable energy and urban design. The
emphasis is on total-city energy planning, based on a universal per capita emissions target
for atmospheric sustainability by 2050. All ISCI cities will use the same per capita target. This
target will be consistent with the Intergovernmental Panel on Climate Change (IPCC) and
2050 population estimation data. It is initially set to two tons of CO2-equivalent in GHGs per
head, which is consistent with a 2050 global target of 450 parts per million by volume of CO2equivalent. ISCI pathfinder cities will provide other cities with much needed practical
examples of cost-effective, sustainable energy planning that can achieve these goals.
The ISCI project has not focused on small-scale isolated demonstrations or expensive, but
ineffective, public relations projects. The goal of the initiative is to support energy and climate
policies by stimulating the interest of cities into becoming benchmark cities that commit
themselves to ambitious emission reduction goals. ISCI will support this goal by:
Developing recommended agreed targets for emissions reduction by 2050.
Encouraging participant cities to use backcasting planning from the 2050 target to
assess the appropriateness of interim targets and projects.
Providing a forum for the exchange of scientific and technical knowledge, developed
by high-level experts in the field of renewable energy and cooperating urban
planners, through journals, electronic media and conferences.
Creating international technical standards for baseline studies, and a scientifically
acceptable modeling framework for the unbiased comparison of different technical
and policy measures.
Assisting the development of regional and international projects and educational
programs that implement renewable energy and energy efficiency.
Assisting member cities in the coordination of global emission targets with other
social, economic, and environmental goals.
Encouraging cities to participate in new business opportunities created through the
use of advanced emissions reduction techniques. The International Council for Local Environmental Initiative
The International Council for Local Environmental Initiatives (ICLEI 2006) was founded by
local governments at the United Nations Headquarters in New York in 1990. ICLEI's mission
is to build and serve a worldwide movement of local governments to achieve tangible
improvements in global sustainability with a special focus on environmental conditions
through accumulative local actions. Recently, over 475 local governments have joined the
association. ICLEI is an open association and the vision is as:
A growing association of local governments dedicated to sustainable development.
A high-energy, flexible movement of local governments working together in networks
to improve performance supported by campaigns, programs and strategic alliances.
An effective sustainability and environmental agency demonstrating creativity and
excellence both in developing cutting-edge solutions and in program design and
ICLEI develops and runs a range of campaigns and program that address local sustainability
issues. It offers the opportunities of joining pilot projects and receiving support and guidance
for implementation. ICLEI’s selected priority program areas include building sustainable
communities and cities, protecting common global interests, and involvement in governance
and sustainability management.
2.2. Factors driving energy consumption
Modern forms of energy empower human beings in countless ways by reducing drudgery,
increasing productivity, transforming food, providing illumination and transportation, powering
industrial processes, conditioning space for households and buildings, facilitating electronic
communications and computer operations, etc. Patterns of energy use vary dramatically, in
ways that reflect and intensify social and economic inequities. However, the prevalent pattern
and profile of energy use today raises important questions about the links between energy
and the economy, environmental protection, society and security. In the previous sections
some of the physical determinants of energy used in the cities, as well as impacts, were
outlined. At the city level, similar to the national level, there is a need to understand the
mechanisms of the driving forces behind energy consumption. A sound understanding of
these factors is essential to follow the energy use patterns and trends. The following subsections describe the factors that drive energy consumption in the city.
2.2.1. Urban demographic changes
There is no question as to whether or not population and the sizes and numbers of
households affect energy use in the city. Tokyo’s population has been stable since the early
1970s. Beijing and Shanghai’s populations have been growing, especially since 1980, but
less rapidly than other similar cities in Southeast Asia. Much of this growth is attributable to
migrant populations. Changes in the political boundaries such as there have been in Beijing
have often resulted in drastic population changes (IGES 2004).
The number of households is obviously increasing primarily due to the rapidly decreasing
size of the average household. The population of Shanghai is greater than that of Tokyo, but
has fewer households. The number of households could be a more important determinant of
energy use than population.
Another important aspect of demographic change is the difference between the daytime
population and that of the night-time. Tokyo attracts a significant number of commuters who
are not counted among the resident population. One-third of the total workforce commutes
from surrounding cities and prefectures utilizing its well-developed railways and subways.
The ratio of the daytime to night-time population increased from 1.15 in 1975 to 1.25 in 1999
in Tokyo, and stabilized after 1990 (TMG 2000; IGES 2004). Bangkok has a similar trend to
Tokyo, during the morning and evening people travel from outer Bangkok Metropolitan Area
(BMA) and outlying cities to inner BMA for their work and schooling. Another issue concerns
the immigrants, who live in Bangkok but are not registered as residents this makes Bangkok
more complex and difficult to predict. In Beijing and Shanghai the situation is in contrast with
Tokyo and Bangkok (Yoon & Araki 2002).
2.2.2. Patterns of urbanization
Patterns of urbanization, such as urban population density, the structure of urban functions,
urban geography, and land use patterns, are important determinants of energy use in a city.
These basics have to meet the fundamental needs of the growing urban population and its
quality of life while using natural resources in a sustainable way. For example, a compact city
may have lower per capita energy consumption due to its compact infrastructure and lower
per capita building floor space, both traits that reduce energy consumption. In Japan, the per
capita energy consumption in dense urban areas is lower than that in non-urban areas.
Developed countries exhibit a similar relationship. In contrast, cities in China and Thailand
report the opposite trend, which is that city residents use more energy (Ichinose et al. 1993;
Harasawa 2002). This can be explained by the income gap between urban and rural areas in
Japan, where density has a clear influence on per capita energy use. In contrast, in China
and Thailand, as well as other developing countries the effect of the significant income gap
surpasses that of density for commercial energy uses and the effect of density is often not
Corresponding to the IGES (2004) report, the energy and CO2 performances in terms of a
cumulative index of economy and per capita energy/emission of Tokyo are better than those
of Beijing and Shanghai, this is partly due to the density effect. This density effect is
especially significant in the central business districts of cities. This is because large cities in
China, such as Beijing, have a combination of urban and rural settlements. The traditional
urban structure of Tokyo is a central core; it plans to adopt a multi-core urban form that might
have significant implications for energy use. It is also characterized by mixed land use. In
Beijing and Shanghai, urban zoning for industrial and other activities has become common
practice in the last few decades. The exact implications for energy use in regards to the
relocating of industries from residential areas to designated industrial zones are difficult to
quantify. Although, such measures in the residential areas of Shanghai and Beijing have
resulted in reductions in the concentration of air pollutants (Stubbs & Clarke 1996).
In Bangkok, the inner city districts are densely populated and the land has already been
utilized to near saturation point. Land use patterns in this area include commercial ventures,
residential quarters, government offices, education establishments and temples.
Intensification of the land usage is characterized by vertical developments in the form of
high-rise buildings for offices and dwellings. Therefore, in this area only the middle and highincome groups can afford to live where a good infrastructure is available, which is significant
for energy use.
According to contemporary models of a sustainable city, three city models are currently
debated: the compact city, the decentralized concentration and the short cycles city (Frey
2004). Their supporters suggest that all of these cities are models for sustainable urban
development. The compact city
Generally, a compact city refers to a city with a high-density population. Supporters of the
compact city suggest that it has urban qualities not provided by sprawling cities. They say
that it is efficient in terms of the distribution of human activities and in its optimal use of its
infrastructure and viable public transport and non car-based transport systems. There is the
preservation of valuable land and energy through dense development, good access to
workplaces, services, facilities and a reduced need to travel. A high quality of life, a safe and
vibrant environment, and support for businesses and services as a result of the concentration
of activities within in urban quarters. There are reduced levels of pollution, as a result of
reduced car usage. There is also potential for socio-cultural vitality and diversity, and many
options for a wide range of activities, social amenities, and cultural facilities (Frey 1999;
Girardet 2001; Lloy-Jones 2004). The decentralized concentration
Supporters of the decentralized concentration or the polycentric city model stress that the
process of a diffusion of people and urban activities has not simply led to urban sprawl but to
the emergence of new more polycentric patterns of urban development with a greater
specialization of functions between centers and a growing significance on networks of cities
(Breheny 1992; van der Valk & Faludi 1992; Knights 1996; Frey 2004; Lloyd-Jones 2004).
An urban structure with compact and dense settlements and centers away from the principal
center and conurbation is seen as a realistic compromise between compaction and dispersal.
This group also argues that the compact city has a number of serious deficiencies. For
example, its negative impact on the countryside and the economic development of rural
communities, increased congestion and pollution, reduced amenity space and privacy, and
increased social segregation due to high-priced housing in the city's core and suburbs (Frey
2004; Lloy-Jones 2004). The short cycle city
This model is a further development of decentralized concentration through the inclusion of
ecological and environmental principles. It is associated with the environmental thrusts of
Local Agenda 21 and an emphasis on achieving local environmental sustainability through
more efficient local use of natural resources and recycling, greater local autonomy, and
smaller ecological footprints (Barton et al. 1995; Lloyd-Jones 2004).
In structural terms this model responds to the existing nature of today's sprawling cities and
can be adapted and implemented in various urban forms like dense city cores and new green
field development. This model aims to increase the quantity, quality and accessibility of
green spaces in the city. To enlarge and integrate the green structures (for example, through
green networks and the landscaping of the public realm), giving additional possibilities for
recreation and leisure, and having an intentional ecological impact on the microclimate of the
city and reducing the impact of pollution (Frey 2004).
2.2.3. Income growth and social change
It is well recognized that rapidly industrialized countries as well as those in economic
transition like China and Thailand yield high economic growth. Tokyo experienced sustained
economic growth from the early 1960s until it went into recession in the early 1990s. Doubledigit growth in Beijing and Shanghai has continued since 1990 (IGES 2004). The financial
collapse in Thailand during 1997-1998 had obvious implications for Bangkok, but Thailand
has regained growth rates of the GDP at approximately 5% per year since 2001 (DEDP
2000). Economic growth in the city has led to an increase in income, which in turn affects the
lifestyles of residents. For they can afford to consume more material goods, follow fashions,
and pursue leisure activities. Hence, they use more energy. Figure 2.2 shows the trend in
GDP per capita in China (not including Hong Kong and Macao), India, Japan, Korea,
Malaysia, Singapore and Thailand.
GDP per capita (US dollars)
Figure 2.2. Trend in GDP per capita in selected Asian countries (IGES 2004).
Currently the studies in public and political concern about energy and its impacts are
increasing. In fact, efforts to empirically examine, theorize, and model the dynamics and
consequences of societal energy use have been pursued for more than 20 years. But
understanding energy consumption is far from being a straightforward matter (IGES 2004). It
is obvious that energy flows are produced and shaped by human actions. This consumption
occurs via a complex of fuel flows, energy technologies and regulatory systems. The flow is
determined by a fairly complex interplay of socio-cultural, geographic, technological and
institutional factors (Stern 1993).
The social sciences have produced fairly good pieces of work on the role of energy and
energy technology in society e.g., Cottrell 1955, Mazur & Rosa 1974, White 1975, Adams
1975, Duncan 1978, Buttel 1979, Humphrey & Buttel 1982 and Olsen 1991. A large amount
of literature also focuses on the connections between social status and consumption,
including the construction of status via the stylized consumption of food, clothing, music,
language, automobiles, housing and appliances (Ewen 1976 & 1988; Cowan 1989;
Featherstone 1990 & 1991; Gartman 1991).
The interdisciplinary literature that is directly concerned with the consumption of energy
suggests that social structure and cultural practice are indeed central to the structuring of
energy consumption. The significant differences in energy use are observed between income
groups, across the stages of a life-cycle and among ethnic subcultures (Lutzenhiser 1992;
Lutzenhiser & Hackett 1993).
According to the study on socio-technical household energy consumption by Lutzenhiser
(1997), using the data from a major survey of housing, energy use and household technology
in northern California. This data showed considerable variation in the energy consumption
across sample households, with distinct differences in consumption by sub-groups defined
on the basis of both social stages e.g., life-cycle, wealth, ethnicity, and also technical stages
e.g., age, type, size of housing and types of appliances. A series of multivariate models were
estimated and these socio-technical models offer a good match in accordance with the data.
They suggested that both the behavior of social groups and their material conditions
contribute to the structuring of consumption in a variety of ways. A second-stage analysis
using regression estimates and subgroup characteristics showed that various combinations
of behavior, housing and technology were responsible for shaping consumption quite
differently across social groups. Rates of energy input and pollution were also found to be
socially variable.
From the study mentioned above it can be concluded that the social system is a primary
determinant of energy consumption. In a system of status-graded lifestyles, volumes of
energy flow provide rough measures of social standing. The poor being excluded from all but
the most modest forms of consumption, while middle class consumption is centered largely
on housing and technologies, and the wealthy are empowered in a variety of ways by high
levels of energy flow.
2.2.4. Structure of economic activities
From the city point of view the volume of the Gross Regional Product (GRP) and energy
demand have a direct co-relation since economies heavily rely on energy sources. In Tokyo,
the growing population, high population density, noise and environmental problems related to
primary industries are forcing industries to relocate away from city centers to the outskirts
and peri-urban areas. Therefore, Tokyo is dominated by service industries and is becoming a
commercial city. The share of the tertiary sector in total value added has risen from 67% in
1980 to 78% in 1997 (Dhakal & Kaneko 2002). There are similar trends in other cities in
China, Japan, Korea and Thailand that are following Tokyo's pattern.
In Bangkok, a gradual shift from primary industries to tertiary and commercial activities has
followed the Tokyo model. The primary industries have been moved from the inner Bangkok
area to the city's outskirts, but there are still many SMEs in the city center. In contrast, Beijing
and Shanghai are industrial cities. As a result, the industrial sector is expected to play a
major role in energy demand in Beijing and Shanghai.
A study by the Asian Development Bank indicated that the high income countries show
strong co-relations between the fall in agricultural employment, generally declining rates of
manufacturing and industrial employment, and a strong growth in employment within the
service sector. The middle-income countries show a more mixed pattern with declines in
agricultural employment countered by a variable increase in the manufacturing and service
industries. The low-income countries remain dominated by agriculture with little growth in
manufacturing but significant growth in the service sector. Within these co-relations there is
clear evidence of the increasing importance of the macro-economy of mega-cities
(Prabhavalkar 2002).
2.2.5. Lifestyle and the level of consumption
Lifestyle and the level of consumption regarding goods and services are significant factors
that can either increase or decrease energy consumption in the residential sector. A number
of determinants, particularly the scale and intensity of cooking, lighting, electrical appliances
and space conditioning devices, influence the use of energy in a household. Consumption
can be viewed as the motor propelling an economy in that the producers will fabricate only
the goods and services that consumers want to buy. Most environmental degradation can be
traced to the extraction of energy resources and other materials with their transformation to
produce (both directly and indirectly) the goods and services valued by consumers (Duchin
1997). The patterns of goods consumption could alter, which can alleviate the pressures on
energy utilization.
The lifestyle and consumption level can be concerned with individuals. An individual's
behavior is tightly linked to his or her employment. In fact, the income earned has to cover
the costs for purchased goods and services. Consumption behavior is also related to other
people's employment and consumption. A household generally comprises of one or more
paid workers. At least a portion of the income they earn is pooled, based on various kinds of
negotiations to pay for both common purchases and financially dependent individuals. A
household's lifestyle refers to the jointly determined work and consumption practices of its
members (IGES 2004). A society often believes that its quality of life generally increases with
higher consumption levels. In other words, the higher the consumption of goods and services
is the better the quality of life is. This belief has to change.
Lifestyles are reflected by consumption patterns, but they also encompass other elements
such as time, social identity, education, employment, family status and cohort. When looking
at the entire life of a person or a household, time and budget constraints become relevant
variables that can be addressed. It is also a more appropriate level in which to understand
social dynamics and wealth accumulation. Cohort effects can then be evaluated and related
back to individual issues (Schipper et al. 1989; Jalas 2002; Perrels 2002).
In most developing Asian countries, significant economic and lifestyle changes have been
taking place. For instance, the Chinese have continued to require a larger quantity and better
quality of energy. People directly consume energy for lighting, cooking and other daily uses.
But they also aspire to a higher quality of life by purchasing fashionable goods and services.
All of these products and services obviously consume energy during their production
processes. On the other hand, they also indirectly consume energy through the act of
purchasing these products.
In the U.S., the Empowerment Institute (EI 2005) claims that their Sustainable Lifestyle
Campaign, which is working for over 30,000 people in the U.S., has achieved the following
average resource savings per year. Namely, 35-51% less garbage sent into the waste
stream, 25-34% less water used, 9-17% less energy used, 16-20% less fuel used for
transportation and US$ 227-389 saved through the more efficient use of resources.
However, it is not clear how representative and reproducible these numbers are.
Recently UNEP has suggested using the life functions of nutrition, mobility, housing, clothing,
health, and education as a way to organize its work on Sustainable Consumption and
Production (UNEP 2002). These functions can be seen as components of lifestyles.
Moreover, the counterparts to these functions are products and services. This is a useful
manner of organizing the analysis of sustainable consumption issues. The discussion about
sustainable consumption suggests focusing more on the following factors: the available
infrastructure, the nature of the energy system, the properties of products and services
available, and through habits, social expectations and long-term decisions, which shape the
long-term development of society (Hertwich 2003).
2.3. The metabolism of a city
As a city is a dynamic system it is therefore important to understand trends in energy use
over time. Like the human body, a city can be characterized by its metabolism, where energy
and materials are used as input and waste as output. The metabolism approach is a powerful
metaphor for the illustration of the processes that mobilize and control the flows of energy
and materials through a city. Understanding how a city works as an ecological system will
help take control of the vital links between human actions and the quality of the environment.
Hence, the knowledge of human-induced energy and material flows with comparison to those
of natural flows is a major step towards the design of sustainable development schemes.
2.3.1. City as an ecosystem
The impact of cities upon the environment has not always been the same. Cities in
developed countries have largely overcome their traditional environmental problems such as
waste water removal, sanitation, water supply, indoor air pollution, etc. Thus the attention
has turned to their impact on ecosystems further away as well as those that are larger in
scale. Cities in the developing world are more concerned with other issues. Challenges for
urban development in developing countries have been divided into two categories: inefficient
modes of resource use (for example, in the water or energy supply) and a limited capacity for
the absorption of pollution and flooding. Brandon & Ramankutty (1993) classify the key urban
environmental challenges in the Asian region as water pollution, air pollution, solid waste
management and inappropriate land use.
A study of urban energy and environmental problems requires an interdisciplinary approach,
which in turn requires an interdisciplinary language. One interesting approach is the system
theory as applied to ecology (Odum 1989). At the heart of the system theory is the definition
of a system in terms of its boundary, the flows across the boundary and the links within the
boundary. Changes in the status of the system depend on positive feedback, which enlarge
or otherwise enhance the system, and negative feedback, which reduce it. At the global
scale, environmental change is studied as a set of bio-geochemical cycles, following such
key elements as water, carbon, nitrogen, sulfur, phosphorous and the other nutrients on
which plants and animals depend (Munn 1986; Lovelock 1991; Nisbet 1991). The rate at
which a particular particle moves through a cycle may be measured through the study of
environmental pathways, which show how much time a particle may stay in various
environmental media (Mackay 1991). One aspect of this study, of interest to plant and animal
health, is the tendency for harmful residues to accumulate in flora and fauna, a process
known as bio-accumulation. On a global scale, this accumulation occurs in sinks such as the
atmosphere, the ocean, or the vegetal mass of plants and trees.
So far, the language sounds more or less biological, but it can easily be transformed into
something more familiar to the engineers. As mentioned earlier, an ecosystem is a biotic
assemblage of plants, animals, and microbes, taken together with their physico-chemical
environment. In an ecosystem the biological cycle of materials is maintained by three groups,
which are producers, consumers and decomposers as shown in Figure 2.3. The producers
are plants and some bacteria capable of producing their own food photosynthetically or by
chemical synthesis. The consumers are animals that obtain their energy and protein directly
by grazing, feeding on other animals, or both. The decomposers are fungi and bacteria,
which decompose the organic matter of producers and consumers into inorganic substances
that then can be reused as food by the producers. They are the recyclers of the biosphere.
Nature is capable of sustaining the producer-consumer-decomposer cycle indefinitely, with
the sun as the energy source. The smallest such entity that is self-sufficient is an ecosystem
(Ayres & Simonis 1994).
External World
External World
Figure 2.3. The movement of chemicals and materials through the natural ecosystem
(adapted from Ayres & Simonis 1994).
Functionally, human activities that perturb the natural environment can also be divided into
three similar components (Figure 2.4). This human-induced system is called an
anthroposystem. Productive activities include energy production (fossil fuels), manufacturing
(non-fuel minerals) and growing food. The consumers are humans and their domestic
animals. Decomposing or recycling activities include treatment of waste water and the
recycling of metals. However, whereas an ecosystem relies on its decomposers for a
complete recycling of its elements, the system created by human activity lacks such efficient
decomposers and recyclers. As such, manufactured materials that are no longer needed and
the by-products of industrial activity are disposed of within the physical environment. The
process of adding unwanted material to the environment is called pollution. The waste
products are taken by the atmosphere and the hydrosphere, and delivered into the biological
and geochemical receptors. In this sense, the anthroposystem is a system more open to the
engineering point of view (Ayres & Simonis 1994).
Figure 2.4. The movement of chemicals and materials through a system resulting from
human activity (adapted from Ayres & Simonis 1994).
The above models provide a convenient framework for comparing an ecosystem to an
anthroposystem. The flow of material in both systems is illustrated quantitatively by the
arrows in the figures. In an ecosystem most of the material is transferred from the producers
(plants) to the recyclers (bacteria); only a small fraction is passed through the consumers to
the recyclers. The decomposers (recyclers) return most of the material to the producers for
reuse. In the anthroposystem the flow from the producers to the recyclers is small or even
non-existent, since it would be pointless to produce (mobilize) material and immediately
recycle it without a consumer within the loop. In this system, much of the mobilized material
is transferred to the rest of the external environment by the producer or by the consumer.
Hence, it is an open system with recycling accounting for only a small fraction of the
mobilized matter.
In an ecosystem, recycling and sustained development (evolution) is facilitated by the close
physical proximity and functional matching between producers and consumers. The physical
proximity of producers, consumers, and recyclers in an ecosystem (e.g., plants, animals, and
bacteria) assures that very little energy is required for the physical transport of matter
between the plant and its symbiotic bacterial population. Also, the physical proximity allows a
reasonably fast mutual adjustment if there is a perturbation in the system. In the
anthroposystem, on the other hand, the consumers play a more significant role. There is
usually a significant physical displacement between the producer and the consumer.
According to this it is the amount of energy required to transfer the matter back to the
producer or to a recycler. This physical separation of consumers, producers, and recyclers
appears to be a major difference between the ecosystem and the anthroposystem. It can be
noted that the anthroposystems differ from ecosystems mainly in that they lack efficient
material recyclers that allow sustainable development (Ayres & Simonis 1994).
All of the flows mentioned by the model also occur within a city. The city can be viewed as an
organism with a metabolism that can be studied. Metabolic studies can provide the basis for
discussions on the desirability of changes within the type or scale of a city’s metabolism, and
how such changes might be best accomplished. Graedel (1995) argued that to study a city
as organisms and ecosystems, and to devise ways to evaluate their environmental
performance could benefit society in three fundamental ways namely, by maintaining human
systems, by maintaining environmental systems and then by redesigning human systems.
2.3.2. Review of the metabolism of a city
As the center of population and human activity, a city is also the center of the flow of
materials. A city gathers resources of all kinds from near and far. For example, Bangkok
receives steel and copper from Sweden, porcelain from China, cars from Japan, fashionable
goods from France and Italy, and machines from the U.S. Some of this material, such as the
steel in buildings, is retained for long periods of time. Other material is transformed within a
short time and its residues discarded. Though waste is seldom disposed of within the urban
area itself, it generally moves a much shorter distance than the distance from which their
progenitors were acquired. Cities are great at attracting but weak at dispersing. The
metabolism of a city is a relatively new area of study, and one where much more data needs
to be gathered before meaningful results can be derived (Ayres & Simonis 1994; Newman
1999; Sahely et al. 2003; Kennedy et al. 2005).
The concept of urban metabolism was first developed by Wolman (1965). He viewed the
urban environment as an ecosystem and began measuring its metabolic activities. This was
in the time when air and water quality was deteriorating rapidly in many American cities.
Wolman used the national usage rates of water, food and fuel with the production rates of
sewage, waste and air pollutants to derive per capita inflow and outflow the rates for a
hypothetical American city of one million people (White 1994). Wolman’s top-down approach
to determining material flow, even with the omission of some inputs such as infrastructure
materials, electricity and other durable goods helped to focus attention on the system-wide
impacts of the consumption of goods and the generation of wastes within the urban
environment (Decker et al. 2000).
One of the earliest and most comprehensive studies was that of Brussels, Belgium by
ecologists Duvigneaud and Denaeyer-Desmet in 1977, which included quantification of urban
biomass and even organic discharges from cats and dogs (Kennedy et al. 2005). The urban
metabolism approach to studying cities continued as part of the UNESCO Man and
Biosphere Project (MAB) in the early 1980s. A significant contribution was made by
Newcombe et al. (1978), who studied the metabolic processes of Hong Kong. Using the
previous mass balance studies, the Hong Kong study took the urban metabolism model a
step further and characterized the physical indicators of the flows through the city and the
social variables that affected the human population such as employment, health, mortality
and satisfaction (White 1994). The study was well ahead of its time in promoting the need for
a more sustainable urban environment.
The pioneering work of Newcombe et al. (1978) has since been updated by Warren-Rhodes
& Koenig (2001). The combination of these studies is powerful in scope, as trends in the
urban metabolic fluxes between 1971 and 1997 could be quantified and closely analyzed. In
its transition from a manufacturing center to a service-based economy, and the addition of
more than three million people since 1971, it seems that Hong Kong paid a high
environmental price for its continued and unrelenting growth. Per capita food, water and
materials consumption had surged by 20, 40, and 149%, respectively, since 1971. In
addition, total air emissions, carbon dioxide outputs, municipal solid wastes, and sewage
discharges had risen by 30, 250, 245, and 153%, respectively (Warren-Rhodes & Koenig
Newman et al. (1996) highlighted the use of metabolism analyses as part of a state of the
environment (SOE) report for Australia. The study compared the metabolic flows for Sydney
between 1970 and 1990. Apart from a few air quality parameters, there was a disconcerting
upward trend in per capita resource inputs and waste outputs over the study period.
A recent study performed by Sahely et al. (2003) estimated the urban metabolism of the
Greater Toronto Area (GTA) and also made an attempt at comparing the urban metabolism
of a few cities worldwide. The most noticeable feature of the GTA metabolism is that the
inputs had generally increased at higher rates than the outputs for the years of study (1987
and 1999). The inputs of water and electricity had increased marginally less than the rates of
population growth (25.6%), and estimated inputs for food and gasoline had increased by
marginally greater percentages than the population. With the exception of CO2 emissions,
the measured output parameters are grew slower than the population; residential solid waste
and wastewater volumes had actually decreased in absolute terms over the 12 year period
from 1987 to 1999. The study also compared the GTA with Sydney and Hong Kong.
Unfortunately, these studies were conducted for different study years, and in many cases the
results are not directly comparable (Sahely et al. 2003). However, to give an overview of the
metabolism of the GTA and Hong Kong, Figure 2.5 shows a systems flow diagram
comparing selected flows for the GTA in 1999 and for Hong Kong in 1997.
(a) the GTA 1999
(b) Hong Kong 1997
Figure 2.5. Comparison of urban metabolisms: (a) the GTA 1999, and (b) Hong Kong 1997
(compiled from Sahely et al. 2003).
In Figure 2.5, food, water, wastewater and residential solid waste are expressed in tons per
capita, CO2 and BOD5 in kg per capita, and electricity in MJ per capita. This type of diagram
is commonly used in material flow analysis and can be used as a communication tool to
convey information to policy-makers as well as to the public.
Studies of full urban metabolism models with trend analysis are few and far between (Sahely
et al. 2003). There are other studies of the metabolism of other cities, including those of
Tokyo, Vienna, Greater London and a part of the Swiss Lowlands (Hanya & Ambe 1976;
Baccini 1987; Hendriks et al. 2000; CIWS 2002). Not all studies have quantified urban
metabolisms as a complete model some of them examine only one perspective. For
instance, Bohle (1994) considers the urban food systems in developing countries. While a
few studies have focused on quantifying the embodied energy in cities (Zucchetto 1975;
Huang 1998), other studies have broadly included fluxes of nutrients and materials and the
urban hydrologic cycle.
From the review of the literature, it can be seen that the model of urban metabolism has
been used by engineers (e.g., civil engineers and mechanical engineers), urban planners
and system ecologists. Interestingly, a thermodynamic approach to urban metabolism
models has been put forward by system ecologists (Odum 1996). This approach quantifies
the flows of embodied energy, which is defined as the total amount of energy needed directly
and indirectly to make any goods or services (Bakshi 2000). Indices and ratios based on
embodied flows can be calculated and used to evaluate different types of systems (Ulgiati &
Brown 1998). This concept is adapted in this thesis; thus, the study will be focused upon
embodied energy rather than a study of full urban metabolism.
2.3.3. Evaluating urban metabolism methodologies
The metabolism of a city can be analyzed in terms of four fundamental cycles, energy,
material, water and nutrients. Differences in the cycles may be expected between cities due
to age, stage of development and cultural factors. Other differences, particularly in the
energy cycle, may be associated with climate or urban density. The detailed methodological
steps of urban metabolism are outlined in the pioneering work of Baccini & Brunner (1991)
and for other studies see Hendriks et al. 2000, Warren-Rhodes & Koening 2001. The major
steps can be concluded as follows: (1) definition of goals and the research questions of the
study, (2) system description, (3) data acquisition, (4) material balances, and (5)
The material balances mentioned in step 4 are common to engineers. In fact, mass flow
analysis has been practiced by engineers for a long time, but emphasis is usually placed on
the flow of a particular substance and not on entire systems (Baccini & Brunner 1991). Since
mass is not destroyed in the system. The total inputs into a process equal the total outputs
plus stock increases. This basic principle holds for urban metabolism models as well. The
main obstacle is obtaining sufficient data to quantify all of the relevant inputs and outputs.
The following mass and energy balance equations were originally put forward by Douglas
(1983) for use in urban metabolism studies (Alberti 1996):
(1) Surface urban energy balance
QS + QF + QI = QL + QG + QE
Eq. (2.1)
where QS is the rate of the arrival of radiant energy from the sun, QF is the rate of the
generation of heat due to combustion and dissipation in machinery, QI is the rate of the heat
arrival from the Earth's interior, QL is the rate of loss of heat by evapotranspiration, QG is the
rate of loss of heat by conduction to soil, building, roads, etc., and QE is the rate of the loss
of heat radiation.
(2) Urban water balance
P + D + A + W = E + RS + S
Eq. (2.2)
where P is the precipitation, D is the dew and hoar frost, A is the water released from
anthropogenic resources, W is the piped water, E is the evaporation, RS is the natural and
piped surface, and the subsurface flow out of the city, and S is the change in water storage
within the urban fabric.
(3) Urban material budget
Ms = Mo + W f + Wa + Mc + Mt
Eq. (2.3)
where M s is the quantity of materials supplied to the city, M o is the materials exported from
the city, W f is the solid and liquid waste materials, W a is the atmospheric pollutants
discharged from the use of materials, M c is the material converted through heat production
or other processes, and M t is the net addition of materials to the urban fabric and stock.
Although the equations 2.1-2.3 outline the major inputs and outputs of an urban region from
a theoretical standpoint they do not necessarily set up a system that can be efficiently
analyzed with the available data. For example, the surface urban energy equation does not
separate man-made energy inputs from natural inputs. In addition, energy fluxes related to a
loss of heat are quite difficult to quantify. They may not be so relevant to analysts. Also, they
would be difficult to communicate to policy-makers and stakeholders (Sahely et al. 2003).
Nevertheless, the idea from the ecologists (Odum 1996), who put forward the
thermodynamic approach to urban metabolism, as mentioned earlier, in addition to the
important idea being, that the interactions between nature and society on which each socioeconomic system depends constitute its relation to the natural environment, is an
understanding of society as a socially organized and thermodynamically open system that
has been termed anthropogenic, social and industrial metabolism (Baccini & Brunner 1991;
Fischer-Kowalki & Haberl 1993; Ayres & Simonis 1994; Kate et al. 2001). A number of
operational tools have been developed to analyze urban metabolism. The following section
gives a brief introduction to the concepts of the well-known analytical frameworks, including
Input-Output analysis (IO-analysis), material flow analysis (MFA), life cycle analysis (LCA)
and ecological footprint. Input-Output analysis
The Input-Output analysis (IO-analysis) is an analytical framework created by Nobel Prize
laureate (economics) Wassily Leontief in the late 1930s (Leontief 1936, 1941, 1970) and was
originally designed to analyze the interdependence of industries in an economy. Today the
compilation of IO-tables is a standard in national accounting statistics in almost all countries
in the world and IO-methods are routinely applied in economic analyses. Since the late
1960s, IO-analysis was extended to address economy-environment relationships, focusing
predominantly on energy use and pollution (Cumberland 1966; Ayres & Kneese 1969;
Leonteif 1970; Griffin 1976; Bullard & Herendeen 1977; Proops 1977; Duchin et al. 1994;
Duchin 1998).
In principle, a standard IO-model is used to calculate gross output and factor inputs required
to satisfy a given final demand. Alternatively, IO-tables represent: (1) the flows of
commodities and services between the industries of an economy in an inter-industry flow
table, (2) the deliveries of commodities and services from the producing sectors to final
demand sectors in a final demand table, and (3) the requirements of primary factors of
production in a factor input or value added table. The variables in the IO-tables may be
measured in physical units such as pieces, tons, joules, in monetary units, or in a mixture of
The advantage of measuring the flows of commodities in physical units as compared to
monetary units lies in the fact that physical units explicitly represent the quantity of the flows,
whereas a measurement in monetary units always combines quantity and unit prices (Duchin
2004). The convention of national accounting, however, is to measure the variables in
monetary units, whereas IO-tables in physical units are available only for a small number of
countries and points in time. Therefore, national statistical offices routinely provide monetary
IO-tables. For instance in Thailand, the National Economic and Social Development Board
(NESDB) produces IO-tables every five years.
To determine resource use, as well as environmental pressures associated with additional
vectors of resource use, energy inputs are needed. Usually, these tables use in a form of a
unit of final energy demand. It can be seen that the IO-tables represent the material flow data
of a society and together with the energy consumption data, the energy metabolism can be
calculated. Hence, from a conceptual point of view, it is clear that the metabolism model can
be adequately understood only if both material and energy are considered. Material flow analysis
This method is an environmental accounting approach aimed at the quantification of
metabolism. It is applicable to various geographic and institutional scales (Hendriks et al.
2000; Grunbuhel et al. 2003; Brunner & Rechberger 2004). MFA at the national level, socalled economy-wide MFA, is probably the most advanced in terms of methodological
standardization and indicator development, but unavailable and uncommon in most
developing countries as compared to IO-tables. The economy-wide MFAs are consistent
compilations of the annual overall material throughput of national economies, expressing all
flows in tons per year (Eurostat 2001).
The MFA was reinvented in the 1990s as a consequence of the growing importance of the
notion of sustainable development. In recent years, methods for economy-wide material flow
accounting have been harmonized and a large number of material flow studies for both
industrial and some developing countries have been published to date (Eurostat 2001; Haas
et al. 2005). MFA makes use of the mass balance principle known as inputs equal outputs
plus stock increase. This requires a sufficiently precise definition of the physical boundaries
of a given economic system and a comprehensive coverage of the inputs, outputs and stock
changes (Fischer-Kowalski 1998; Eurostat 2001). For the purpose of material flow analysis,
highly aggregated indicators are derived from MFA. The convention is to aggregate all solid
materials i.e., biomass, gross ores, industrial and construction minerals, fossil fuels and the
physical volume of traded commodities that cross the defined boundary of a system, but not
water and air because the socio-economic throughput of these materials exceeds that of all
others by order of magnitude (Eurostat 2001).
Overall, these indicators are intended to represent a proxy for aggregated environmental
pressure comparable to aggregated energy use. However, for the purpose of evaluating
environmental pressures associated to a city’s metabolism, MFA cannot be used directly as it
does not specify the material requirements of final demand categories. Therefore, an
additional step in the empirical analysis is needed, one that makes use of other methods
such as NAMEA (National Accounting Matrix Including Environmental Accounts) in
combination with IO-analysis (Eurostat 2001; Haas et al. 2005). Life cycle assessment
LCA is a tool that assesses the impacts of product systems and services, accounting for the
resources used during the production, distribution, use and disposal of a product (ISO 1997).
LCA has developed from the analysis of cumulative or embodied energy demand (Boustead
& Hancock 1979; Casler & Wilbur 1984). It uses physical process analysis sometimes in
combination with monetary IO-analysis. LCA consists of three distinct analytical steps: (1) the
determination of processes involved in the life-cycle of a product, (2) the determination of
environmental pressure such as resource use produced in each of those processes and (3)
the assessment of environmental impacts and aggregation as impact indicators. The ISO
14040 standard for LCA defines the first two steps as inventory analysis and the third step as
impact assessment (ISO 1997). ISO defines two additional procedural steps, goal and scope
definition (i.e., planning the LCA) and interpretation (i.e., discussion and conclusion). LCA
can be seen as constructing a causal link between production processes, associated
environmental stress and the produced products (Haas et al. 2005).
The causal link can be constructed in different ways (Baumann & Tillman 2004):
One can divide all of the existing emissions by the total number of products produced
over a period. This is the more common, attribution mode, which attributes
responsibility for the existing emissions evenly across the produced products.
One can ask what happens when an additional product is produced. This marginal
perspective is relevant, for example, when looking at the production of electricity,
where the existing base load of coal or hydropower stations has significantly different
emissions from the newly built gas or wind power plants.
LCA practice today can build on the cumulative effort of data collection. Standard LCA
software already existing includes databases for many basic materials and a number of
important commodities. More extensive databases, such as EcoInvent, are available for
purchase. Some industry associations have produced their own data. SimaPro, the most
widely used software tool now also contains data from IO-analysis, so that hybrid
assessments can be constructed. The data represents conditions in industrialized countries.
Data from developing and emerging countries, however, is still lacking. Hence, there is a lack
of data especially on a number of agricultural products and manufactured products, and also,
the available data may be biased (Haas et al. 2005). Ecological footprint
The original ecological footprint was conceived as a simple and elegant method for
comparing the sustainability of resource use among different populations (Rees 1992). The
consumption of these populations is converted into a single index: the land area that would
be needed to sustain that population indefinitely. This area is then compared to the actual
area of productive land that the given population inhabits, and the degree of un-sustainability
is calculated as the difference between the available and the required land. Unsustainable
populations are simply populations with a higher ecological footprint than available land.
Consumption is divided into 5 categories: food, housing, transportation, consumer goods and
services. Land is divided into 8 categories: energy land, degraded or built land, gardens, crop
land, pastures, managed forests and land of limited availability. The ecological footprint is
calculated by compiling a matrix in which a land area is allocated to each consumption
category. In order to calculate the per-capita ecological footprint, all land areas are added up,
and then divided by the population, giving a result in hectares per capita.
A number of researchers have criticized the method as it was originally proposed, for
example Ayres 2000, Moffatt 2000, Opschoor 2000, van Kooten & Bulte 2000. The criticisms
largely refer to ecological footprints as being an oversimplification of the complex task of
measuring the sustainability of consumption, leading to the comparisons of populations
becoming meaningless or the result of a single population being significantly underestimated.
In addition, the aggregated form of the final ecological footprint makes it difficult to
understand the specific reasons for the un-sustainability of the consumption of a given
population, and to formulate appropriate policy responses (Rapport 2000). In response to the
problems, the concept has undergone significant modification. The development of and
debate about the method are continuing.
2.3.4. Strengths, weakness, and extending of the metabolism modeling
The study of urban metabolism can be used as part of a national environmental report, for
example the study by Newman et al. (1996). One of the strengths of the urban metabolism
model is that it provides parameters that meet the criteria for good sustainability indicators.
The model is scientifically based on the principles of the conservation of energy and mass. In
addition, Hendriks et al. (2000) highlighted that urban metabolism is a useful tool for early
recognition, as urban metabolism does not rely on signals of environmental stress but rather
highlights potential future problems by demonstrating changes in flows and stocks of the
region. Urban metabolism can also increase the effectiveness of policy making by forcing
decision-makers to consider the whole system and setting priorities accordingly.
The major limitation of the approach of urban metabolism is the availability and accuracy of
data at the city level, especially in developing counties where most data is at the national
level. Even in developed countries information is quite scattered and needs to be
synthesized. Decker et al. (2000) provided a comprehensive review of available data,
highlighted data gaps, and emphasized the need for more interdisciplinary research and
integrative data analysis. A further limitation has been a tendency for the framework to focus
on the biophysical environment with less emphasis on the social and economic issues.
The tool would be more useful to policy-makers and decision-makers if it incorporated socioeconomic considerations. Newman (1999) stated that it is possible to define the goal of
sustainability in a city as the reduction of the city's use of natural resources and the
production of waste while simultaneously improving its livability. Thus, it can fit better into the
capacities of the local, regional and global ecosystems. He also proposed an extended
metabolism model as shown in Figure 2.6, which includes livability measures such as
indicators of health, employment, income, education, housing, leisure, accessibility, urban
design quality and community.
The extended metabolism model (Figure 2.6) indicates how the basic concept of metabolism
can be extended to include the dynamics of settlements and the livability of those
settlements so that the economic and social aspects of sustainability are integrated with the
environmental. In this model, it is possible to specify the physical and biological processes of
converting resources into useful products and wastes. It is based on the laws of
thermodynamics, which shows that anything that comes into a biological system must pass
through it and that the amount of waste is therefore dependent on the amount of resources
required. This model becomes more of a human eco-system approach. Sustainability of a
city is thus not only the reduction of metabolic flows i.e., resource inputs and waste outputs, it
must also be about increasing human livability, including social amenities and health.
Building material
Other resources
Dynamics of
Transportation priorities
Economic priorities
Cultural priorities
Leisure activities
Urban design quality
Solid waste
Liquid waste
Air pollutants
Greenhouse gases
Waste heat
Figure 2.6. The extended metabolism model of the city (adapted from Newman 1999).
The extended metabolism model can be applied to a range of levels and to a range of
different human activities. For example, the industrialists can examine the input of resources
and the output of waste while measuring their usual economic parameters and other matters
such as worker health and safety. A household can make an assessment of their metabolic
flows and livability, and together make attempts to do better with both. An urban planner can
be assessed in regards to sustainability using this model to demonstrate projects.
Businesses can use the model to design sustainability plans. Also, a city can make
comparisons by analyzing indicators for resource use, waste, and livability in other cities
(Newman 1999).
2.4. References
Adams R. (1975), Energy and Structure: A Theory of Social Power. University of Texas
Press, Austin, Texas.
Akbari H., Davis S., Dorsano S., Huang J., Winnett S. (1992), Cooling Our Communities: A
Guidebook on Tree Planting and Light-Colored Surfacing. U.S. EPA Office of Policy
Analysis, U.S. Superintendent of Documents, Washington, D.C., p.18.
Alberti M. (1996), Measuring Urban Sustainability. Environmental Impact Assessment
Review 16, pp.381-424.
Ayres R.U., Kneese A.V. (1969), Production, Consumption and Externalities. American
Economic Review 59 (3), pp.282-297.
Ayres R.U., Simonis U.E. (1994), Industrial Metabolism. United Nations University Press,
Ayres R.U. (2000), Commentary on the utility of the ecological footprint concept. Ecological
Economics 32, pp.347-349.
Baccini P., Brunner P. (1991), Metabolism of the Anthroposphere. Springer-Verlag, Berlin.
Baccini P.A. (1987), A City's Metabolism: Towards the Sustainable Development of Urban
Systems. Journal of Urban Technology 4 (2), pp.27-39.
Bakshi B.R. (2000), A Thermodynamic Framework for Ecologically Conscious Process
Systems Engineering. Computers and Chemical Engineering 24, pp.1767-1773.
Barton H., Davis G., Guise R. (1995), Sustainable Settlements: A Guide for Planners,
Designers, and Developers. University of West of England and The Local Government
Management Board.
Baumann H., Tillman A.M. (2004), The Hitch Hicker's Guide to LCA. Student litteratur,
Lund, Sweden.
Bohle H.G. (1994), Metropolitan Food Systems in Developing Countries: the Perspective of
Urban Metabolism. GeoJournal 34 (3), pp.245-251.
Boustead I., Hancock G.F. (1979), Handbook of Industrial Energy Analysis. Ellis Horwood,
Chichester, UK.
Brandon C., Ramankutty R. (1993), Toward an Environmental Strategy for Asia. World
Bank Discussion Paper 224, World Bank, Washington D.C.
Breheny M.J. (1992), Sustainable Development and Urban Form: An Introduction. In
Breheny, M.J. (ed.) Sustainable Development and Urban Form. London.
Bretz S., Akbari H., Rosenfeld R. (1998), Practical Issues for Using Solar Reflective
Materials to Mitigate Urban Heat Island. Atmospheric Science 32 (1), pp.95-101.
Brunner P.H., Rechberger H. (2004), Practical Handbook of Material Flow Analysis. Lewis
Publishers, New York.
Bullard C., Herendeen R.A. (1977), The Energy Cost of Goods and Services: An InputOutput Analysis for the USA, 1963 and 1967. In Thomas J.A.G. (eds.), Energy Analysis.
IPC Science and Technology Press Ltd., Colorado.
Buttel F.H. (1979), Social Welfare and Energy Intensity: A Comparative Analysis of the
Developed Market Economies. In Unseld C., Morrison D., Sills D., Wolf C. (eds.),
Sociopolitical Efficient Energy Use Policy. National Academy of Sciences, Washington D.C.
Castells M. (1993), Why the Mega-Cites Focus: Mega-cites in the New World Disorder.
Mega-Cities 7th Annual Coordinators Meeting, August 1-7, 1993, Jakarta, Indonesia.
Casler S.D., Wibur S. (1984), Energy Input-Output Analysis. Resource and Energy 6,
CEG (2004), Sustainable Consumption and Production and The Energy Sector. Clean
Energy Group, UNEP Industry and Environment, October - December 2004, pp.15-19.
CESA (2004), Clean Energy States Alliance, as accessed 21.10.2004.
CIWS (2002), A Resource Flow and Ecological Footprint Analysis of Greater London.
Chartered Institute of Wastes Management, London.
Cottrell F. (1955), Energy and Society: The Relation Between Energy, Social Change and
Economic Development. McGraw-Hill, New York.
Cowan R.S. (1989), The Consumption Junction: A Proposal for Research Strategies in the
Sociology of Technology. In Bijker W., Hughes T.P., Pinch T. (eds.), The Social
Construction of Technological Systems. M.I.T. Press, Cambridge, Massachusetts.
Cumberland J. (1966), A Regional Interindustry Model for Analysis of Development
Objectives. Papers of the Regional Science Association (17), pp.65-94.
Decker E.H., Elliot S., Smith F.A., Blake D.R., Rowland F.S. (2000), Energy and Material
Flow Through the Urban Ecosystem. Annual Review of Energy and Environment 25,
DEDP (2000), Thailand energy situation 2000. Department of Energy Development and
Promotion, Ministry of Science, Technology and Environment, Thailand.
Dhakal S., Kaneko S. (2002), Urban Energy Use in Asian Mega-Cities: Is Tokyo a
Desirable Model? In Proceedings of IGES/APN International Workshop on Policy
Integration and Industrial Transformation Towards Sustainable Urban Energy Use for Cities
in Asia, January 23-25, 2002, Institute for Global Environmental Strategies, Kanagawa,
Douglas I. (1983), The Urban Environment. Edward Arnold, London.
Duchin F. (1997), Structural Economics: A Stragegy for Analyzing the Implications of
Consumption. In Stern P.C., Dietz T., Ruttan V.W., Socolow R.H., Sweeney L. (eds.),
Environmentally Significant Consumption. National Academy Press, Washington D.C.
Duchin F. (1998), Structural Economics: Measuring Change in Technology, Lifestyles, and
the Environment. The United Nations University, Island Press, Washington D.C.
Duchin F. (2004), Input-Output Economics and Material Flows. Rensselaer Working
Papers in Economics, Rensselaer Polytechnic Institute, Department of Economics, US.
Duchin F., Lange G.M., Thonstadt K., Idenburg A. (1994), The Future of the Environment.
Ecological Economics and Technological Change. Oxford University Press, New York Oxford.
Duncan O.D. (1978), Sociologists Should Reconsider Nuclear Energy. Social Forces 57,
EC (1994), European Sustainable Cities. European Commission Expert Group on the
Urban Environment, European Commission, Brussels.
EC (2000), Sustainable Urban Design. European Commission, Brussels.
ECOASIA (2001), Towards A Sustainable Asia and the Pacific. Report of ECO-ASIA
(Environment Congress for Asia and the Pacific). Long-term Perspectives Project, Institute
for Global Environmental Strategies, Kanagawa, Japan.
EI (2005), Sustainable Lifestyle Campaign. Empowerment Institute. Available at, as accessed 5.7.2005.
Eurostat (2001), Economy-wide Material Flow Accounts and Derived Indicators. A
Methodological Guide. Eurostat, European Commission, Office for Official Publications of
the European Communities, Luxembourg.
Ewen S. (1976), Captains of Consciousness: Advertising and the Social Roots of the
Consumer Culture. McGraw-Hill, New York.
Ewen S. (1988), All Consuming Images: The Politics of Style in Contemporary Culture.
Basic Books, New York.
Featherstone M. (1990), Global Culture: Nationalism, Globalization and Modernity. Sage,
Featherstone M. (1991), Consumer Culture and Postmodernism. Sage, London.
Fischer-Kowalski M. (1998), Society's Metabolism. The Intellectual History of Material Flow
Analysis, Part I, 1860-1970. Journal of Industrial Ecology 2 (1), pp.61-78.
Fischer-Kowalski M., Haberl H. (1993), Metabolism and Colonization. Modes of Production
and the Physical Exchange between Societies and Nature. Innovation the European
Journal of Social Sciences 6 (4), pp.415-442.
Frey H. (1999), Designing the City: Towards a More Sustainable Urban Form. E & FN
Spon, Routledge, London & New York.
Frey H. (2004), The Search for a Sustainable City: An Account of Current Debate and
Research. Keynote Speaker at Plea 2004 - The 21st Conference on Passive and Low
Energy Architecture, September 19-22, 2004, Eindhoven, The Netherlands.
Gartman D. (1991), Culture as Class Symbolization or Mass Reification? A critique of
Bourdieu's Distinction. American Journal of Sociology 97, pp.421-447.
Girardet H. (2001), Creating Sustainable Cities. Devon, Green Books Ltd.
Graedel T.E., Allenby B.R. (1995), Industrial Ecology. Prentice Hall, Englewood Cliffs, N.J.
Griffin J. (1976), Energy Input-Output Modelling. Electric Power Research Institute, Palo
Grunbuhel C.M., Haberl H., Schandl H., Winiwarter V. (2003), Socio-economic Metabolism
and Colonization of Natural Processes in SangSaeng Villiage: Material and Energy Flows,
Land Use, and Cultural Change in Northeast Thailand. Human Ecology 31 (1), pp.53-86.
Haas W., Hertwich, E., Hubacek K, Korytarova K., Ornetzeder M., Weisz H. (2005), The
Environmental Impact of Consumption: Research Methods and Driving Forces. Reports
and Working Papers from Norwegian University of Science and Technology (NTNU),
Industrial Ecology Programme (IndEcol), Trondheim, Norway.
Hanya T., Ambe Y. (1976), A Study on the Metabolism of Cities. In Science for a Better
Environment, pp.228-233.
Harasawa H. (2002), Compact City Project. In Proceeding of IGES/APN International
Workshop on Policy Integration and Industrial Transformation Towards Sustainable Urban
Energy Use for Cities in Asia, January 23-25, 2002, Institute for Global Environmental
Strategies, Kanagawa, Japan.
Hendriks C., Obernosterer R., Muller D., Kytzia S., Baccini P., Brunner P.H. (2000),
Material Flow Analysis: A Tool to Support Environmental Policy Decision Making. Case
Studies on the City of Vienna and the Swiss lowland. Local Environment 5, pp.311-328.
Hertwich E. (2003), The Seeds of Sustainable Consumption Patterns. In Proceedings of
the 1st International Workshop on Sustainable Consumption in Japan, Society for NonTraditional Technology, May 19-20, 2003, Tokyo, Japan.
Huang S. (1998), Urban Ecosystems, Energetic Hierarchies, and Ecological Economics of
Taipei Metropolis. Journal of Environmental Management 52, pp.39-51.
Huesemann M.H. (2003), The Limits of Technological Solutions to Sustainable
Development. Clean Technology Environmental Policy 5, pp. 21-34.
Humphrey C.R., Buttel F.H. (1982), Environment, Energy and Society. Wadsworth,
Belmont, California.
Ichinose T., Hanaki K., Matsuo T. (1993), International Comparison of Energy
Consumption in Urban Area. Proceedings of Environmental Engineering Research, Japan
Society of Civil Engineers (JSCE) 30, pp.371-381.
ICLEI (2006), Strategic Plan 2007-2012 (The Cape Town Plan). The International Council
for Local Environmental Initiatives. As approved by the ICLEI Council in Cape Town, South
Africa, March 3, 2006.
IGES (2004), 2004 Top News on Environment in Asia. Institute of Global Environmental
Strategies (IGES), Japan.
IGES (2005), Urban Environmental Management Challenges in Asia. Institute for Global
Environmental Strategies (IGES), Japan.
ISCI (2004), International Solar Cities Congress 2004. International Solar Cities Initiative.
Available at, as accessed 21.10.2004.
ISO (1997), ISO 14040: Environmental Management - Life Cycle Assessment - Principles
and Framework. International Standard Organization, Geneva.
Jalas M. (2002), A Time Use Perspective on the Materials Intensity of Consumption.
Ecological Economics 41, pp.109-123.
Kates R.W., Clark W.C, Corell R.H., Jaeger C.C., Lowe I., Schellnbuber H.J., Bolin B.,
Faucheux S., Grubler A., Kasperson R.E., Mabogunje A., Mooney H.A., Moore B., Svedin
U. (2001), Sustainability Science. Science 292, pp.641-642
Kennedy C.A., Cuddihy J., Yan J. (2005), The Changing Metabolism of Cities. Department
of Civil Engineering, University of Toronto, Canada.
Knights C. (1996), Economic and Social Issues. In Jenks, M., Burton, E., Williams, K.
(eds.), The Compact City: A Sustainable Urban Form? E & FN Spon, London.
Landberg H. (1981), The Urban Climates. Progress in Physical Geography 8 (1), pp.1-31.
Leontief W. (1936), Quantitative Input-Output Relations in the Economic System. Review of
Economics and Statistics 18, pp.105-125.
Leontief W. (1941), The Structure of American Economy. Oxford University Press, New
Leontief W. (1970), Environmental Repercussions and the Economic Structure. An InputOutput Approach. Review of Economics and Statistics 52 (3), pp.262-271.
Lloyd-Jones T. (2004), Urban Design for Sustainability. Final Report of the Working Group
on Urban Design for Sustainability to the European Union Expert Group on the Urban
Environment, January 2004.
Lovelock J. (1991), Gaia. The Practical Science of Planetary Medicine. Gaia Books Ltd.,
Lutzenhiser L. (1992), A Cultural Model of Household Energy Consumption. Energy the
International Journal 17, pp.47-60.
Lutzenhiser L. (1997), Social Structure, Culture, and Technology: Modeling the Driving
Force of Household Energy Consumption. In Stern P.C., Dietz T., Ruttan V.W., Socolow
R.H., Sweeney L. (eds.), Environmentally Significant Consumption. National Academy
Press, Washington D.C.
Lutzenhiser L., Hackett B. (1993), Social Stratification and Environmental Degradation:
Understanding Household CO2 Production. Social Problem 40, pp.50-73.
Mackay D. (1991), Multimedia Environmental Models. The Fugacity Approach. Lewis
Publishers, Chelsea MI.
Mazur A., Rosa E.A. (1974), Energy and Lifestyle. Science, pp.607-610.
Moffatt I. (2000), Ecological footprints and sustainable development. Ecological Economics
32, pp.359-362.
Munn R.E. (1986), Global Environmental Prospects. Geography, Resources and the
Environment. Themes from the Work of Gilbert F. White. (eds.), University of Chicago
Press, Chicago, pp.326-338.
Newcombe K., Kalina, J.D., Aston A.R. (1978), The Metabolism of a City: the Case of Hong
Kong. Ambio 7, pp.3-15.
Newman P., Kenworthy J. (1989), Cities and Automobile Dependence. Gower Technical,
Newman P.W.G., Birrell R., Holmes D., Mathers C., Newton P., Oakley G., O'Connor A.,
Walker B., Spessa A., Tait D. (1996), Human Settlements. In State of the Environment
Australia 1996. State of the Environment Advisory Council, CSIRO Publishing, Melbourne,
Australia, pp.1-57.
Nisbet E.G. (1991), Leaving Eden. To Protect and Manage the Earth. Cambridge
University Press, Cambridge.
Odum E.P. (1989), Ecology and Our Endangered Life-Support Systems. Sinauer
Associates, Sunderland.
Odum H.T. (1996), Environmental Accounting: Emergy and Environmental Decision
Making. John Wiley & Sons Inc., New York.
Olsen M.E. (1991), The Energy Consumption Turnaround and Socioeconomic Well-being
in Industrial Societies in the 1980s. In Freese L. (eds.), Advances in Human Ecology. JAI
Press, Greenwich.
Opschoor H. (2000), The ecological footprint: measuring rod or metaphor? Ecological
Economics 32, pp.363-365.
Perrels A. (2002), Understanding Consumption Patterns - Including Time Use, Skills, and
Market Failure. Life-cycle Approaches to Sustainable Consumption. Workshop
Proceedings, Luxemburg, Austria, International Institute for Applied Systems Analysis,
Prabhavalkar N.S. (2002), Megacity Mumbai. International Conference on Megacities –
nightmare or chance for developing countries? June 17, 2002, Hannover, Germany.
Proops J.L.R. (1977), Input-Output Analysis and Energy Intensities: A Comparison of
Some Methodologies. Applied Mathematical Modelling 1, pp.181-186.
Ramanathan R. (2000), Urban Transportation. In India Development Report 1999-2000,
Parikh K.S. (eds.), Oxford University Press, Mumbai.
Rapport D.J. (2000), Ecological footprints and ecosystem health: complementary
approaches to a sustainable future. Ecological Economics 32, pp.381-383.
Ree W.E. (1992), Ecological footprints and appropriated carrying capacity: what urban
economics leaves out. Environment and Urbanization 4 (2), pp.121-130.
Sahely H.R., Dudding S., Kennedy C.A. (2003), Estimating the Urban Metabolism of
Canadian Cities: Greater Toronto Area Case Study. Canadian Journal of Civil Engineering
30, pp.468-483.
Schipper L., Bartlett S., Hawk D., Vine E. (1989), Linking Life-Styles and Energy Use: A
Matter of Time? Annual Review of Energy 14, pp.273-320.
SCP (1999), The SCP Source Book Series: Urban Air Quality Handbook. The Sustainable
Cities Programme, United Nations Centre for Human Settlements (UNCHS), Nairobi,
Smil V. (1984), The Bad Earth: Environmental Degradation in China. M.E. Sharpe, New
Steemers K. (2003), Energy and the City: Density, Buildings and Transport. Energy and
Building 35, pp.3-14.
Stern P.C. (1993), A Second Environmental Science: Human-Environment Interactions.
Science 260, pp.1897-1899.
Stubbs J., Clarke G. (1996), Mega-city Management in the Asian and Pacific Region:
Policy Issues and Innovative Approaches. Asian Development Bank.
TMG (2000), Transport Demand Management Action Plan. Tokyo Metropolitan
Ulgiati S., Brown M.T. (1998), Monitoring Patterns of Sustainability in Natural and Manmade Ecosystems. Ecological Modelling 108, pp.23-26.
UN (1993), Earth Summit - Agenda 21. United Nations, New York.
UN (2001), 2001 Review of Developments in Transport and Communications in the
ESCAP Region 1996-2001. United Nations-ESCAP, Bangkok, Thailand.
UN (2002), World Urbanization Prospects: The 2001 Revision. Population Division,
Department of Economic and Social Affairs, United Nations Secretariat, New York.
UNEP (2002), UNEP Contribution to Framework on Promoting Sustainable Consumption
and Production Patterns. Division for Technology, Industry and Economics, United Nations
Environment Programme, Paris.
UN-HABITAT (1996), An Urbanizing World Global Report on Human Settlements. United
Nations Centre for Human Settlements, Oxford University Press, Oxford.
UN-HABITAT (2001), The State of the World's Cities Report 2001. United Nations Human
Settlements Programme. United Nations Publications, New York.
van der Valk A., Faludi A. (1992), Growth Regions and the Future of Dutch Planning
Doctrine. In Breheny, M.J. (eds.), Sustainable Development and Urban Form, London.
van Kooten G.C., Bulte E.H. (2000), The ecological footprint: useful science or politics?
Ecological Economics 32, pp.385-389.
Warren-Rhodes K., Koenig A. (2001), Escalating Trends in the Urban Metabolism of Hong
Kong. Ambio 30 (7), pp.429-438.
White L. (1975), The Concept of Cultural Systems: A Key to Understanding Tribes and
Nations. Columbia University Press, New York.
White R.R. (1994), Urban Environmental Management: Environmental Change and Urban
Design. John Wiley & Sons Ltd., Toronto.
Whitelegg J. (1993), Transport for a Sustainable Future: the Case for Europe. Wiley,
WHO (1996), Creating Healthy Cities in the 21st Century. World Health Organization,
WHO (1997), Sustainable Development and Health: Concepts, Principles and Framework
for Action for European Cities and Towns. European Sustainable Development and Health
Series, No. 1, World Health Organization Regional Office for Europe, Copenhagen.
Wolman A. (1965), The Metabolism of Cities. Scientific American 213 (3), pp.179-190.
World Bank (2004), World Development Indicators 2003. World Bank, Washington D.C.
Yeung Y.M. (2000), Globalization and Networked Societies: Urban-Regional Change in
Pacific Asia. University of Hawai'i Press, Honolulu, pp.190.
Yoon S., Araki K. (2002), Scale of Economy in Urban Energy Use. In Proceedings of
IGES/APN International Workshop on Policy Integration and Industrial Transformation
Towards Sustainable Urban Energy Use for Cities in Asia, January 23-25, 2002, Institute
for Global Environmental Strategies, Kanagawa, Japan.
Zucchetto J. (1975), Energy, Economic Theory and Mathematical Models for Combining
the Systems of Man and Nature. Case study, the urban region of Miami. Ecological
Modelling 1, pp.241-268.
The significance of energy use in cities is not very well understood, especially in Asian
countries. Although a number of studies on energy consumption from various sectors
(transport, industry, residential and commercial) have been conducted, most are from a
national point of view. Energy demand is not considered important at the city level. Also,
when analyzing energy or other social and economic systems, there are a number of factors
that need to be taken into account. Computer models can help humans as a useful tool in
analysis. This chapter deals with energy modeling as a tool designed to help with analyzing
the energy situations, particularly at the city level. It is followed by the embodied energy
analysis in selected Asian mega-cities. This analysis is based on the concept of urban
metabolism as described in chapter 2.
3.1. Modeling the energy system
When analyzing energy or other social and economic systems, it is common to use computer
models due to the factors that need to be considered such as time, modularity and
transparency of structure, the level of complexity, the level of technological detail, etc. The
purpose of integrated energy modeling is to inform discussions and decision-makers in a
coherent manner and to develop insights into energy systems for reasons such as marketing
services. It is important to look at the bigger picture in terms of such modeling. Basically, the
models are not constructed for the sake of modeling itself rather they are tools designed to
help with the analysis of some real life situations (Alfstad 2005). A model is created for a
distinct purpose and it is supposed to be applied to a particular problem. There is a wide
range of options and techniques available to energy analysts. The following sections
describe the classification of energy system models, as well as selected well-known models
used throughout the world. According to Sterman (1991) has summarized the benefits of
models as:
They are explicit. Assumptions that go into a model can be stated in written
documentation and be subject to review. For complex systems, skilled professionals
from various professions can critically assess the data that has been used.
Computer models do not make logical errors. Computer models infallibly compute the
logical consequences of the modeler's assumptions.
They are comprehensive. The models can be all-inclusive and are able to relate a
great number of factors simultaneously.
In order to decide on the use of one specific model, it is important to explain the differences
between model structure, data, and the modeling method. A model is a simplified abstraction
of a real technical energy system. The structure of the model is defined by processes and the
energy and material flows between the processes. Some models can also distinguish
different regions and economic sectors. A model needs technical and economic data to
describe processes and flows. It also needs a set of mathematical equations to describe the
behavior of the system.
The main consideration of a selected model is the method because the model can be a
simulation or an optimization. Simulation is better suited to energy demand analyses or
explorative analyses, while optimization is especially suited to calculate least-cost strategies
under certain boundary conditions. As a general rule, simulation is easier to understand and
to apply. The interpretation of the results is straightforward. Conversely, optimization is more
complex and it takes more time to get meaningful results. Interpretation of results and error
detection requires considerable experience. The results are sometimes unexpected, but give
new insights into the system's behavior (IEA 2000; Alfstad 2005). Other considerations for
the selection of a model are the possible time horizon, limitations on the number of possible
processes and flows allowed by the modeler, cost and hardware requirement of the software,
and the necessary engineering skills and computer literacy of users (IEA 2000).
3.1.1. Classification of energy system models
Models are built for various purposes and consequently have different characteristics and
applications. They are classified according to the model category, although many models
contain elements of several types and cannot unambiguously be said to belong to only one
such category (Alfstad 2005). The most fundamental distinction is related to the driver of the
system. In a Top-Down model, the functional details of the system are derived from
aggregated macro-economic parameters such as labor, capital, interest rate, etc. In contrast,
in a Bottom-Up model the driver is energy service demand and the results are produced by
the structure of a detailed system of technologies. A Bottom-Up model is thus rich in
technological detail, and aggregated values are based on a projection of the energy service
demand and the properties of these technologies. They are mostly used in evaluation of the
impacts of technology choices. In general, the available techniques to classify the energy
models fall into four broad categories, ranging from economy-wide Top-Down models to
disaggregated detailed Bottom-Up models (Harnish et al. 2002; Howells et al. 2002). The
following section describes the basic distinctions of four different models: simulation model,
optimization model, general equilibrium model, and input-output model, respectively. Table
3.1 summarizes the characteristics of the different approaches to energy modeling. The simulation model
This model is a mimic of real systems. It has zero degree of freedom. Therefore, the solution
is directly predetermined by the user. The purpose of the simulation may be foresight
(predicting how systems might behave in the future under a particular set of assumed
conditions) or policy design (designing new strategies for decision-making and evaluating
their effect on the behavior of the systems). In other words, simulation models are a “what if”
tool. They calculate what would happen under given assumptions of consumption forecasts
and policies. Such a model, however, allows the user to explore different hypotheses via
scenarios, and typically capture the area of interest at a macro-economic level. These
models are used to investigate technologically-oriented measures where macro-economic
interactions (i.e., effects of price) are less important. Simulation models have two main
components, firstly, a representation of the problem being studied and secondly, a set of
decision-making rules (Howells et al. 2002; Alfstad 2005).
A particular type of simulation model is achieved through accounting models. These models
set up an accounting balance for the flow of energy through an economy for each given time
period, usually one year. They are like assets that have equal liabilities in a financial balance
sheet; supply needs to equal consumption in an energy balance. It can be explained
mathematically by equation 3.1:
P + I − X − DS = L + C f + Cne
Eq. (3.1)
where P is total indigenous production, I is imports, X is exports, DS is stock changes, L
is losses and own consumption within the energy sector, C f is total end use, and Cne is nonenergy consumption.
51 The optimization model
This model uses technology databases containing detailed information on the intended area
of application and the relevant cost aspects involved. It is also designed to determine how to
make the best of a given situation. The optimization model has a goal or an objective
represented by a function usually referred to as an objective function, which is to be
maximized according to the alternatives given and the constraints imposed. Although this
kind of model is flexible, a high level of detail often needs to be incorporated into the model
for the simulation to be realistic. This requires information such as load curves and
technological requirement profiles, which are not always easy to acquire. A model of this
nature usually implements a form of linear programming, and tries to find an optimal solution
subject to a range of constraints. The model is usually applied when considering technologyrelated economic research questions (Howells et al. 2002; Alfstad 2005).
The optimization model is prescriptive rather than descriptive and tells the user how to create
the best situation for a related predefined goal. As opposed to the accounting model, the
optimization model has several degrees of freedom. Therefore, there may be more than one
feasible solution to the problem. In terms of forecasting and predicting actual behavior, one
should be very careful when applying the optimization model. The output of this model
should be seen as the best way of accomplishing a goal, rather than as a prediction. So,
instead of being a “what if” tool, it is a “how to” tool. For instance, if the goal is to minimize
cost, then users should use the optimization model. The optimization model is most useful
when the situation of the problem is to choose the best from a set of well-defined options.
Furthermore, a model can be static, quasi-dynamic or dynamic. A static model generates a
solution to a problem at a single point in time, neglecting any temporal development. Quasidynamic and dynamic models analyze the system over longer periods of time. In quasidynamic modeling the system is optimized for one period at a time, with the solution for one
period forming the starting point for the solution of the next period. In a fully dynamic model
the solutions for all periods are calculated simultaneously and an optimum for the entire time
horizon is found. The general equilibrium model
The general equilibrium model has a strong theoretical basis of market equilibrium, and
adopts a micro-economic view of consumer and producer behavior. Demand elasticity is
introduced and users specify a demand curve rather than a single value for the demand.
Based on the analysis, the demand level can change in response to price changes. The
production function used and the supported elasticities play a well-defined role in determining
the results. The empirical foundation, however, is weak, and it is often questionable as to
whether or not the scenarios captured are realistically represented. This type of model
focuses on macro-economic research questions of national, multi-national and global
significance (Howells et al. 2002; Alfstad 2005). The input-output model
The input-output model (IO-model) is based on macro-economic interaction matrices, energy
balances and labor market statistics. An IO-model is an accounting system showing
economic transaction. Activities are explained against the backdrop of sectoral development,
energy carrier consumption and emissions development. It is useful in analyzing the
economic relationship of links among major sectors of an economy. Using IO-analysis, it is
possible to project output requirements that must be met by the economic sectors, given a
change in output in the energy sector of the economy. IO-model is an equilibrium model in
that it assumes no surplus production or consumption. It is also a static model so that it is
difficult to incorporate changing conditions into a model of this type, especially if the
simulation period is over a long term. It is rather a snap-shot of the economy at one point in
time. This type of model focuses on the formulation of macro-economic and sectoral
research questions (Howells et al. 2002; Karkacier & Goktolga 2005).
As described above, there are several types of models pursuing different objectives, such as
demand sectors analyses, minimizing goods and services costs, measuring pollutant
emissions, etc. Furthermore, the modeling procedure must be iterated with respect to new
issues arising from the decision framework and from the analysis of intermediate results.
Such an iterative procedure must be set up to regularly update the model database, to
analyze and compare different development scenarios, and to perform a sensitivity analysis
that allows the users to identify the key-parameters of the case studies and to point out the
effect of their variables. Table 3.2 summarizes a number of well-known and user-friendly
models for energy system analysis. They are built upon different approaches e.g., simulation
or optimization, and different objectives such as comprehensive and sub-system analyses.
Table 3.1. Energy model characteristics (Howells et al. 2002).
Critical factors
Level of detail of the
energy systems
System boundaries
Flexibility in terms of
a sectoral question
Interaction and
feedback with the
entire economy
Classical question
Rationality and
market balances
Development of
reference scenarios
General equilibrium model
Medium to long term
Input-output model
Short to medium term
Usually many years
Quality of historical time
Optimization model
Short to long term
Technology energy
system with cost
One reference year
Additional conditions
Simulation model
Short to long term
Technology system with
specific general
conditions and barriers
One reference year
Quality of tech. and
Partially high
Usually one reference year
Nesting structure,
Entire economy
Entire economy
Energy system
Energy system
Not implicit, only with
Not considered
Macro-economic effects of
Sectoral effects on
environmentally economic
In principle assumed
Not relevant
Identification of priorities
for a mix of technological
Only in part, not implicitly
Dependent on level of
endogenisation, usually
considered endogenous
Aggregated at the level of
interacting structures
Technology and
Model generator
For the most part,
combined together to single
or few tech.
Closed theoretical structure
Small empirical basis, often
low level of sectoral
Implementation of
the modelling
Decisions corresponding to
nesting and elasticities
Flexibility in terms of
technically detailed
Flexibility in terms of
the scope of
Modelling supply and
Modelling demand
and consumption
Broad empirical
foundation, sectoral
disaggregation of ind.
Statistical theoretical
background, founded
solely upon historical
analyses, extensive model
preparation and
Historical analysis of
interaction matrix
Econometric estimation of
the interconnections of the
Fundamentally possible,
low for existing models
Model inherent
Implemented in different
Interlacing structure via
Endogenous, in part also
Function of production with
nesting and elasticities
Demand elasticities
Implicit for future
Plausible expert
With considerable
exogenous guidelines
As separate tech. and
explicit estimations of
each future develop.
Mostly yes
Applicable to tech. total
sys., flexible application
As separate tech. and
explicit estimation of each
future develop.
Mostly no
Also usable without
targeted entities for
Implicitly rational
optimization decisions,
strongly influenced by
Economic influences
underrepresented, based
considerably on the
quality of expert
Optimization with regard
to tech.-economic
Technological database
with optimization
High, dependent upon
the level of detail of the
tech. database
Primarily tech.
determinism of energy
Technological database,
expert knowledge
High for limited
Explicit via specific
Explicit via specific
In part, exogenous via
scenarios, in part
connected to economic
On the basis of scenarios,
coming out of economic
Table 3.2. A summary of models used in energy-environmental planning
(partially extracted from IEA 2000).
Model name
World Bank,
GLOBAL 2100,
Type of model
Other information
Energy supply and energy system
Energy information system
Energy information system
A model for the simulation of energy supply,
belongs to the ENPEP family
CO2 database
Database and technology chain analysis
Energy supply and energy system
Model for life-cycle assessment of
power systems
Modular planning instrument
Energy Flow Optimization model
Modular planning instrument
Energy-economic model
Energy-economic model
Energy-economic model
Model for the analysis of energy
conservation potential
Modular planning instrument
Model for the analysis of energy
Energy supply and energy system
Energy-economic model
District heating model
Model for the analysis of energy
Modular planning instrument
Energy supply and energy system
Energy supply and energy system
Energy system optimization model
Energy-economic model
Energy supply and energy system
Environmental Manual: a simulation model
It couples a macro-economic model with a
simulation model of energy sectors
Energy and Power Evaluation Program
Energy Technology Assessment, a dynamic
model which couples the macro-economic
MACRO with the aggregated energy system
model ETA
Computable General Equilibrium Model for
Dynamic models based on energy technology
assessment with 5 world regions
An Excel-based with database model
Long-Range Energy Alternatives Planning, a
simulation model with environmental database
Model for the Analysis of Energy Demand, a
module of the ENPEP planning tool
MARKet ALlocation model with an user support
Linked models for energy-economy analysis
A simulation model for district heating
Modèle d`Evaluation de la Demand En
Energie, a bottom-up model
Modular Energy System Analysis and Planning
Optimization Model for Energy Supply Systems
and Their General Environmental Impact
A modular simulation model
Minimization of capital and operation costs of
energy supply and demand side management
Quasi-dynamic model with
an hybrid
representation (bottom-up and top-down) of the
technologies of the industry sector
Long-term energy system simulation
USA Department of Energy
International Institute for Applied Systems Analysis, Laxenburg, Austria
International Atomic Energy Agency
European Union
Gesellschaft für Technische Zusammenarbeit mbH, Germany
United Nations, Department of Technical Co-operation for Development
Electric Power Research Institute, Palo Alto, California, USA
Organization for Economic Co-operation and Development, Paris, France
Projektinriktad Forskning och utveckling-PROFU, Göteborg, Sweden
Stockholm Environment Institute-Boston, USA
Institut für Kernenergetik und Energiesysteme, University of Stuttgart, Germany
Energy Technology System Analysis Project
Brookhaven National Laboratory
Institut Economique et Juridique d’Energie, France
Institut für Energiewirtschaft und Rationelle Energianwendung, University of Stuttgart, Germany
IKP Energy System Institute of Technology, Linköping, Sweden
Table 3.2. (continued) A summary of models used in energy-environmental planning
(partially extracted from IEA 2000).
Model name
Type of model
Other information
Energy supply and energy system
Energy-economic model
Technology assessment model
Prospective Outlook on Long-term Energy
Systems, a simulation model
A computable Price-Driven Partial Equilibrium
Model of the Energy System and Markets for
Strategic Assessment Framework for the
Implementation of Rational Energy, a
simulation model for heat and power supply
at the local and regional level for European
The Sustainable Energy Systems Analysis
Model for energy systems planning at local
and regional scale
TERI Energy Economy Simulation and
Evaluation model
The Integrated MARKAL-EFOM System, an
optimization model that produces least-cost
solutions. It is intended to replace MARKAL
which has its origin in the late 1970 and no
longer meets modern requirements and
Wien Automatic System Planning, an
optimization model
Modular planning instrument
Modular planning instrument
Energy supply and energy system
Electricity supply model
Aalborg University, Denmark
Tata Energy Research Institute, India
Energy Technology Systems Analysis Programme, Italy
3.1.2. The selected model in this thesis
As mentioned earlier, data about energy in developing countries, especially at the city level is
quite sparse. Most of the data is from the national point of view. For example, in Bangkok the
detailed breakdown of energy demand cannot be drawn from the statistical data, which is
primarily at the national level. The amount of data recorded in Bangkok is usually restricted
to the total consumption of the various energy forms e.g., diesel oil, gasoline, and electricity
(Phdungsilp 2005). One of the primary objectives of this study is to provide a user-friendly
tool for decision-makers, especially at the city level, to perform demand and supply analysis,
develop forecasts and perform impact assessments. There is a wide range of options and
techniques available to energy analysts who wish to use such models. Several models have
been developed and used in recent years for energy planning. They vary from econometric
models using linear programming to techno-economic models that analyze sectoral energy
consumption at a detailed level. In this study, the LEAP model has been used to simulate the
current energy situation for a given area and to develop forecasts for the future under certain
The LEAP model was developed by the Stockholm Environment Institute (SEI-Boston) in
1997, when SEI-Boston joined forces with five leading international energy research and
training institutions to create a new software tool for integrated energy and environmental
analysis. For more detail of LEAP development see SEI-Boston 1995 and Heaps 2004.
LEAP is a general purpose energy modeling tool that can be used for a wide variety of tasks
ranging from the preparation of energy balances and energy forecasts to policy analyses. It
is a scenario-based energy-environment model. Its scenarios are based on the
comprehensive accounting of how energy is consumed, converted and produced in a given
region or economy under a range of alternative assumptions about population, economic
development, technology, price and so on. Scenarios are developed by asking “what if”
questions. For instance what if the population growth slows down? What if transportation
systems are improved, What if renewable energy technologies are introduced? With its
flexible data structures, LEAP allows for analysis as rich in technological specification and
end-use detail as the user chooses (Heap 2004).
Important advantages of LEAP are its flexibility and ease-of-use, which allow decisionmakers to move rapidly from policy ideas to policy analysis without having to resort to using a
more complex model. LEAP is designed as a powerful decision support system (DSS) with
extensive data management and reporting capabilities to support this requirement. More than
200 government agencies, NGOs, and academic organizations have used LEAP for a variety
of tasks, including energy forecasting, GHG mitigation analysis and integrated energy
planning (Heaps 2004). Analyses have been conducted for different spatial levels, including
large cities, states and countries. It can be seen that LEAP is probably one of the best
models for this study since it allows for a transparent arrangement of data and various
flexible scenarios and energy system configurations can be developed. This model is useful
in both developed and developing counties, even if data at the city level is not readily
3.1.3. LEAP model framework
LEAP is an accounting tool that balances production and consumption of energy in an
energy system model. It is not a model of a particular energy system, but rather a tool that
can be used to create models of different energy systems. It supports a wide range of
methodologies, including simple trend projections, econometric forecasts, end-use analyses
and engineering-based simulations or a combination of these techniques. LEAP is
deterministic, in the sense that all of the outcomes are specified by the user. As such, it is a
tool that calculates the implications of a set of assumptions and tells the user what would
happen if these were true. Scenario analysis is at the heart of using LEAP, and LEAP
provides a wide range of tools for quickly constructing, evaluating and comparing alternative
policy scenarios in terms of their energy requirements, their social costs and benefits, as well
as their environmental impacts.
Based on the assumptions provided by the user, LEAP balances energy flow equations,
thereby identifying energy transformation and primary energy supply requirements. The
requirements are back-calculated from a set of final energy demands, which form the fixed
side of the first set of the equations in the accounting process. The entire system can be
included within the model and the level of detail is decided by the user.
The central concept of LEAP is an end-use driven scenario analysis. Additionally, the model
includes the technology and environmental database (TED), which provides extensive
information on the technical characteristics, costs and environmental impacts of a wide range
of energy technologies. The LEAP framework mainly contains five modules: energy demand,
transformation, resource analysis, cost-benefit analysis and non-energy sector effects (SEIBoston 1995; Heap 2004). Figure 3.1 shows the overview of the LEAP model calculation
flows. The following sections give a brief description of different modules of LEAP.
(Pollutant Emissions)
Environmental Loadings
Integrated Cost-Benefit Analysis
Non-Energy Sector
Emissions Analysis
Figure 3.1. LEAP calculation flows (adapted from Heaps 2002). Demand module
Demand analysis is disaggregated in a hierarchical tree structure of levels. It can range from
highly disaggregated end-use oriented structures to highly aggregated analyses. Typically, a
structure would consist of sectors: households, industry, transport and agriculture, each of
which might be broken down into different sub-sectors, end-uses and fuel-using devices.
Users can adapt the structure of the data to their purposes based on the availability of data,
the types of analyses they want to conduct and unit preference. It is also possible to create
different levels of disaggregation in each sector (Heaps 2002). Figure 3.2 illustrates the
example of the tree structure in the energy demand module. This example shows a typical
approach to disaggregated demand data structure in four levels representing sectors, subsectors, end-uses and devices. In the energy demand module the energy intensity values,
along with the type of fuel used in each device, are required to estimate the energy
requirements at sector, sub-sector and end-use levels.
(number of HH)
(number of bld)
Energy Intensity
Inner BMA
(% share of HH)
Outer BMA
(% share of HH)
(% share of HH)
(% share of HH)
(% share of HH)
Compact fluorescent
(% share of HH)
Electricity (kWh/HH)
(% share of bld)
(% share of bld)
(% share of bld)
(% share of bld)
(% share of bld)
High efficiency
(% share of bld)
Electricity (kWh/m2)
Food & Beverage
(% share of GPP)
(% share of GPP)
(% share of GPP)
(% share of GPP)
(% share of GPP)
High efficiency
(% share of GPP)
Electricity (kWh/Baht)
(% share of pass-km)
Mass transit
(% share of pass-km)
(% share of pass-km)
Urban taxi
(% share of pass-km)
(% share of pass-km)
High efficiency
(% share of pass-km)
Gasoline (liter/pass-km)
Electricity (kWh/HH)
Electricity (kWh/m2)
Electricity (kWh/Baht)
Gasoline (liter/pass-km)
Notes: HH stands for household, BMA stands for Bangkok Metropolitan Area, bld stands for
building, and GPP stands for Gross Provincial Product.
Figure 3.2. Example of the tree structure in the energy demand module.
Similarly, the module offers a choice of methodologies that can be applied to the energy
demand analysis:
Activity level analysis, which consists of either final energy demand analysis or useful
energy demand analysis, in which energy consumption is calculated as the product of
an activity level and an annual energy intensity.
Stock analysis, in which energy consumption is calculated by analyzing the current
and projected future stocks of energy-using devices with the annual energy intensity
of each device.
Transport analysis, in which energy consumption is calculated as the product of the
number of vehicles, the annual average mileage (i.e., distance traveled per vehicle),
and the fuel economy of the vehicles (e.g., liters per km).
In fact, users can mix and match these different methodologies within each set of data. For
example, useful energy analysis can be applied to the analysis of industrial and commercial
heating, while final energy analysis can be employed for all other sectors. In each case,
demand calculations are based on a disaggregated accounting for the various measures of
social and economic activity such as the number of households, vehicle-km, tons of industrial
production, commercial value added, etc. These activity levels are multiplied by the energy
intensities of each activity. In addition, the emission factors of different pollutants in the TED
module are linked to the device level to appraise the environmental emission from energy
utilization during the planned horizon. TED contains emission factors for hundreds of energy
consuming and energy producing technologies, including the default emission factors
suggested by the Intergovernmental Panel on Climate Change (IPCC).
In the demand module, LEAP requires data for the base year at least (and any for future
years). Then, using the functions, such as interpolation, extrapolation and growth rate
method, the future energy demand and emissions are estimated for the other years.
59 Transformation module
Before being used, the primary energy has to be transformed through secondary energy into
final energy. In a transformation analysis, the processes with their efficiencies and losses are
incorporated within the model, in order to calculate the total amount of primary energy that is
required to produce the final energy demand. This module simulates the conversion and
transportation of forms of energy from the point of extraction of the primary resources and
imported fuels all the way through to the point of final consumption. The different levels of
complexity within the transformation processes can be distinguished from the simple to the
more complex processes with multiple inputs, outputs and efficiencies (Heaps 2002).
The general structure of transformation modules is shown in Figure 3.3. A module is a
branch representing an energy conversion sector such as the generation of electricity, oil
refining, district heating, transmission, distribution, etc. Each module contains a number of
processes that represent the individual technologies, which convert energy from one form to
another or transmit or distribute energy. Each module has one or more output fuels; the
module’s processes are dispatched to try and meet any of the requirements for its output
Auxiliary Fuel Use
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Feedstock Fuel
Auxiliary Fuel Use
Fuel (e.g Heat)
Figure 3.3. The general structure of transformation modules (adapted from Heaps 2002).
The transformation calculations are driven by the results of the demand analysis. Similar to
the demand module, alternative scenarios can be used to represent different future
transformation configurations, reflecting alternative assumptions about policies and
technologies. This feature is useful when implementing renewable energy supply policies, for
instance, the target of having 15% of electricity generated from renewable energy sources by
60 Resource analysis
The resource analysis is used for entering data on the availability of primary resources,
including fossil and renewable resources, as well as information on the costs of indigenous
production and imports, and exports of both primary resources and secondary fuels.
Resource branches are always sub-divided into two categories: primary resources and
secondary fuels. Fossil fuel resources require the total available reserve of the resource,
while renewable energy resources require only that which is available from the resource.The
availability of each resource is specified for the area as a whole. Resource availability can
build up from a more disaggregated analysis, in which the total availability is sub-divided by
region or some other type of classification. In terms of renewable resources analysis, this
approach is useful for keeping track of resources such as biomass. It provides a
comprehensive framework for the whole of the energy flow from resources (biomass) through
conversion to end-use consumption (Heaps 2002).
An example of the structure for resource analysis is given in Figure 3.4 as a wood resources
structure in the biomass model. Similar to demand analysis, the study area is divided into
sub-areas, zones and land use types. The area is divided into two regions: Northern and
Central, and both regions have two zones: forest and non-forest, each with respective land
use types. For all land use types the acreage, the productivity (i.e., standing stock and
average annual yield of wood or crop productivity) and the access fraction has to be
specified. The access fraction represents the maximum fraction of annual yields and stock
that can be used for energy purposes.
Land use type
Stock, yield & access fraction
Natural forest (14,000 ha)
Forest plantations (6,500 ha)
Watershed forest (3,200 ha)
125 ton/ha, 1.2 ton/ha/yr, 20%
60 ton/ha, 4.0 ton/ha/yr, 40%
100 ton/ha, 1.0 ton/ha/yr, 15%
Village land (4,500 ha)
Fruit plantations (6,000 ha)
Crop land (21,000 ha)
37 ton/ha, 1.5 ton/ha/yr, 80%
45 ton/ha, 2.5 ton/ha/yr, 60%
10 ton/ha, 0.5 ton/ha/yr, 80%
Forest plantation (9,500 ha)
55 ton/ha, 3.5 ton/ha/yr, 40%
Village land (3,300 ha)
Fruit plantations (5,700 ha)
Crop land (17,500 ha)
City (1,300 ha)
31 ton/ha, 1.4 ton/ha/yr, 80%
53 ton/ha, 2.7 ton/ha/yr, 60%
12 ton/ha, 0.6 ton/ha/yr, 80%
1 ton/ha, 0.1 ton/ha/yr, 50%
Figure 3.4. Example of a wood resources structure (adapted from SEI 1995). Cost-benefit analysis
LEAP can perform integrated social cost-benefit analysis on the scenarios created. The costbenefit analysis calculates the costs within each part of the energy system: the capital and
operating maintenance costs of the purchase and use of the technologies in the demand and
transformation system; the costs of extracting primary resources and importing fuels, and the
benefits from exporting fuels. In addition, it allows examination of the environmental
externalities by assigning costs to the emission of pollutants and any other direct social and
environmental impacts of the energy system.
Cost-benefit analysis is based on the social costs of resources, not the final prices of the
energy in respect to the consumer. It centers on the costs to society of a given set of actions.
It does not take the perspective of a particular consumer or producer. Social costs and prices
need to be the same. For example, electricity prices may differ from the costs of producing
electricity, due to subsidies, transfer payments and market distortions.
In the LEAP model, the cost-benefit analysis is not intended to provide an analysis of
financial viability. Instead, it helps identify a range of socially acceptable policy scenarios.
Users specify a costing boundary e.g., a whole system including resource costs or a partial
system and the costs of the fuels delivered to a module. Figure 3.5 shows the specifying
limited boundary in different modules. Cost-benefit analysis calculates the Net Present Value
(NPV) of the differences in costs between scenarios. NPV sums all costs in all years of the
study discounted to a common base year (SEI-Boston 2005).
(costs of saved energy,
device costs, other non-fuel
(Capital and O&M costs)
Primary Resource Costs
Delivered Fuel Costs
Externality Costs
Figure 3.5. An example of a costing boundary in cost-benefit analysis
(adapted from SEI-Boston 2005). Non-energy sector effects
LEAP also has an option to create inventories and scenarios for non-energy related effects.
Typically, this feature is used for inventories and scenarios of GHG emissions and as a
complement to the analysis of energy sector emissions and mitigation measures such as
cement processing emissions, landfill emissions, etc.
Unlike an energy demand analysis in which total effects are calculated as the consumption of
the energy as product and an emission factor, in the non-energy sector users specify total
annual effect loadings. Users also have the ability to analyze any number of different
emission scenarios.
3.2. Estimation of urban energy use
Typically the estimation of energy demand is performed centrally. Energy production is
normally accomplished in large sized facilities, which are intended to provide energy for large
areas and for a long time horizon. Centralized planning goes together with a macroscopic
view of an energy system. Energy representations are still highly aggregated and do not
examine possible variations within the spatial distribution of the energy demand and of the
energy supply sources. Since one of the objectives of this thesis is to try to integrate the
maximum use of renewable energy at the city level, a much more detailed approach to
energy demand needs to be developed in order to effectively integrate renewable energy into
an urban energy system. In addition, the need for a city perspective within an urban energy
system is intended to identify the most advantageous technologies in order to maximize
economic benefits and minimize environmental damage. This section presents an overview
of energy use in cities i.e., the estimation of useful energy demand, final energy consumption
and demand forecasting. Considering useful energy the benefits are twofold: it will provide
opportunities to integrate renewable energy into an urban energy system; and it will be useful
for further study, e.g., optimization modeling such as MARKAL, which requires useful energy
demand as the primary driver in the model.
3.2.1. Energy demand estimation
In many countries the detailed breakdown of the city's energy demand cannot be drawn from
national statistics, this is particularly so in developing countries. The amount of data recorded
at the city level is usually restricted to the total consumption of the various energy forms.
Therefore, it is useful to develop a Bottom-Up approach in order to attain the required degree
of detail. In this study the detailed disaggregated energy demand will be further used to
illustrate the final energy demand, as well as to calculate indirect energy consumption.
The rationale behind the Bottom-Up approach is that there is no need for electricity, fuel, oil,
gasoline, etc. What is needed are energy services for accomplishing various activities which
are undertaken in each sector. This service is provided by means of an end-use device e.g.,
boiler, engine, etc., which transforms fuels, electricity or other energy forms into useful
energy (Finon & Lapillone 1983). Such an approach not only provides the basis for
establishing energy balances but also allows for a better approximation of the likely
penetration of renewable technologies in end-use activities such as solar collectors for water
heating (Sarafidis et al. 1999). However, energy services are immaterial services for the
procurement of what energy is actually used for; for example, the transportation of a person
or goods from A to B or the provision of air-conditioned rooms or a well-lit work place, etc.
Useful energy can be defined as the energy equivalent to the work actually performed in the
process of providing energy services. For example, the amount of heat released by a heating
system in a room that is to be kept at a desired temperature, the drive-power actually applied
to accelerate a vehicle or the light emitted by a lamp. Useful energy can be calculated from
data on final energy use by multiplying the amount of final energy used for a certain
appliance or process by the efficiency of this appliance or process.
Usually, the estimation of energy demand is at first common disaggregation demand and
distinct end-use activities, as shown in Table 3.3. For each end-use activity energy
requirements are estimated by means of technical models relating present or future energy
demand to a number of known or predictable key parameters. These parameters can be
classified by the following categories: (a) structural characteristics such as population, type,
number and magnitude of dwellings, (b) natural characteristics such as climatic conditions
and their seasonal variations, (c) technological characteristics such as the type and efficiency
of the end-use devices, (d) behaviour characteristics such as lifestyle and mode of
technologies used (NTUA 1996; Sarafidis et al. 1999). The procedures for estimating the
useful energy demand as shown in Table 3.3 can be explained as follows (Stoll 1989):
Space heating method: used for estimating energy needs for space heating within the
building sector. It relies on a simplified model of the thermal heat balance of
dwellings. Heat losses are calculated by a set of equations, which represent heat
losses through the building shell and air infiltration. The key parameters used in the
estimation procedure are the number and characteristics of buildings.
Appliance saturation method: used for estimating energy needs for air-conditioning
and other electrical appliances. The key parameters for the implementation of this
method are the number of households and the stock of appliances (saturation level
per type of appliance; technical characteristics, mode of usage).
Floor space method: used principally for the electricity uses of the tertiary sector. It is
similar to the appliance saturation method, except that end-use activities are defined
on the basis of the building’s cross-sectional area. Electricity needs (e.g., for lighting,
air-conditioning, etc.) are expressed on a per unit of area basis. The estimation in the
agricultural sector can be based on cultivation areas instead of building areas.
Thermophysical law: is particularly used for water heating. The key parameters used
are population data and other users of hot water such as numbers of tourists, water
consumption per capita, and comfort standards (temperature of hot water).
Statistical records of energy data: some statistical data for fuels and electricity
consumption are usually available at the city level (e.g., the transportation and
industrial sector), so that energy demand can be drawn directly from statistical
However, the estimation of energy demand always depends on how much detailed primary
data can be obtained from statistical records, and if so, does the data available match the
objective. The useful energy demand established through LEAP is based upon the
projections of relevant drivers such as population growth, GDP growth, number of
passengers-kilometers, number of buildings, etc.
Table 3.3. Energy demand disaggregation and the estimation methods
(compiled from Sarafidis et al. 1999; Howells et al. 2002).
End-use activities
Space heating
Useful energy
Useful energy
Water heating
Useful energy
Useful energy
Space heating
Water heating
Useful energy
Useful energy
Useful energy
Useful energy
Useful energy
Useful energy
Appliance saturation
Floor space or appliance
saturation or space heating
Thermophysical law or floor
Floor space or appliance
Appliance saturation
Appliance saturation
Floor space
Floor space
Floor space
Floor space and/or base year
Thermophysical law or
appliance saturation
Thermophysical law or floor
Floor space
Floor space or appliance
Appliance saturation
Appliance saturation
Base year data or appliance
Base year data or appliance
Base year data
Base year data
Base year data
High temperature Useful energy
process heat
Useful energy
Water heating
Useful energy
Useful energy
Industry (and Thermal uses
Electric uses
Useful energy
Useful energy
Useful energy
Land pass-km
Air pass-km
Land ton-km
Land passenger
Air passenger
Land freight
Useful energy
In the transport sector, various modes of transport, such as petrol cars, diesel trucks or
electric trains, convert fuel into an energy service. It is more useful to consider this service
rather than the useful energy. The conversion ratio between the fuel and the service
delivered is therefore referred to as intensity, rather than efficiency (Howells et al. 2002). The
services delivered in the transport sector are passenger-kilometer and ton-kilometer for the
passenger and freight sectors, respectively.
Hence, the final energy consumption can be estimated from the useful energy demand of the
technical models compiled, with the fuels available for providing the energy required, the
share of the various conversion technologies used in each activity, and the efficiency of each
ECi, j = ∑
E j pi, j,k
Eq. (3.2)
ei, j,k
where EC is the energy consumption, E is useful energy demand, p is share of
technologies, e is the efficiency of technologies, i is fuel, j is end-use activity, and k is
The energy demand is estimated for a base year for which all the data needed for calculating
useful energy demand and final energy consumption are available. It is necessary to
compare the resulting aggregated figures of consumption with existing statistical records in
order to check the consistency of the modeling. This calibration procedure provides a safe
basis for the energy modeling and allows for more reliable future estimates.
Energy demand forecasting typically starts by estimating for the examined time horizon the
values of considered key parameters included within the technical models, as well as the
expected changes in the shares of the conversion technologies and in their efficiencies. In
the transportation sector for example the modal shift from private cars to mass transport is
frequently considered.
3.2.2. Representing an urban energy system
For the purpose of illustrating the modeling process, the energy system can be visualized by
simulating the flow of energy in various forms (energy carriers) from the source of supply
through the transformation systems to the demand devices, which satisfy end-use demands.
In this study a Reference Energy System (RES) type of representation was chosen. RES is a
format for the graphical display of energy balances. It was first developed at Brookhaven
National Laboratory (BNL), USA. It has been used for energy assessments and energy policy
and planning studies throughout the world. It is a way of representing the activities and
relationships of an energy system. RES describes the flow of energy from the sources to the
final uses. It shows all flows of energy from the primary energy supply through central
conversion and different forms of distribution, and then onto its final use within different
sectors. Additionally, the RES usually contains useful information on energy demand and
even energy services. A principle physical flow of a RES is presented in Figure 3.6.
However, RES is not a geographical representation of the energy system. RES
representation, facilitates the learning process for people who are involved in planning, since
it clearly shows how different parts of the energy system interact with each other (Kanpal
1998; IEA 2000).
Figure 3.6. A physical representation of the RES (adapted from IEA 2000).
Through RES it is possible to see how energy flows and how energy conversion technologies
influence the fuel-technology chains in an energy system. This overall perspective is
particularly important when one evaluates the demand side, energy conservation
technologies (i.e., the balance between supply and conservation measures) or the costefficiency of a proposed investment to control emissions. While the RES is a graphical
representation of energy flows within an energy system, an energy balance contains the
values of all energy flows. These can be included in the graphical form or be presented in
separate tables. In fact, RES may contain more conversion levels like distribution, end use
technologies and useful energy demands, which are not normally included in an energy
balance table. It is preferred that RES it is built-up according to certain practical
recommendations (IEA 2000):
Sources and primary energy supply: the RES begins at the far left of the diagram with
the input flows of energy e.g., oil, natural gas, coal, petrol and imported electricity.
Processes: next follows the processes that modify the fuels e.g., oil refining and the
preparation of pellets from the biomass.
Conversion technologies: next the flow of energy enters the large energy conversion
technologies e.g., electricity production plants, district heating and cooling plants, and
combined heating and power plants.
Distribution systems: large scale conversion is followed by distribution systems for the
different energy forms e.g., electricity, district heating and cooling, and natural gas.
End-use technologies: the next step is the small-scale energy conversion
technologies e.g., oil fired boiler for multi-family houses, solar heating systems for
single family houses, electrical appliances, petrol fuelled cars; and small scale
combined heating and power plants. All of these technologies are supplied by final
energy sources.
Useful energy demand: the energy that is needed for different kinds of applications
e.g., space heating and cooling, lighting and cooking.
Figure 3.7 shows an example of part of the RES in an energy system. In addition to the
conventional energy balance and the RES, there is another representation tool that is used
frequently: the Sankey diagram. In the Sankey diagram the flows of energy from input of
energy to final use is illustrated by lines of different width, where the width is proportional to
the size of the energy flow. This gives an immediate feeling of the relative importance of the
energy flows. The energy system must often be more simplified than the RES, in order to fit
all flows into one diagram. The RES representation will generally be the basis for further
analyses. For presentation purposes, it could very well be supplemented by energy balances
or Sankey diagrams. Another way is to use it in conjunction with optimizing models such as
MARKAL. RES can be used to show different aspects of the energy system. It can also cover
the total energy system but in order to make this possible the RES should be somewhat
simplified. Otherwise, it will be too large and complicated for practical use (IEA 2000).
Figure 3.7. Example of part of the RES (redrawn from IEA 2000).
n y
r s
il i o rg
O d ia t n e
w e Ga
P o te d
te R a r
o r
p o l a r cl e a l d r p o
I m S o N u Co Hy I m
Primary Energy
of Oil
of Gas
District Heating Genearation
Hydro Power
Power Plant
Coal Power
Power Generation
ea y
o r s t r i c ct r i
il f a i s
O G D E le
Final Energy
Oil Central
El. Heat
El. Storage
Useful Energy
Low energy
of a ce
re s p
Energy Service
3.3. Direct and indirect energy demand analysis
There has been limited research at the level of the city and most studies consider direct or
end-user energy consumption. The local effects on resource use in cities also need to be
better understood. Since the function of a city is to serve the lives of their residents, indirect
energy use in city or energy embodied in the consumption of goods and services by its
residents can be regarded as being as important as direct energy use (Phdungsilp 2005).
However, physical models of a city are extremely complex and have difficulty in dealing with
boundary issues i.e., indirect resource requirements. The urban metabolism approach, as
described earlier, can be used in order to obtain a comprehensive energy breakdown.
In this study, the concept of urban metabolism has been applied to analyze the energy
demand (direct and indirect) at the level of the city. The concept of metabolism implies the
processes of physical exchange or energy and material flows between human societies and
their natural environment, as well as the internal energy and material flows of human
societies (Ayres & Simonis 1994; Fischer & Huttler 1998). From the perspective of the
metabolism concept, two types of problems can be addressed: resource scarcity on the input
side and emission on the output side. It is obvious that the metabolism of a city can only be
clarified if both the energy and material flows are considered. This thesis is not intended to
study full urban metabolism, but rather focus on analyzing the energy flow within a city. In
order to do so, the concept of embodied energy is employed. Embodied energy analysis has
been used by several authors to distinguish between the indirect energy use for the
production of goods and the direct energy use (van Engelenburg et al., 1994; Subak 1995;
Coley & Goodliffe 1997; Phdungsilp 2005). Indirect energy use can also imply the economic
dependency of the city upon the outer regions and especially highlight the environmental
load displacement upon outer regions. Additionally, the ratio of direct energy use or indirect
energy use can also provide information about economic structural change within the city.
Furthermore, it can be used to determine the associated environmental impacts and to
predict some of the consequences of social and technological development.
Indirect energy consumption can illustrate how a city relies upon the outside. Certainly, the
direct energy demand can be taken as an indicator of how a city relies upon factors from the
outside. However, as energy demand is a proxy indicator, indirect energy demand may
provide some very interesting information that cannot easily be drawn from direct energy
demand (Kaneko et al. 2003).
A material-related indicator can clarify the dependency upon outside factors. The
direct energy demand of each sector, as a pure energy-related indicator, can give
little information about material reliance outside of the city. But this indicator can
contribute to this query in some sense.
Indirect energy demand is directly related to the characteristics of the sectors
themselves, which can easily clarify the sectoral differences and characteristics in
terms of energy or material. Consequently, implications regarding the industry
structure transition can be drawn.
Since indirect energy demand is a reflection of the industrial structure of the city, a
comparison among cities at different stages of development can indicate how the
evolution of industry influences the energy reliance of the city.
Indirect energy demand illustrates how a city relies upon the outside or is relied upon
by the outside. This indicator serves to clarify the role of the city in the sense of
energy consumption beyond the geographical scope of the city.
This section will firstly propose a method to account for indirect energy consumption based
upon the concept of urban metabolism, which provides a framework for understanding the
interactions between economics and urban energy use, and then uses this method for the
estimation of the embodied energy analysis. Finally, it also makes an attempt to compare
direct and indirect energy demand of selected Asian mega-cities based upon the available
3.3.1. Model for indirect energy consumption based on embodied energy analysis
To simulate the direct and indirect energy flows within the metabolism perspective, the
concept of embodied energy is employed in this research. The term embodied energy use
means the total amount of energy needed directly or indirectly to make any product or
service (Bakshi 2000). According to the laws of thermodynamics, the metabolism model
implies that anything that comes into a system (city) must pass through it and that the
amount of waste is therefore dependent on the amount of resources required. Without
violating laws there should be an energy balance for each production sector within the
system. This is simply demonstrated in Figure 3.8.
Direct energy input
Indirect energy input
(embodied by the goods
produced within the city)
Sector j
Indirect energy input
(embodied by the goods
produced outside the city)
Export to outside
Indirect energy output
(embodied by the products
produced in this sector)
Consume domestically
Figure 3.8. Energy balance for sector j on a city scale (adapted from Kaneko et al. 2003).
Due to the limited data of the disposal and recycle parts, this study neglects these two parts,
implying that there is no material recycling or disposal by the sectors. The basic closed
embodied energy model can therefore be expressed in Eq. 3.3 as follows:
E j + ∑εi x ij = ε j X j
Eq. (3.3)
where E j is the direct energy input into sector j , εi is the embodied energy per unit
production of sector i within the city, x ij is the goods and services flow from sector i to
sector j within that city, ε j is the embodied energy per unit production of sector j within the
city, and X j is the gross production of sector j .
Solving the Eq. 3.3 for ε can be expressed in matrix form. The embodied intensity or
embodied energy per unit production of each sector can therefore be calculated as Eq. 3.4:
ε = E * ( X − x )−1
Eq. (3.4)
where (X-x)-1 is referring to (I-A)-1, Leontief’s Inverse Matrix - a fundamental matrix of an
Input-Output (IO)-analysis that identifies any ripple effects among economic sectors. Detailed
IO-analysis is looked at in the next section.
As mentioned earlier, Equations (3.3) and (3.4) represent a closed system. However, cities
are open systems that exchange goods and services with other cities and countries. For a
city level analysis, the establishment of appropriate boundary conditions can be problematic.
Specifically, goods and services consumed within a city can be classified as : (a) local goods
produced in the city, (b) goods imported from other areas of the country, and (c) goods
imported from other countries. This makes it necessary to establish a model with not only
boundaries between the city and the country, but also between the country and the rest of
the world. Based on the available data it is difficult to calculate the embodied energy of
goods imported from other countries. In this study, the embodied energy is assumed to be
equal between imported goods from other areas in the country and imported goods from
other countries if they are in the same sector, but different between local and imported
goods. Therefore, this becomes a single-boundary model with only a boundary between
locally produced goods and goods produced from outside of the city (Kaneko et al. 2003;
Phdungsilp 2005). When distinguishing between local and imported goods, Eq. 3.3 can be
changed as follows:
E j + ∑εi x ij + ∑ζ i x ij = ε j X j
Eq. (3.5)
where ζ i is the embodied energy per unit production of imported goods, and x ij is the flow of
imported goods and services from sector i to sector j .
In fact, the second term in Eq. 3.5 represents the embodied energy of intermediate-input
local goods, and the third term represents the embodied energy of intermediate-input
imported goods.
The model in Eq. 3.5 can be applied to a competitive-imports type IO-table that clearly
distinguishes between local and imported goods, but not to a non-competitive-imports type
IO-table that does not distinguish between local and imported goods. For example, city level
IO-tables in Japan as well as in China are usually of the competitive-imports type and rarely
of the non-competitive-imports type. Therefore, the degree of local self-sufficiency of each
sector is used to distinguish between local and imported products (Kaneko et al. 2003). Then
the energy balance is:
i =1
i =1
E j + ∑ ε iγ j xij + ∑ ζ i (1 − γ j ) xij = ε j X j
Eq. (3.6)
where γ j is the degree of self-sufficiency of product j , which can be expressed as
γ i = X j /( X j + M j ) while M j stands for the import of product j .
The essence of the embodied energy is a kind of indirect reflection on the behavior that
follows after the consumption of direct energy. To emphasize the indirect energy
consumption, meaning where the goods are finally consumed, the energy embodied in those
goods should be accounted as a kind of indirect energy consumption or indirect energy
demand. In terms of indirect energy consumption, the end-user should somehow take
responsibility for the energy consumption and corresponding emissions.
3.3.2. Calculation method for embodied intensity based on Input-Output (IO) analysis
The IO-analysis technique is applied to recur the energy related economic activities. An IOmodel is useful in analyzing the economic relationship of links among the major sectors of an
economy. This model is a static model or a snap-shot of the economy at a fixed point in time.
In IO-tables, exchanges of goods and services among industrial sectors are presented in
matrix form. Most of the tables actually available are specified in currency units. Energy and
resource flows among industries can be analyzed on the assumption that goods are
transferred in direct proportion to their monetary value. IO-tables have been applied to the
analyses of environmental issues. More information and details on IO-theory can be found in
articles by Leontief (1953), Duchin (1992), and Dixon (1996).
In principle, a standard static IO-model is used to calculate the gross output and factor inputs
required to satisfy a given final demand. Alternatively, final demand can be deduced for a
given gross output. In most cases an IO-model is based on an IO-table of the general form
shown in Table 3.4.
Table 3.4. Principle of an input-output (IO) table.
Value added
Total output
Intermediate demand or intermediate
Final demand
Total output
As shown in Table 3.4, an IO-table consists of four main parts, including transaction table
(xij), value added (Vj), final demand (Yi), and total output (Xj). Similarly, this kind of model has
three basic tables; the transaction table, technical coefficients, and the interdependence
coefficients matrix.
The transaction table is constructed to present the relationships among major sectors of an
economy. Each producing sector within the economy has a certain amount of output, which
may be used within the sector, sold as inputs to other producing sectors, or sold as final
demand to the consumer. The transaction table of sectors can be written as a simultaneous
set as follows:
x11 + x12 + x13 + Y1 = X 1
x21 + x22 + x23 + Y2 = X 2
Eq. (3.7)
x31 + x32 + x33 + Y3 = X 3
Where xij is a sale from sector i (rows) to sector j (column), Yi is a sale from sector i to
final demand, and X i is the total output of sector i .
Eq. (3.7) represents an example of three sectors within the economy, so that it explains the
inter-industry relationship among sectors as aij = xij / X j . This expression may be
rearranged to x ij = aij X j , which is interpreted to mean that the level of sales from sector i to
sector j depend upon the level of output in sector j .
The technical coefficients show the value of the input purchased from all sectors in the
economy per monetary unit (e.g., Baht, SEK, Euro, etc.) of output in a particular sector.
Technical coefficients can be derived by dividing all of the entries in each sector’s column by
the total output of that sector as shown in the following equation:
⎡ a11
⎢ 21
⎢⎣ a31
a 22
a13 ⎤
a 23 ⎥⎥
a33 ⎥⎦
Eq. (3.8)
Let xij symbolize the value of sales from sector i to sector j , and Xj the total output of
sector j . Then the technical coefficient or input coefficient for each sector is calculated by
using the following equation:
aij =
x ij
Eq. (3.9)
Then, the Eq. (3.9) can be written as follows:
x ij = aij X j
Eq. (3.10)
Substitution Eq. (3.10) into Eq. (3.7), rewrites the equations for the producing sectors as
shown below:
a11 X1 + a12 X 2 + a13 X 3 + Y1
a21 X1 + a22 X 2 + a23 X 3 + Y2 = X 2
a31 X1 + a32 X 2 + a33 X 3 + Y3 = X 3
Eq. (3.11)
Eq. (3.11) shows the interdependence of each sector on all of the others because it reveals
that the level of output in any sector is dependent upon the level of output in other sectors,
the input requirements of each sector, and the level of its final demand. Treating the final
demand (Yi) as exogenous to the sector, then:
X1 − a11 X1 − a12 X 2 − a13 X 3
= Y1
−a21 X1 + X 2 − a22 X 2 − a23 X 3
−a31 X1 − a32 X 2 + X 3 − a33 X 3
= Y2
= Y3
Eq. (3.12)
(1 − a11 ) X 1 − a12 X 2 − a13 X 3
− a 21 X 1 + (1 − a 22 ) X 2 − a 23 X 3
− a31 X 1 − a32 X 2 + (1 − a33 ) X 3
= Y1
= Y2
= Y3
Eq. (3.13)
or matrix notation
− a12
− a13 ⎤
⎡(1 − a11 )
⎢ −a
(1 − a 22 )
− a 23 ⎥⎥ *
⎢⎣ − a31
(1 − a33 )⎥⎦
− a32
⎡Y1 ⎤
⎡ X1 ⎤
⎢ X ⎥ = ⎢Y ⎥
⎢ 2⎥
⎢ 2⎥
⎢⎣Y3 ⎥⎦
⎢⎣ X 3 ⎥⎦
Eq. (3.14)
or simply
A* X = Y
Eq. (3.15)
In Eq. (3.15), X is the output matrix and A and Y are the coefficient matrix and the final
demand matrix, respectively. The aij elements of the matrix ( A* ) of Eq. (3.15) have the same
values as the technical coefficients calculated from Eq. (3.9). Then, matrix A* is the
difference between two matrices, an identity matrix (I) minus the matrix of technical
coefficients. Therefore, Eq. (3.15) can be written as:
( I − A) X = Y
Eq. (3.16)
where ( I − A) = A* . The matrix ( I − A) is called the Leontief Matrix and then Eq. (3.16) may
be rearranged as:
X = ( I − A)−1Y
Eq. (3.17)
Eq. (3.17) is the solution to the IO-model. The matrix ( I − A)−1 is so-called the Leontief’s
Inverse Matrix or simply the Inverse Matrix, and is a fundamental matrix for IO-analysis. It
identifies any ripple effects among economic sectors. The elements of the inverse matrix
measure the direct and indirect output levels from each sector of the economy necessary to
satisfy the given levels of final demand.
It should be noted that there are two different types of prices in IO-tables: the market
shipment price of the sector that has produced the relevant item, the so-called producer
price, and the other called the purchaser price, which is based upon the sum of producer
price, domestic transportation charges and trade margins. Most calculated embodied
intensities are expressed on the basis of producer price (Nansai et al. 2002). An IO-table as
described above is an accounting system that shows economic transaction, an application of
the IO-technique can allow one to trace through an economy and the direct and indirect
energy impact of changes in the final demand. According to IO-model, the embodied
intensities can be obtained using producer prices in IO-tables, and others are obtained using
purchaser prices. In this thesis, the embodied analysis is calculated on producer price, which
is based on the factory gate value of a product. In the following section the calculation
method of embodied intensities based on IO-tables is presented.
The treatment of imports in the IO-tables has a significant effect on the basic IO-model,
regardless of whether it is based on producer price or purchaser price. If the environmental
burdens (of energy use) related to the production of imported products are assumed to be
identical to those of the very same domestic products, and are referred to the basic closed
embodied energy model in Eq. (3.3), the total energy requirements generated by processes
in sector j are the sum of the direct and indirect uses from intermediate demands that satisfy
the following equation (Miller & Blair 1985; Nansai et al. 2002):
ε 1 x1, j + ε 2 x 2, j + ... + ε k x k , j + ... + ε n x n , j + E j = ε j X j
Eq. (3.18)
where ε j is the embodied intensity and indicates energy use in sector j , generated directly
or indirectly per unit production (e.g., toe/million US$) producer price, x i, j is indicated input
quantity from sector i into sector j (monetary unit), E j is the direct energy use by activities
in sector j (energy unit), and X j is domestic production of sector j (monetary unit).
Dividing both sides of Eq. (3.18) by Xj, and using input coefficient or technical coefficient, aij ,
and direct energy use per unit production ej gives Eq. (3.19):
a1, jε1 + a2, jε2 + ...ak , jεk + ...+ an , jεn + e j = ε j
Eq. (3.19)
where ai, j equals to x i, j / X j , and e j equals to E j / X j .
Expressing these with vector and matrix for sector j = 1 to n gives Eq. (3.20) as follows:
(ε 1
⎛ a11
⎜ a21
ε 2 ... ε n )⎜
⎝ n ,1
an , 2
L a1,n ⎞
L a 2,n ⎟
+ (e1
O M ⎟
L an ,n ⎟⎠
... en ) = ( ε 1 ε 2 ... ε n )
Eq. (3.20)
Then, expressing Eq. (3.20) with embodied intensity vector ε , direct energy use per unit
production e , and input coefficient matrix A in Eq. (3.21), (3.22), and (3.23), respectively,
gives Eq. (3.24) as shown below:
ε = (ε1 ε 2 L ε n )
Eq. (3.21)
e = (e1 e2 L en )
Eq. (3.22)
⎛ a11
⎜ a21
⎝ n ,1
Eq. (3.23)
an , 2
L a1,n ⎞
L a2,n ⎟
O M ⎟
L an ,n ⎟⎠
ε∗ A + e =ε
Eq. (3.24)
Solving this equation for ε provides Eq. (3.25), which gives embodied intensity.
ε = e ∗ (I − A)−1
Eq. (3.25)
This method is now widely used, since it is difficult to make accurate estimates of energy
uses for imported products. As described above, ( I − A)−1 refers to Leontief’s Inverse Matrix,
where I is the identity matrix. However, the above method involves calculating the energy
use for only domestic production activities and excludes inputs from imported products.
Conversely, the following method of calculating involves imported products.
Import coefficient mi that represents percentages of imported products with respect to
intermediate demand and domestic final demand in sector i is defined as the following
equation (Miller & Blair 1985; Nansai et al. 2002; Kaneko et al. 2003):
mi =
i, j
Eq. (3.26)
X j + Fi
j =1
where M i is the imports in sector i , n
is number of sectors, and Fi is domestic final
Subtracting the energy use for imported products from Eq. (3.19) gives Eq. (3.27):
a1, j ε 1 + a2, j ε 2 + L + ak , j ε k + L an , j ε n − (a1, j m1ε 1 + a2, j m2ε 2 + L ak , j mk ε k +
L + a n , j mn ε n ) + e j = ε j
Then, expressing import coefficient mi using diagonal matrix M
A ∗ε − M ∗ A ∗ε + e = ε
Eq. (3.27)
gives the following
Eq. (3.28)
Solving Eq. (3.28) for ε gives the embodied intensity for domestic producer goods equation
(Eq. 3.29) as shown below, which gives an accurate value for the actual burdens
domestically generated.
ε = e{I − (I − M ) A}
Eq. (3.29)
The data from IO-tables comprises a useful data source for calculating embodied energy,
which is the energy required to provide a product both directly and indirectly through all the
processes upstream i.e., traceable backwards from the finished product to the consideration
of raw materials. In particular, these data are very significant since all products and services
are covered. However, it should be recognized that there are limitations to the application of
data from IO-tables. For example, in the Thai IO-tables all commodities and services are
classified into a very limited 180 categories, as compared to Japanese IO-tables, which are
classified into 400 categories (Nansai et al. 2002; NESDB 2003). Therefore, one category
contains many different products. Individual sectors are provided for typical materials,
including steel, glass, resin and paper; but a large number of different manufacturers
produce these materials and they consist of an immense number of different types and
qualities. Therefore, IO-analysis provides only an average value. When it comes to more
highly processed products, such as machinery, a large number of different types of products
are lumped together in one sector. Typically, the electric household machines and equipment
sector include a wide range of products such as microwave ovens, air-conditioners, washing
machines, refrigerators and vacuum cleaners. In this case, it is obvious that multiplying the
price of each product by the intensity per price in this sector will provide a rough estimation.
One increasingly used approach to overcome these problems with IO-analysis is to use a
method that combines the process analysis method and IO-analysis called the hybrid
method. According to Bullard et al. (1978) and Lave et al. (1995), they have classified
methods for embodied analysis into three groups: (a) IO-analysis, (b) process analysis, and
(c) hybrid analysis. Process analysis methods are generally the most accurate, but they are
often incomplete due to the amount of detail required for the evaluation of the main
production process and of the complexity of associated upstream processes. IO-analysis,
while comprehensive, is subject to various errors. For example, it does not differentiate
between different production processes. Hybrid analysis combines both of these methods
with the aim of reducing errors that are typically found among both. However, previous hybrid
analysis methods have tended to be somewhat limited by either of the methods on which
they are based. Details of hybrid analysis can be found in Treloar 1997, Treloar et al. 2000;
3.3.3. Comparative study on indirect energy use of selected Asian mega-cities
There are many studies of energy use and related emissions on a national level that have
been published; conversely, at the city level there are very few such studies. City-scale
studies that cover urban sectors comprehensively are still at the methodological stage of
development. Such analyses typically focus on the urban energy of specific cities, most of
which are outside of Asia. Furthermore, most of the interest was confined to the improvement
of energy efficiency or the reduction in the use of fossil fuels. Obviously, studies on indirect
energy consumption in cities are rare in Asia, as well as the rest of the world (ICLEI 1997;
McEvoy et al. 1997; Kates et al. 1998; Baldasano et al. 1999; Newman 1999; Bennett &
Newborough 2001; IGES 2004; Kennedy et al. 2005). It should be noted that when
considering energy consumption as well as environmental protection measures at the city
level the production, consumption and other activities of cities are closely linked with those of
other cities and countries. City-level measures should always be considered from the
viewpoint of interdependence between the factors within the city and those outside of it. The
main problem area to be considered is the indirect energy consumption, or the embodied
energy in goods and services that are imported from outside of the city. For instance, if a city
does not produce goods but consumes goods from the outside, the direct energy
consumption in the city is small. Even when the apparent environmental load in the city is
small, the actual environmental load outside of the city may be large. Therefore, it cannot be
concluded that the policies of a city with low direct energy consumption are favorable to the
global environment.
As regards energy consumption statistics at the city level, some of the city's administration
has provided this kind of data; however, it is difficult to interpret variations in the indirect
energy consumption from the statistics released to the general public. This section presents
a comparative study upon the indirect energy consumption of selected Asian mega-cities,
including Bangkok, Beijing, Shanghai and Tokyo. The indirect energy use in those targeted
cities has been estimated using the same methodology (IO-table and embodied energy
analysis) as described in the earlier section. The studies on Beijing, Shanghai and Tokyo
were extracted from Urban Energy Use and Greenhouse Gas Emissions in Asian MegaCities: Policies for A Sustainable Future (IGES 2004), while the Bangkok case is compiled
from one of the author’s papers (see Phdungsilp 2005). A detailed estimation of indirect
energy consumption in Bangkok will be described in chapter 5. Unfortunately, all of these
studies were conducted for different study years, and in many cases the results are not
directly comparable, given data gaps and assumptions made by the analysts. However,
some interesting comparisons with caveats can be made.
The consumption of large amounts of goods and services by cities has an indirect effect
upon places outside of the cities where the manufacturing and resource extraction take
place. On a global scale, it does not matter where environmental loading originates: cities
should be judged by their contributions to the total environmental load. A city does not only
consume goods and services, it also supplies them in the form of exports. The relationship
between the direct and indirect energy consumption for which a city is responsible differs
from city to city depending upon its scale of industrialization and type of industries.
Therefore, the emissions for which a city is responsible are those which are emitted as a
result of direct emissions plus the net value of emissions embodied in material goods that are
consumed in a city after subtracting exported material goods. Figure 3.9 shows the
proportion of the direct and indirect energy consumption in selected Asian mega-cities.
Indirect (ktoe)
Direct (ktoe)
Bangkok (1998)
Tokyo (1995)
Beijing (1997)
Shanghai (1997)
Figure 3.9. Direct and indirect energy consumption of selected Asian mega-cities
(compiled from IGES 2004; Phdungsilp 2005).
According to the Figure 3.9, indirect energy consumption is more significant than direct
energy demand in Bangkok, Shanghai and Tokyo. During 1990-1995, the total amount of
energy demand within Tokyo decreased by 6.34%, while indirect energy demand decreased
by 10.74% and direct energy demand increased by 9.36%. During 1992-1997 the total
amount of energy demand of Shanghai increased by about 1.66%, among which the direct
energy demand increased by 25.37% and the indirect energy demand decreased by 11.94%.
For the Beijing case, the direct energy demand is greater than the indirect energy demand.
During the period 1992-1995, the total amount of energy demand increased by 25.77%,
among which the direct energy demand increased by 33.57% and simultaneously the indirect
energy demand increased by 14.89% (IGES 2004). These differences can be explained by
the fact that Bangkok, Shanghai, and Tokyo have a great reliance upon the outside in terms
of energy demand, while Beijing is more self-sufficient.
Indirect energy uses can identify the degree to which the economic dependency of the city is
upon the outer regions and especially it can highlight the environmental load displacement of
the highly developed mega-cities to the outer regions. Additionally, the ratio of direct energy
to indirect energy demand can also provide information about economic structural changes
within the city.
Apart from indirect emissions responsibility, the indirect energy consumption can be
addressed from other viewpoints, including that of waste management and the creation of a
material-cycle society, which is found in an advanced form in Europe. These can reduce
emissions as well as the consumption of natural resources.
The lack of quantitative and qualitative information at the city level poses a problem in
regards to the Asian context. The obstacles that stand in the way of a comparative analysis
include the incomparability of information, problems in defining data and differences in the
methods used to organize information and data. It can be seen that scientific as well as
energy analysts need to create consistent information.
3.4. References
Alfstad T. (2005), Development of a Least Cost Energy Supply Model for the SADC
Region. Master of Science in Engineering Thesis, Energy Research Centre, Faculty of
Engineering and the Built Environment, University of Cape Town, South Africa.
Ayres R.U., Simonis U.E. (1994), Industrial Metabolism. Untied Nations University Press,
Bakshi B.R. (2000), A Thermodynamic Framework for Ecologically Conscious Process
System Engineering. Computers and Chemical Engineering 24, pp.1767-1773.
Baldasano J.M., Soriano C., Boada L. (1999), Emission Inventory for GHG in the City of
Barcelona, 1987-1996. Atmospheric Environment 33, pp.3765-3775.
Bennett M., Newborough M. (2001), Auditing Energy Use in Cities. Energy Policy 29,
Bullard C.W., Penner P.S., Pilati D.A. (1978), Net Energy Analysis: A Handbook for
Combining Process and Input-Output Analysis. Resources and Energy 1, pp.267-313.
Coley D.A., Goodliffe E. (1997), The Embodied Energy of Food: the Role of Diet. Energy
Policy 26, pp.455-459.
Dixon R. (1996), Inter-industry Transactions and Input-Output Analysis. Australian
Economic Review 3 (115), pp.327-336.
Duchin F. (1992), Industrial Input-Output Analysis: Implications for Industrial Ecology. In
Proceedings of the National Academy of Science of the USA 89, pp.851-855.
Finon D., and Lapillone B. (1983), Long-term Forecasting of Energy Demand in the
Developing Countries. European Journal of Operational Research 13 (12).
Fischer K.M., Huttler W. (1998), Society's Metabolism, the Intellectual History of Material
Flow Analysis, Part II, 1970-1988. Journal of Industrial Ecology 2, pp.107-137.
Harnisch J., Koch M., Höhne N., Blok K. (2002), Prospects for the Application of Energy
Models in the Design of Climate Policies. The 6th Greenhouse Gas Control Technologies
Conference, International Energy Agency, Japan.
Heaps C. (2002), Integrated Energy-Environment Modeling and LEAP. Stockholm
Environment Institute - Boston and Tellus Institute.
Heaps C. (2004), A Tool for Sustainable Energy Analysis. reCOMMEND (1), pp.12-15.
Howells M.I., Alfstad T., Cross N., Jeftha L.C. (2002), Rural Energy Modelling. Energy
Research Institute, Department of Mechanical Engineering, University of Cape Town,
South Africa.
Howells M.I., Kenny A.R., Solomon M. (2002), Energy Outlook 2002: Modelling Energy in
South Africa. Energy Research Institute, Department of Mechanical Engineering,
University of Cape Town.
ICLEI (1997), Local Government Implementation of Climate Protection: Case Studies.
International Council for Local Environmental Initiatives. Available at, as
accessed 1.8.2002.
IEA (2000), Advanced Local Energy Planning (ALEP) - A Guidebook. International Energy
Kaneko S., Nakayama H., Wu L. (2003), Comparative Study on Indirect Energy Demand,
Supply and Corresponding CO2 Emissions of Asian Mega-Cities. In Proceedings of the
International Workshop on Policy Integration Towards Sustainable Urban Energy Use for
Cities in Asia, East West Center, Honolulu, Hawaii.
Kanpal T.C. (1998), Lecture Note on Energy System Analysis. Renewable Energy System
Programme, United Nations University (UNU), Japan at Centre for Energy Studies, Indian
Institute of Technology, Delhi, India.
Karkacier O., Goktolga Z.G. (2005), Input-Output Analysis of Energy Use in Agriculture.
Energy Conversion and Management 46, pp.1513-1521.
Kate R.W., Mayfield M.W., Torrie R.D., Witcher B. (1998), Methods for Estimating GHG
from Local Places. Local Environment 3 (3), pp.278-298.
Kennedy C.A., Cuddihy J., Yan J. (2005), The Changing Metabolism of Cities. Department
of Civil Engineering, University of Toronto, Canada.
Lave L.B., Cobas-Flores E., Hendrickson C.T., McMichael F. (1995), Life Cycle
Assessment: Using Input-Output Analysis to Estimate Economy-wide Discharges.
Environmental Science and Technology 29, pp.420-426.
Leontief W. (1953), Introduction. In Leontief W., Chenery H.B., Clark P.G., Duesenberry
J.S., Ferguson A.R., Grosse A.P., Grosse R.N., Holzman M., Isard W., Kistin H. (eds.),
Studies in the Structure of the American Economy, Oxford University Press, New York,
NY, USA, pp.3-16.
McEvoy D.W., Longhurst K.W.S., Gibbs D.C. (1997), A Framework for the Construction of
Local CO2 Inventories. In Power H., Tiranasai T., Brebbia C.A. (eds.), Air Pollution Vmodelling, Monitoring and Management, Computational Mechanics Publications,
Southampton, pp.3-13.
Miller R.E., Blair P.D. (1985), Input-Output Analysis: Foundations and Extensions.
Prentice-Hall, Englewood Cliffs, New Jersey.
Nansai K., Moriguchi Y., Tohno S. (2002), Embodied Energy and Emission Intensity Data
for Japan Using Input-Output Tables (3EID). National Institute for Environmental Studies,
Newman P.W.G. (1999), Sustainability and Cities: Extending the Metabolism Model.
Landscape and Urban Planning 44, pp.219-226.
NESDB (2003), Thailand Input-Output Table 1975 - 2000. National Economic and Social
Development Board, Bangkok, Thailand.
NTUA (1996), Implementing Large Scale Integration of Renewables (REPLAN). National
Technical University of Athens, CEC, DGXII, APAS Programme, CT940005, Athens.
Phdungsilp A. (2005), Towards Sustainable Urban Energy Use in Cities: A Metabolism
Approach. In Proceedings of the 2005 World Sustainable Building Conference, September
27-29, 2005, Tokyo, Japan.
Sarafidis Y., Diakoulaki D., Papayannakis L., Zervos A. (1999), A Regional Planning
Approach for the Promotion of Renewable Energies. Renewable Energy 18, pp.317-330.
SEI-Boston (1995), LEAP User Guide, Version 95.0. Stockholm Environment Institute,
SEI-Boston (2005), User Guide for LEAP 2005. Stockholm Environment Institute, Boston.
Sterman J.D. (1991), A Sceptic's Guide to Commuter Models. Massachusetts Institute of
Technology, USA.
Stoll H.G. (1989), Least-cost Electric Utility Planning. John Wiley & Sons, USA.
Subak S. (1995), Methane Embodied in the International Trade of Commodities:
Implications for Global Emissions. Global Environmental Change 5, pp.433-446.
Treloar G.J. (1997), Extracting Embodied Energy Parts From Input-Output Tables:
Towards an Input-Output-Based Hybrid Energy Analysis Method. Economic Systems
Research 9 (4), pp.375-391.
Treloar G.J., Love P.E.D., Faniran O.O., Iyer-Raniga U. (2000), A Hybrid Life Cycle
Assessment Method for Construction. Construction Management and Economics 18,
Treloar G.J., Love P.E.D., Holt G.D. (2001), Using National Input-Output Data for
Embodied Energy Analysis of Individual Residential Buildings. Construction Management
and Economic 19, pp.49-61.
van Engelenburg B.C.W., van Rossum T.F.M., Blok K., Vringer K. (1994), Calculating the
Energy Requirements of Household Purchase. Energy Policy 22, pp.648-656.
Tradeoffs among conflicting objectives lie at the heart of much energy planning. Engineers,
economists, policy-makers and government agencies have many objectives, goals, criteria,
attribute or performance indices they use to judge possible courses of action. The problem is
that it is frustratingly rare that all of these desiderata can be met or maximized by a single
alternative. Instead, some options will be good according to some criteria whereas other
alternatives will do better against differing criteria. Choosing one of the alternatives over the
others means that the priorities must have been set in such a way that accomplishing some
goals would sacrifice others. Setting priorities is hard for an individual to do, and even more
difficult when there are many people involved in a decision-making process, each with their
own values and perspectives. Confronting a decision consists of understanding how well
different options might perform based on the given objectives (Hobbs & Meier 2000). In the
energy sector, tradeoffs are particularly evident because of the large environmental, social
and economic costs frequently involved in providing reliable and convenient energy supplies.
The main purpose of this chapter is to introduce a systematic approach to examining
tradeoffs and expressing value judgments that can yield critical insights, facilitate
constructive discussion among conflicting interests, and effectively build a consensus. MultiCriteria Decision-Making (MCDM) is presented through actual application, illustrating how it
can help bring insight into problems and facilitate agreement among a range of stakeholders.
This thesis emphasizes the use of this technique in dealing with sustainable energy systems
(i.e. the evaluation of scenarios). The reviews of the literature show that MCDM techniques
are gaining popularity within sustainable energy management. These techniques provide
solutions to problems that involve conflicting and multiple objectives.
4.1. Overview of multi-criteria decision-making
It is well known from psychological research that the human brain can at any one time
consider only a limited amount of information. The very nature of multiple criteria problems is
that there is much information of a complex and conflicting nature, often reflecting differing
viewpoints and often changing with time. One of the principal aims of MCDM methods is to
help decision-makers organize and synthesize such information in a way that leads them to
feel comfortable about making a decision. Belton (1990) stated that the context in which
MCDM is useful does not fit into the traditional optimization paradigm of operational
research. The concept of an optimum does not exist; there is no such thing as a “right
answer”. The aim of MCDM methods is to help decision-makers learn about the problems
they face, to learn about their own and other parties' value systems, to learn about
organizational values and objectives by exploring them within the context of the problem,
which can guide them through a process of learning, understanding, information processing,
assessing, and defining of the problem and its circumstances.
MCDM method can provide solutions to increasingly complex energy problems. Traditional
single criteria decision-making is normally aimed at the maximization of benefits with a
minimization of costs. These methods provide a better understanding of the inherent features
within decision-making problems, promote the role of the participants in the decision-making
processes, facilitate compromise and collective decisions, and provide a good platform in
which understanding the perception of the model and analysis in a realistic scenario can take
place. These methods help to improve the quality of decisions by making them more explicit,
rational and efficient. They deal with the process of making decisions in the presence of
multiple objectives. A decision-maker is required to choose among quantifiable or nonquantifiable and multiple criteria. The objectives are usually conflicting, and therefore, the
solution is highly dependent upon the preferences of the decision-maker and therefore must
be a compromise. In most cases, different groups of decision-makers are involved within the
process. Each group brings different criteria and points of view, which must be resolved
within a framework of understanding and mutual compromise. Many applications of MCDM to
energy engineering can be found within the literature (e.g., Afgan et al. 1998, Afgan &
Carvalho 2000, Afgan et al. 2000a, Hobbs & Meier 2000, Afgan & Carvalho 2002,
Phdungsilp & Martinac 2004). In addition, the applications include areas such as integrated
manufacturing systems, evaluation of technology investment, water, and agriculture
management (e.g., Putrus 1990, Boucher & McStravic 1991, Ozelkan & Duckstein 1996,
Raju & Pillai 1999). Based on the intention of MCDM methods, Hobbs & Meier (2000)
concluded that these methods have six basic functions that support the overall goal:
1. To structure the decision process. The MCDM approach helps analysts and decisionmakers think systematically about a problem by providing a logical framework for
defining the alternatives, comparing their performance on important objectives, and to
consider different viewpoints.
2. To display tradeoffs among criteria. An example of a tradeoff is “a decrease of 10,000
tons in SO2 emissions would require a cost increase of US$ 2,000,000.” Such
displays aim to help planners, regulators, and the public understand the relative
advantages and disadvantages of alternatives. Two common approaches include X-Y
plots, where individual points represent the performance of alternatives upon two
objectives, and value paths in which a series of vertical bars or scales represent the
degree of achievement for each criterion. A single alternative is represented as a path
that connects points upon those scales.
3. To help people reflect upon, articulate, and apply value judgments concerning
acceptable tradeoffs, resulting in recommendations concerning alternatives.
Recommendations can be of several types: choosing the best single option,
constructing a portfolio consisting of several options, defining two or more alternative
strategies representing different viewpoints, and screening out undesirable options.
The objective is to help people understand the implications of their value judgments
and to inspire confidence in the soundness of the decision without being
unnecessarily difficult.
4. To help people make more consistent and rational evaluations concerning risk and
uncertainty. Psychological studies show that people have difficulty being consistent in
decisions concerning which risks are acceptable. Some MCDM methods attempt to
measure decision-makers' attitudes towards risk using utility functions, and then use
those functions to evaluate the alternatives.
5. To facilitate negotiation. MCDM methods accomplish this in two ways: firstly, by
quantifying and communicating the priorities held by different stakeholders, and
secondly, by moving the discussion away from alternatives and towards fundamental
objectives and tradeoffs among those objectives. A focus on values facilitates
negotiation because it encourages people to think about their common interests and
avoids the defensive discussions that often result from each stakeholder anchoring
on a preferred alternative. Understanding whether differences of opinion arise from
disagreements over facts or over values is difficult in emotionally charged debates
over issues such as nuclear power, but is critical to making progress in negotiations.
A discussion of values can also help define new alternatives that better satisfy group
6. To document how decisions are made. By detailing how each of the steps of MCDM
analysis has been applied, an agency or firm can communicate the basis of its
decision to stakeholders. Indeed, some of the first applications of formal MCDM
methods by the US Government required agencies to document the rationale for the
decisions within environmental impact statements.
MCDM is a branch of a general class of operations research models that deal with the
problems of decision-making under the presence of a number of decision criteria. According
to Climaco (1997), this class can be further divided into Multi-Objective Decision-Making
(MODM) and Multi-Attribute Decision-Making (MADM). There are several methods in each of
the above categories. Priority based, outranking, distance based and mixed methods are
also applied to various problems. Each method has its own characteristics and the methods
can also be classified as deterministic, stochastic and fuzzy. There may be combinations of
the above methods, depending upon the number of decision-makers the methods can be
classified as single or group decision-making methods. Decision-making under uncertainty
and decision support systems are also prominent decision-making techniques (Gal & Hanne
These methodologies share common characteristics of conflict among criteria, incomparable
units, and difficulties in the selection of alternatives. The best alternative is usually selected
by making comparisons between the different alternatives in respect to each attribute. The
overview of the multi-criteria decision process is shown in Figure. 4.1. The different weighting
methods are described in the following sub-sections.
Formulation of
Selection of
Selection of Decision Process
Performance Evaluation
Decide Decision Parameters
Application of the Method
Evaluation of Result
Figure 4.1. Multi-criteria decision process
(adapted from Pohekar & Ramachandran 2004).
4.1.1. Weighted sum method (WSM)
The weighted sum method (WSM) is the most commonly used approach, especially in single
dimensional problems. If there are M alternatives and N criteria then the best alternative is
the one that satisfies the following expression (Solnes 2003):
= Max∑ aij w j for i = 1,2,3...M
Eq. (4.1)
is the WSM score of the best alternative, N is the number of decision criteria,
where AWSM
aij is the actual value of the ith alternative in terms of the jth criterion, and w j is the weight of
importance of the jth criterion.
The total value of each alternative is equal to the sum of the products. Difficulty with this
method emerges when it is applied to mutli-dimensional decision-making problems. In
combining different dimensions, and consequently different units, the additive utility
assumption is violated.
4.1.2. Weighted product method (WPM)
The weight product method (WPM) is very similar to WSM. The main difference is that
instead of addition in the model there is a multiplication. Each alternative is compared with
the others by multiplying a number of ratios, one for each criterion. Each ratio is raised to the
power equivalent to the relative weight of the corresponding criterion. In general, in order to
compare the alternatives AK and AL the following product is obtained (Chang & Yeh 2001):
R ( AK / AL ) = ∑ (a kj / a Lj )
Eq. (4.2)
j =1
where N is the number of criteria, akj is the actual value of the ith alternative in terms of jth
criterion. If R( AK / AL ) is greater than one, then alternative AK is more desirable than
alternative AL (in the maximization case). The best alternative is the one that is better than or
at least equal to all the other alternatives.
4.1.3. Analytical hierarchy process (AHP)
Analytical Hierarchy Process (AHP) has been developed by Saaty (1980; 1992). The
essence of the process is the decomposition of a complex problem into a hierarchy with a
goal or objective at the top of the hierarchy, criteria and sub-criteria at levels and sub-levels
of the hierarchy, and decision alternatives at the bottom of the hierarchy. Elements at a given
hierarchical level are compared in pairs to assess their relative preference with respect to
each of the elements at the next higher level. The verbal terms of Saaty’s fundamental scale
of 1-9 are used to assess the intensity of preference between the two elements. The value of
1 indicates equal importance, 3 moderately more, 5 strongly more, 7 very strongly, and 9
extremely more important. The values of 2, 4, 6, and 8 are allotted to indicate compromise
values of importance. The ratio scale and the use of verbal comparisons are used for
weighting of quantifiable and non-quantifiable elements. The method computes and
aggregates their eigenvectors until the composite final vector of weight coefficients for
alternatives is obtained. The entries of a final weight coefficients vector reflect the relative
importance (value) of each alternative with respect to the goal stated at the top of hierarchy.
A decision-maker may use this vector in accordance with his particular needs and interests.
To elicit pair-wise comparisons performed at a given level, a matrix A is created in turn by
putting the result of pair-wise comparison of element i with element j into the position aij as
⎡ a11
A = ⎢⎢a 21
⎢⎣ a n1
. a1n ⎤
. a 2 n ⎥⎥
. a nn ⎥⎦
a 22
an 2
Eq. (4.3)
After obtaining the weight vector, it is then multiplied with the weight coefficient of the
element at a higher level (that was used as criterion for pair-wise comparisons). The
procedure is repeated upwards for each level, until the top of the hierarchy is reached. The
overall weight coefficient, with respect to the goal for each decision alternative is then
obtained. The alternative with the highest weight coefficient value should be taken as the
best alternative. One of the major advantages of AHP is that it calculates the inconsistency
index as a ratio of the decision-maker's inconsistency and randomly generated index. This
index is important for the decision-maker to assure that his/her judgments were consistent
and that the final decision is made well. The inconsistency index should be lower than 0.10.
Although a higher value of inconsistency index requires re-evaluation of pair-wise
comparisons, decisions obtained in such cases could also be taken to be the best alternative
(Pohekar & Ramachandran 2004).
4.1.4. Preference
This method uses the outranking principle to rank the alternatives, combined with the ease of
use and decreased complexity. It performs a pair-wise comparison of alternatives in order to
rank them with respect to a number of criteria. Brans et al. (1986) have offered six
generalized criteria functions for reference, namely usual criterion, quasi criterion, criterion
with linear preference, level criterion, criterion with linear preference and indifference area,
and Gaussian criterion. The method uses preference function Pj (a, b) , which is a function of
the difference d j between two alternatives for any criterion j; i.e., d j = f ( a, j ) − f (b, j ) ,
where f (a, j) and f (b, j) are values of two alternatives a and b for criterion j . The
indifference and preference thresholds q’ and p’ are also defined depending upon the type of
criterion function. Two alternatives are indifferent for criterion j as long as d j does not
exceed the indifference threshold q’. If d j becomes greater than p’, there is a strict
preference. Multi-criteria preference index π (a,b) , a weighted average of the preference
functions Pj (a, b) for all the criteria, is defined as follows:
π ( a, b) =
j =1
w j Pj (a, b)
Eq. (4.4)
j =1
Φ+ ( a) = ∑ π ( a, b)
Eq. (4.5)
Φ− ( a) = ∑ π (b, a)
Eq. (4.6)
Φ (a) = Φ + (a) − Φ − (a)
Eq. (4.7)
where w j is the weight assigned to the criterion j , Φ+ ( a) is the outranking index of a in the
alternative set A, Φ− ( a) is the outranked index of a in the alternative set A, and Φ( a) is the
net ranking of a in the alternative set A. The value having the maximum Φ( a) is considered
to be the best.
a outranks b if Φ( a) > Φ(b) , a is different to b if Φ( a) = Φ(b)
Eq. (4.8)
4.1.5. The elimination and choice translating reality (ELECTRE)
This method is capable of handling discrete criteria that is both quantitative and qualitative in
nature and provides a complete ordering of the alternatives. The problem is to be formulated
so that it chooses alternatives that are preferred over most of the criteria but that do not
cause an unacceptable level of discontent for any of the criteria. The concordance,
discordance indices and threshold values are used in this technique. Based on these indices,
graphs for strong and weak relationships are developed. These graphs are used in an
iterative procedure to obtain the ranking of alternative (Pohekar & Ramachandran 2004).
This index is defined in the range (0-1), provides a judgment about the degree of credibility
of each of the outranking relations, and represents a test to verify the performance of each
alternative. The index of global concordance Cik represents the amount of evidence to
support the concordance among all criteria, under the hypothesis that Ai outranks Ak . It is
defined as:
j =1
j =1
Cik = ∑W j c j ( Ai Ak ) / ∑W j
Eq. (4.9)
where W j is the weight associated with jth criteria.
The ELECTRE method yields a whole system of binary outranking relations between the
alternatives. Because the system is not necessarily complete, the ELECTRE method is
sometimes unable to identify the preferred alternative. It only produces a core of leading
alternatives. This method has a clearer view of alternatives by eliminating less favorable
ones, especially convenient while encountering a few criteria with a large number of
alternatives in a decision-making problems (Goicoechea et al. 1982; Pohekar &
Ramachandran 2004).
4.1.6. The technique for order preference by similarity to ideal solutions (TOPSIS)
This method was developed by Huang & Yoon (1981) as an alternative to ELECTRE. The
basic concept of this method is that the selected alternative should have the shortest
distance from the negative ideal solution in a geometrical sense. The method assumes that
each attribute has a monotonically increasing or decreasing utility. This makes it easy to
locate the ideal and negative ideal solutions. Thus, the preference order of alternatives is
yielded through comparing the Euclidean distances. A decision matrix of M alternatives and
N criteria is formulated first. The normalized decision matrix and the construction of the
weighted decision matrix is carried out. This is followed by the ideal and negative-ideal
solutions. For the benefit criteria the decision-makers want to have the maximum value
among the alternatives and for cost criteria they want minimum values amongst alternatives.
This is followed by a separation measure and a calculating of the relative closeness to the
ideal solution. The best alternative is the one that has the shortest distance to the ideal
solution and the longest distance to the negative ideal solution (Pohekar & Ramachandran
4.1.7. Compromise programming (CP)
Compromise programming (CP) defines the best solution as the one within a set of efficient
solutions whose point is the closest to an ideal point. The aim is to obtain a solution that is as
close as possible to the ideal. The distance measure used in CP is the family of Lp-metrics
and is given as (Zeleny 1982):
L p (a) = ∑ w pj f j* − f (a) / M j − m j
Eq. (4.10)
where L p (a ) is the L p metric for alternatives a , f ( a) is the value of criterion j for
alternative a , M j is the maximum (ideal) value of criterion j in set A, m j is the minimum
(anti-ideal) value of criterion j in set A, f j* is the ideal value of criterion j , w j is the weight
of the criterion j , and p is the parameter reflecting the attitude of the decision-maker with
respect to compensation between deviations. For p = 1 , all deviation from f j* are taken into
account in direct proportion to their magnitudes, meaning that there is full (weighted)
compensation between deviations.
4.1.8. Multi-attribute utility theory (MAUT)
Multi-Attribute Utility Theory (MAUT) takes into consideration the decision-maker’s
preferences in the form of the utility function, which is defined over a set of attributes. The
utility value can be determined by the determination of single attribute utility functions
followed by a verification of the preferential and utility independent conditions and the
derivation of multi-attribute utility functions. The utility functions can be either additively
separable or multiplicatively separable with respect to single attribute utility. The
multiplicative form of equation for the utility value is defined as (Keeny & Raiffa 1976):
1 + ku( x1, x 2,...x n ) = Π (1 + kk j u j ( x j ))
Eq. (4.11)
where j is the index of attribute, k is overall scaling constant (greater than or equal to -1),
k j is the scaling constant for attribute j , u(.) is the overall utility function operator, and uj(.)
is the utility function operator for each attribute j .
4.2. Developing of multi-criteria decision-making in energy systems assessment
The MCDM has been applied in several energy-related areas: renewable energy planning,
energy resource allocation, building energy management, transportation energy
management, energy planning, electric utility planning, and others. These areas have
common features of the minimization of cost benefit rations, high degrees of uncertainties in
the formulation of the problems, incommensurable units, and the need to handle socioeconomic aspects in planning.
Renewable energy planning and energy resource allocation refer to the compilation of
feasible energy plans and the dissemination of various renewable energy options. The key
factors applicable are investment planning, energy capacity expansion planning, and the
evaluation of alternative energies. Building energy management refers to design, selection,
installation and building energy management options within a multi-criteria environment. The
application normally deals with quantitative issues. Transportation system applications
include the evaluation of alternative strategies for pollution control, the elimination of old
polluting vehicles, and choosing between private and public transport, etc. The key features
of transportation applications are of a high concern for reasons of a socio-economic nature.
Project planning refers to site selection, technology selection and decision-making support in
renewable energy harnessing projects. Miscellaneous applications include desalination plant
selection, solid waste management, etc.
It can be observed from the literature surveyed and a review of application of MCDM to
sustainable energy planning by Pohekar & Ramachandran (2004) that AHP is the most
popular method for prioritizing the alternatives, followed by PROMETHEE and ELECTRE.
According to the aim of this thesis which is to provide a user-friendly tool for aiding energy
planners as well as policy-makers regarding policy interventions in regards to energy
modeling. The following section describes a technique used for sustainability assessment of
energy scenarios based upon the MCDM approach, which enables individual as well as
group decision-making to select an appropriate mix of energy technologies or intervention
4.2.1. A framework of the multi-criteria decision-making process
The purpose of the MCDM method is to produce the means by which decision teams as well
as energy planners are better able to understand and handle a holistic approach. It is a tool
for organizing information required for decision-making. A framework of the MCDM method,
corresponding to the evaluation of energy scenarios, can be seen as two main phases.
In the first phase, the participants or the decision team decide on the criteria they want to use
and determine their relative importance. Since there are usually quite a lot of criteria to
consider, it is suggested that the criteria is organized into five to eight main criteria, each with
several sub-criteria (Andresen 2000). This is the first step and should be done at the
beginning before there are scenario designs to be considered. In this thesis the main criteria,
as well as sub-criteria, were selected from several studies (see Afgan et al. 2000b, Hobbs &
Meier 2000, Afgan & Carvalho 2002, Phdungsilp & Martinac 2004, Cavallaro & Ciraolo
2005). In this phase, the scales can also be established that will be used later in scoring the
various criteria. Both qualitative and quantitative values can be incorporated within the
In the second phase, the group uses the method to judge the relative merits of the energy
scenarios (alternatives). This is done by determining scores for each scenario for each
criterion, using the measuring scales defined in the first phase. The scores can be
aggregated into several overview presentations. For example, a single score for each
scenario, a star diagram for each scenario that shows its score graphically, and a bar chart
for each scenario which gives more detail about the weighted results.
For practical reasons it is recommended that the team designates one of its members to be
responsible for organizing the information and following the steps. This person becomes the
resident MCDM expert. Other members of the team do not need to become familiar with the
mechanics of aggregating the information and running the MCDM calculation, which can be
done by various MCDM software applications; but they do need to understand the principles
involved so that they develop faith in the method.
In general, the proposed method can be conducted in a user-friendly seven-step procedure.
The first three being carried out in phase one, while the last four are carried out in phase two:
Step 1: selecting the main criteria and sub-criteria.
Step 2: developing measurement scales for the sub-criteria.
Step 3: generating energy scenarios.
Step 4: weighting the main criteria and sub-criteria.
Step 5: predicting the performance of the scenarios.
Step 6: aggregating the scores.
Step 7: analyzing the results and making decisions.
It should be noted that during this process, criteria may be added, removed, or reformulated,
which may require the team to go back and redo a part or all of the procedure several times.
This should be considered a useful outcome, for it indicates that the discussion and analysis
of the problem and the objectives has produced a deeper understanding within the decision
Also, it is important to remember that the primary goal of this method is not to provide
definitive answers, but to enhance the ability of all participants to comprehend the problem at
4.2.2. Selected main criteria and sub-criteria
It is important to specify the objective first and then select the main criteria and sub-criteria,
respectively. In this thesis, the objective is to propose sustainable energy systems; therefore,
the main criteria and sub-criteria have been selected from a set of sustainability indicators for
energy system assessment as presented in various studies.
The criteria for the energy system assessment have to reflect three main aspects, namely
environmental, economic and social. Some criteria or sub-criteria can be quantifiable such as
annual resource use. Others can be qualitative such as integration in the urban context. As
any other complex system, the energy system is defined by the constraints that reflect its
function, technology, geography, property and capacity.
In order to cope with the complexity of sustainability related issues for different systems the
criteria have to reflect the wholeness of the system, as well as the interaction of its
subsystems. Consequently, criteria have to measure the intensity of the interactions among
the different elements in the system and its environment. In this view, there is a need for the
criteria sets to be related to the interaction processes that allow an assessment of the
complex relationship within the system. This implies that the complexity of criteria is defined
reflecting links among the internal parameters and external parameters of the system.
In this study, the main criteria for the sustainability assessment are: resource use,
environmental loading, financial and economic, social, and practical. In this respect, the
design criteria and sub-criteria used are shown in Table 4.1, while Figure 4.2 shows the
overall structure of criteria use.
Table 4.1. The design main criteria and sub-criteria.
Main criteria
Resource use
Environmental loading
Financial and economic
Practical (political)
Resource depletion
Annual electricity
Annual fuels
Annual CO2 emissions
Annual SO2 emissions
Annual NOx emissions
Construction cost
Annual operation cost
Annual maintenance cost
Market maturity
Job creation
Public acceptable
Human health impacts
Integration in urban context
Compatibility with political, legislative and
administrative situation
Data availability adequate
Consistence with local technical and economic
The number of main criteria and sub-criteria must be considered carefully. If there is a large
amount of criteria, the problem may be difficult to handle, and one may lose the overview.
Also, with a large number of criteria, the importance weights end up small, and thus the
meaningfulness of the weights is less acute. Alternatively, if there are only a small number of
criteria, it may lead to an oversimplification of the real world (Andresen 2000).
Main goal
(e.g., sustainable energy system)
Main Criteria
(e.g., resource use)
(e.g., annual fuels)
(e.g., kWh/m2 ann.)
Figure 4.2. The overall structure of criteria and sub-criteria use.
The number of criteria may affect the value balance of the hierarchy. The greater the level of
detail pertaining to an objective, possibly reflected in the number of sub-criteria, the more
likely it is that it will be attributed a high level of importance. Miller’s theory (Miller 1956)
suggested that it may be wise to limit the number of main and sub-criteria to approximately
seven. Miller claimed that seven plus or minus two represents the greatest amount of
information that an observer can give about an object on the basis of an absolute judgment.
This theory is widely accepted within the behavioral sciences.
4.2.3. Generating energy scenarios
The energy scenarios (alternatives) depend upon each individual case since it is basically a
craft; little formal guidance can be given for this task. In this thesis, the energy demand is
organized into four sectors: residential, commercial and building, industry, and transport. In
addition, the supply-side is taken into consideration as well. A detailed analysis will be
discussed in the case study in chapter 5.
One of the ultimate goals of the thesis is to promote the maximum use of renewable energy
within a large urban area. Therefore, individual policies (scenarios) have been created for
various sectors to support this objective. For example, in the residential sector, scenarios
simulate the impacts of solar water heaters, passive design, daylighting applications, and
photovoltaic installations in households. The policy question addresses the implications of,
what if the policies are implemented according to the assumptions stated in this section.
It is important to consider constraints to implementation in the selection of policies, because
the analysis of policies will provide a useful input into the city’s energy modeling. It may thus
be useful to categorize the policies into main groups and sub-groups, according to the
purpose and the individual sectors.
4.2.4. Performance prediction
The performance prediction is done in accordance with the developed scales for the subcriteria. A scale for each sub-criterion is necessary to be able to measure the performance of
individual scenarios. A measurement scale is a way to convert a value into a score. A value
can be a number or a phrase, depending on whether the criterion is quantitative or
Quantitative values are used for criteria that can be measured directly with numbers, such as
annual energy use, carbon emissions, etc. Qualitative values are words or phrases that can
be used to characterize how well a scheme rates against a particular criteria where the rating
is more a matter of judgment, not normally subject to quantification. These are quality issues,
such as public acceptance or integration into an urban context. Some criteria can be
characterized in either a qualitative or quantitative way.
All criteria are ultimately converted to a qualitative scale. For example in the 4-to-10 scale,
the upper and lower ends have particular meanings. The upper end, a score of 10, means
that the scenario rates as “excellent”. To be more exact, the 10 means that the scenario is
the “best reasonable attainable” with regard to the particular criteria. The lower end, a score
of 4, means that it is just marginally possible to construct such an energy system.
The next step is to create a measurement scale for each of the main criteria, indicating the
assessment of the merit of achieving particular scores. In this study, the main criteria are
evaluated by using a questionnaire on a 4-to-10 scale or by simply saying that a bigger
number is better. Table 4.2 illustrates the example of a measurement scale for annual
electricity use.
Table 4.2. Example of measurement scale for annual electricity use.
Good to Excellent
Fair to Good
Acceptable to Fair
Marginally acceptable
Annual electricity use (kWh/m2)
However, the process of creating measurement scales should generate much discussion. It
will often come to light that the same words have different meanings for different individuals.
Scaling the objectives of a problem in this manner not only helps the team arrive at uniformity
of measurement scales, but is also a way to define the general nature and context of the
problem. This process also involves the collective participation of the entire team and allows
each team member to express his or her own values and expertise to the group as a whole.
Regarding the levels of the predicted performance of the proposed scenarios, with respect to
the criteria, are determined. The performance prediction can be done based on computer
simulation (energy modeling), databases, rules of thumb, experience or expert judgment. It is
best if the scoring for the performances is decided upon by the whole team.
4.2.5. Weighting of criteria and aggregating scores
The main criteria weights reflect the central priorities of the project. The weights chosen will
be critical in comparing alternative schemes. There are different ways of eliciting weights.
The grading method works with the weights directly. The criteria weights are determined on a
10-point scale similar to the one used for scoring the performances. The decision team
expresses the importance of criteria in grades on the scale 10, 9, 8, and so on down to 4.
The most important criterion receives a grade of 10. All other criteria are compared to this
e.g., if a criterion is felt to be somewhat less important than the most important one, it
receives a grade of 8.
Another method is a complex mathematical technique, The so-called AHP as mentioned in
section 4.1.3. Furthermore, there is a range of software available that can implement the
MCDM model. In this thesis, the computer program Web-HIPRE (HIerarchical PREference
analysis on the World Wide Web) has been used in this process. The Web-HIPRE has been
developed by the Systems Analysis Laboratory at the Helsinki University of Technology,
Finland. Since it is a web-based software, it can be accessed from anywhere in the world
and provides a common platform for individual and group decision support (Hämäläinen
2000). Detailed information will be given in the next section.
Andresen (2000) suggested that in order to present the overall performance of an energy
scenario a so-called “star diagram” is a useful tool to graph the weights. In this diagram it is
possible to show multiple dimensions; thus, all individual performance measures can be
gathered into one picture. Each finger (see Figure 4.3) represents the scale for one criterion.
The performance on each criterion is plotted on each finger. The center of the star usually
designates the lowest/worst performance, or the minimum score of 4, for each criterion. The
outer unit polygon represents the maximum score of 10 for each criterion.
Figure 4.3. An example of a star diagram with indicated individual criteria scores and
reference scores (adapted from Andresen 2000).
However, there is a limit to the number of dimensions that can be presented without making
the graphics look too cluttered. According to Miller’s theory (Miller 1956), a maximum of nine
fingers is probably the limit. Therefore, only the main criteria should be plotted in the star
diagram. The sub-criteria may be indicated below the main criterion.
4.3. The Web-HIPRE decision-support software
Web-HIPRE is Internet software for multi-criteria decision analysis based on the well-known
decision support software HIPRE 3+ (Hämäläinen & Lauri 1995). Web-HIPRE provides
implementations of AHP and multi-attribute value theory (MAVT) to support the different
phases of decision analysis, prioritization and analysis of the results. Individual models can
be integrated into a group model via the Internet. In general, Web-HIPRE includes decisionmaking, group collaboration and computer support.
Web-HIPRE is a Java-applet and can be accessed from any location with a Java-enabled
web-browser (e.g., Internet Explorer and Netscape) connected to the Internet. When using
Web-HIPRE, the browser loads the applet in its local memory, from where it is operated.
When the browser is closed, no files remain on the user's computer. An essential feature is
the possibility of defining links to other Internet addresses. These links can refer to graphical
or any other forms such as sound or video to describe the criteria or alternatives, which can
improve the quality of decision support dramatically.
The methods used in Web-HIPRE are based on the AHP and on MAVT. In the area of
MAVT, it supports direct weighting, Simple Multi-Attribute Rating Technique, SMART
(Edwards 1977; von Winterfeldt & Edwards 1986; Edwards & Barron 1994), SWING (von
Winterfeldt & Edwards 1986) and the rank based SMARTER technique (Edwards & Barron
1994). Figure 4.4 presents the hierarchy of the residential case for evaluating the energy
scenarios. In the hierarchy window, different weighting or rating methods can be used in the
local subtrees, and the selected methods are shown by abbreviations including: DR = Direct,
SM = SMART, SW = SWING, SR = SMARTER, PW = AHP (Pair-wise Comparisons), and VF
= Value Functions.
Basically, the method of SWING weighting is to elicit weights for the criteria. This is based on
comparisons of differences: how does the swing from 0 to 100 on one preference scale
compare to the 0 to 100 swing on another scale? Regarding SMART, it has undergone one
important change and one potentially useful extension. The change was the introduction of
the swing-weights procedure for weight assessment to replace the one originally
recommended. Thus, SMART became SMARTS (SMART with Swings). Edwards & Barron
(1994) have also developed a way of implementing SMARTS that is less demanding in
regards to the information input requirements from the decision-maker. This procedure is
named SMARTER (SMART Exploiting Ranks).
Figure 4.4. An example of value tree window of the residential case
(extracted from Web-HIPRE,
4.3.1. The analytic hierarchy process method
Besides the original form of AHP, Web-HIPRE also supports a continuous preference scale.
Preferences can be given, for example graphically with the slider or numerically by typing a
value (Figure 4.5). When all the pair-wise comparisons have been given, elicited weights for
each attribute are shown by numbers and bar-graphs.
Verbal statements can also be used to define preferences by choosing an appropriate
expression from the list. Web-HIPRE uses the original mapping from words to ratios, which
is: equal preference = 1, moderate preference = 3, strong preference = 5, very strong
preference = 7, and extreme preference = 9 (Saaty 1980). However, it has been shown that
the use of the 9/9 – 9/1 scale (Ma & Zheng 1991) or the balanced scale (Salo & Hämäläinen
1997) gives a better equivalency to verbal expressions (Pöyhonen et al. 1997). These variant
scales can be used by selecting the corresponding rations from the continuous scale.
However, the original consistency ratio (CR) of AHP (Saaty 1980) is not applicable with the
general scales. When using different discretions of a scale or a continuous scale, a scaleinvariant consistency measure should be used. In Web-HIPRE the consistency measure
(CM) is used (Salo & Hämäläinen 1997). It is derived by transforming the inconsistent replies
into an extended set of feasible preference statements, and using the properties of this set to
measure the inconsistency of the pair-wise comparison matrix.
Figure 4.5. An example of the AHP weighting window
(extracted from Web-HIPRE,
In CM, inconsistent replies are transformed into a non-empty set of feasible preferences and
the properties of this set can be used to measure the inconsistency of the original matrix.
More precisely, the consistency measure is defined as (Salo 1993):
CM =
r (i, j ) − r(i, j )
n ( n − 1) i〉 j (1 + r(i, j )(1 + r(i, j ))
Eq. (4.12)
where r (i, j) , maxka(i,k)a(k,j) stands for the extended bound of the comparison matrix a(i,j)
or the element in the ith row and the jth column, and r(i, j ) is the inverse of r (i, j ) .
Thus, the consistency measure is an indicator of the size of this extended region formed by
the set of local preferences such that w i ≤ r(i, j )w j for all i, j ∈ {1,..., n}. This measure
increases with the inconsistency of the elements of the pair-wise comparisons matrix.
4.3.2. The multi-attribute value theory method
In Multi-Attribute Value Theory (MAVT) the problem is structured into a hierarchical form
similar to AHP. The objective is to obtain general value scores for each alternative. The value
scores are composed of the ratings or the alternatives with respect to each attribute and of
the weights of the attributes. By using an additive value function, the overall value score of
an alternative x is (Mustajoki & Hämäläinen 1999):
v (x ) = ∑ w iv i (x i )
Eq. (4.13)
where v i (x i ) is the component value of an attribute rating x i and w i is the weight associated
with an attribute i . The component value functions vi (⋅) and the weight w i get values
between 0 and 1, and the weights are normalized to sum up to one. The weight w i indicates
the relative importance of the event of an attribute i changing from its worst level to its best
level compared to the corresponding changes on the other attributes.
In MAVT the weights of the attributes can be given directly, or some sophisticated methods
like SMART, SWING or SMARTER can be used. On the lowest level, value scores can be
given directly or, for example value functions can be used to transform the ratings of the
alternatives into the value scores.
4.3.3. Combined use of the methods
A unique feature of Web-HIPRE is the possibility to use different weighting methods within
the same hierarchy. Thus, under each element of the hierarchy the decision-maker can
select the most suitable method. Users can apply different methods, but only one of them is
active. All of the prioritizations made by the other methods are stored, which makes the
comparison of different weighting methods easy.
AHP and MAVT methods can be combined freely in criteria weighting. In respect to the
alternatives, AHP weights can be converted to the compatible 0-1 value scale by setting the
lowest priority weight to zero, the highest priority weight to one, and scaling the intermediate
weights proportionally to this scale (Dyer 1990). More precisely, the converted weight w ci for
the alternative is (Mustajoki & Hämäläinen 1999):
w ci =
w i − w min
w max − w min
Eq. (4.14)
where w max = max( w1,..., w n ) and w min = min( w1,..., w n ) are the original maximum and
minimum AHP weights. These converted AHP weights can now be treated as value scores.
Alternatively, value scores can be normalized to sum up to one in order to make them
compatible with the AHP weights.
4.3.4. Group decision-making
There are different ways of approaching multi-attribute preference aggregation and
combination in group-decision-making. In methods allowing incomplete information, such as
preference programming (Salo & Hämäläinen 1995), PAIR (Salo & Hämäläinen 1992a), and
PRIME (Salo & Hämäläinen 1992b), it is possible to define intervals for the weight ratios
instead of exact number estimates. The intervals can be interpreted to denote the decisionmaker's preferential uncertainty. Interval models can also be used in group-decision-making
by forming intervals so that they include the opinions of all the members (see e.g.,
Hämäläinen & Pöyhönen 1996). As the local weights are presented as intervals, the overall
priorities are also intervals.
Another way to combine individual models is to use a direct aggregation method. In MAVT,
the weighted sum of individual values can be taken to get the group values, but generally this
requires the explicit comparison of interpersonal preferences. The AHP literature proposes
two aggregation methods: the geometric mean method (Aczel & Saaty 1983) and the
weighted arithmetic mean method (Dyer & Forman 1992). The weighted arithmetic mean
method satisfies the commonly accepted axioms for social choice except for the
independence of irrelevant alternatives, which is also violated by AHP itself (Ramanathan &
Ganesh 1994). However, it again requires the explicit weighting of group members.
In Web-HIPRE, individual weights can be aggregated into a group model within the weighted
arithmetic mean method. This naturally includes the weighted sum of MAVT as a special
case. In the group hierarchy each decision-maker is graphically represented by an element,
which actually presents the whole hierarchy of the individual decision-maker. The composite
group priorities are generated as a weighted sum of individual priorities, which are obtained
from the individual models. This feature is very useful for decision-making in regards to
energy for many stakeholders should be involved in the decision process. Then, the
outcome of these group member elements is the overall score for the alternatives. One
should note that, if necessary, the individual members can work in an asynchronous
distributed mode i.e., in different locations and at different times.
4.3.5. Analyzing the results
In Web-HIPRE, the composite priorities (i.e., the overall scores for the alternatives) are
shown both by numerical values and by bar graphs (see Figure 4.6). Bars can be divided into
segments in different ways indicating the relative importance of the criteria and sub-criteria.
Any criterion can be chosen as the goal. Then, Web-HIPRE calculates the composite
priorities from the sub-hierarchy under a selected goal element.
For group decision-making, the overall weights in the group model are represented similarly
as in an individual model. The bar graphs can be divided according to the relative importance
of the decision-makers.
Figure 4.6. An example of the overall values of the alternatives
(extracted from Web-HIPRE,
The main advantage of MCDM is that it provides a systematic way of assessing the
importance of criteria. In addition, MCDM is an attempt to measure the preferences of a
diverse group of people. The MCDM approach can be considered as an alternative to
considering only a handful of predefined plans. Most applications in energy planning have
emphasized the minimization of cost. However, in the MCDM method the environmental and
other non-cost objectives can be added to such models in order to allow the planner to
generate many alternatives representing a range of tradeoffs.
4.4. References
Aczel J., Saaty T.L. (1983), Procedures for Synthesizing Ratio Judgments. Journal of
Mathematical Psychology 27, pp.93-102.
Afgan N.H., Carvalho M.G. (2000), Sustainable Assessment Method for Energy Systems.
Kluwer Academic Publishers, Boston.
Afgan N.H., Carvalho M.G., Hovanov N.V. (2000a), Sustainability Assessment of
Renewable Energy Systems. Energy Policy 28, pp.603-612.
Afgan N.H., Carvalho M.G. (2002), Multi-Criteria Assessment of New and Renewable
Energy Power Plants. Energy 27, pp.739-755.
Afgan N.H., Carvalho M.G., Hovanov N.V. (2000b), Energy System Assessment with
Sustainability Indicators. Energy Policy 28, pp.603-612.
Afgan N.H., Gobaisi D., Carvalho M.G., Cumo M. (1998), Sustainable Energy
Management. Renewable and Sustainable Energy Review 2, pp.235-286.
Andresen I. (2000), A Multi-Criteria Decision-Making Method for Solar Building Design.
Ph.D. Thesis at Department of Building Technology, Faculty of Architecture, Planning and
Fine Arts, Norwegian University of Science and Technology, Trondheim, Norway.
Belton V. (1990), Multiple Criteria Decision Analysis – Practically the Only Way to Choose.
Operational Research Tutorial Papers, Department of Management Science, University of
Strathclyde, UK.
Boucher T.O., McStravic E.L. (1991), Multi-attribute Evaluation within a Present Value
Framework and its Relation to Analytic Hierarchy process. The Engineering Economist 37,
Brans J.P., Vincke P.H., Mareschal B. (1986), How to select and how to rank projects: The
PROMETHEE method. European Journal of Operations Research 24, pp.228-238.
Cavallaro F., Ciraolo L. (2005), A Multicriteria Approach to Evaluate Wind Energy Plants on
An Italian Island. Energy Policy 33, pp.235-244.
Chang Y.H., Yeh C.H. (2001), Evaluating airline competitiveness using multi-attribute
decision making. Omega 29(5), pp.405-415.
Climaco J. (1997), Multicriteria analysis. Springer-Verlag, New York.
Dodd D.E., Lesser J.A. (1994), Can Utility Commissions Improve on Environmental
Land Economics 70, pp.63-76.
Dyer J.S. (1990), Remarks on the Analytic Hierarchy Process. Management Science 36
(3), pp.249-275.
Dyer R.F., Forman E.H. (1992), Group Decision Support with the Analytic Hierarchy
Process. Decision Support Systems 8, pp.99-124.
Edwards W. (1977), How to Use Multiattribute Utility Measurement for Social Decisionmaking. IEEE Transactions on Systems, Man and Cybernetics, 1977, pp.326-340.
Edwards W., Barron F.H. (1994), SMARTS and SMARTER: Improved Simple Methods for
Multiattribute Utility Measurement. Organizational Behaviour and Human Decision
Processes 60, pp.306-325.
Gal T., Hanne T. (1999), Multicriteria decision making: advances in MCDM models,
algorithms, theory, and applications. Kluwer Academic Publishing, New York.
Goicoechea A., Hansen D., Duckstein L. (1982), Introduction to multi objective analysis
with engineering and business application. Wiley, New York.
Hobbs B.F., Meier P. (2000), Energy Decisions and the Environment: A Guide to the Use
of Multicriteria Methods. Kluwer Academic Publishers, Boston.
Huang C.L., Yoon K. (1981), Multi attribute decision making: methods and application.
Springer-Verlag, New York.
Hämäläinen R.P. (2000), Decisionarium – Global space for decision support. Systems
Hämäläinen R.P., Helenius J. (1997), Winpre - Workbench for Interactive Preference
Programming. Systems Analysis Laboratory, Helsinki University of Technology.
Hämäläinen R.P., Lauri H. (1995), HIPRE 3+ User's Guide. System Analysis Laboratory,
Helsinki University of Technology. Available at,
as accessed 21.8.2005.
Hämäläinen R.P., Pöyhönen M. (1996), On-Line Group Decision Support by Preference
Programming in Traffic Planning. Group Decision and Negotiation 5, pp.485-500.
Jones M.R. (1989), The Potential of Decision Analysis as a Means of Integrating the
Environment into Energy Policy Decision-Making. Environment and Planning 21, pp.13151327.
Keeney R.L., Raiffa H. (1976), Decisions with Multiple Objectives: Preferences and Value
Tradeoffs. John Wiley & Sons, Inc.
Ma D., Zheng X. (1991), 9/9-9/1 scale method of AHP. In Proceedings of the 2nd
International Symposium on the Analytic Hierarchy Process (ISAHP'91), Pittsburgh, PA,
Vol. I, pp.197-202.
Miller G.A. (1956), The Magic Number Seven, Plus or Minus Two. Psychological Review
Mirasgedis S., Diakoulaki D. (1997), Multicriteria Analysis vs. Externalities Assessment for
the Comparative Evaluation of Electricity Generation Systems. European Journal of
Operational Research 102 (2), pp.364-379.
Mustajoki J., Hämäläinen R.P. (1999), Web-HIPRE - A Java Applet for AHP and Value
Tree Analysis. 5th International Symposium on the Analytic Hierarchy Process (ISAHP'99),
August 12-14, 1999, Kobe, Japan.
Ozelkan E.C., Duckstein L. (1996), Analyzing Water Resources Alternatives and Handling
Criteria by Multicriterion Decision Techniques. Journal of Environmental Management 48,
Phdungsilp A., Martinac I. (2004), A Multi-Criteria Decision-Making Method for the
Retrofitting of Designated Building in Thailand. In Proceedings of the 21th Conference on
Passive and Low Energy Architecture, Eindhoven, September 19-22, 2004, The
Pohekar S.D., Ramachandran M. (2004), Application of multi-criteria decision making to
sustainable energy planning – A review. Renewable and Sustainable Energy Reviews 8,
Putrus P. (1990), Accounting for Intangibles in Integrated Manufacturing-non-financial
Justification Based on Analytical Hierarchy Process. Information Strategy 6, pp.25-30.
Putta S. (1991), Methods for Valuing and Incorporating Environmental Costs in Electric
Resource Planning and Acquistion. In Hohmeyer O., Ottinger R.L. (eds.), External
Environmental Costs of Electric Power, Springer-Verlag, NY.
Pöyhönen M., Hämäläinen R.P., Salo A.A. (1997), An Experiment on the Numerical
Modelling of Verbal Ratio Statements. Journal of Multi-Criteria Decision Analysis 6, pp.110.
Raju K.S., Pillai C.R.S. (1999), Multicriterion Decision Making in Performance Evaluation of
Irrigation Projects. European Journal of Operational Research 112 (3), pp.479-488.
Ramanathan R., Ganesh L.S. (1994), Group Preference Aggregation Methods Employed in
AHP: An Evaluation and An Intrinsic Process for Deriving Members "Weightages".
European Journal of Operational Research 79, pp.249-265.
Saaty T.L. (1980), The Analytic Hierarchy Process. McGraw-Hill, Inc.
Saaty T.L. (1992), Decision making for leaders. RWS Publications, Pittsburgh.
Saaty T.L. (1994), Fundamentals of Decision Making and Priority Theory with the Analytic
Hierarchy Process. RWS Publications, Pittsburgh, PA.
Salo A.A., Gustafsson J., Gustafsson T. (1999), Prime Decisions. Systems Analysis
Laboratory, Helsinki University of Technology.
Salo A.A. (1993), Inconsistency analysis by approximately specified priorities. Mathematic
Computing Modelling 17, pp.123-133.
Salo A.A., Hämäläinen R.P. (1992a), Preference Assessment by Imprecise Ratio
Statements. Operations Research 40 (6), pp.1053-1061.
Salo A.A., Hämäläinen R.P. (1992b), PRIME - Preference Ratios in Multiattribute
Evaluation. System Analysis Laboratory Research Reports A43, Helsinki University of
Salo A.A., Hämäläinen R.P. (1995), Preference Programming Through Approximate Ratio
Comparisons. European Journal of Operational Research 82, pp.458-475.
Salo A.A., Hämäläinen R.P. (1997), On the Measurement of Preferences in the Analytic
Hierarchy Process. Journal of Multi-Criteria Decision Analysis 6, pp.309-343.
Solnes J. (2003), Environmental quality indexing of large industrial development
alternatives using AHP. Environmental Impact Assessment Review 23(3), pp.283-303.
von Winterfeldt D., Edwards W. (1986), Decision Analysis and Behavioural Research.
Cambridge University Press.
Watson S.R. (1981), Decision Analysis as a Replacement for Cost/Benefit Analysis.
European Journal of Operational Research 7, pp.242-248.
Zeleny M. (1982), Multiple criteria decision making. McGraw-Hill, New York.
This case study is built upon developing methodologies, including indirect energy
consumption based on embodied energy analysis, energy modeling, and multi-criteria
decision-making, as described in the previous chapters.
5.1. Analysis of the energy situation
Over the past two decades, the economic development in Thailand along with relevant
policies have transformed the country from an agricultural-based economy into a semiindustrialized economy. Consequently, energy consumption has rapidly increased. However,
due to the limited data at the city level especially from disaggregated sectors, the energy
analysis is somehow described from both national and city perspectives.
5.1.1. Energy organization in Thailand
The energy sector is undergoing a period of restructuring and privatization. The Thai electric
utilities and petroleum industries, which have historically been state-controlled monopolies,
are currently being restructured. Previously, all energy related organizations operated under
a decentralized structure, each divided into separated ministries and departments. This
decentralization was caused by various things, among which, were differing objectives and
varying circumstances. Certain organizations, such as the Metropolitan Electricity Authority
(MEA) and the Provincial Electricity Authority (PEA), were established as public utility service
providers to local towns. Therefore, they were assigned under the Ministry of Interior. Some
organizations were established at a time when there were no authorities assigned to manage
and control energy production. They were assigned under the Office of the Prime Minister,
such as the Electricity Generating Authority of Thailand (EGAT).
In 1986, the government took into consideration the necessity of aligning policies and
supervising the decentralized organizations in order for them to work towards same goal. As
a result, the Prime Minister's Direction 1992 (B.E. 2535) was issued, establishing the
National Energy Policy Office (NEPO) to assume such responsibility as a department under
the Office of the Prime Minister. NEPO was chaired by the Prime Minister, and consisted of
ministers from related ministries as board members. The Board's Secretary-General acted as
both member and secretary to the Board. The Board had the authority to make decisions,
establish policies for the cabinet, and assign government agencies or state enterprises to
implement plans and projects.
There were energy related agencies allocated in over 20 units across nine ministries. This
hindered the overall operation and caused differing policies and legislation governing each
agency. The government attempted to unify these various independent units in order to help
facilitate energy management. In November 2001, the Bureau of Energy was established.
Subsequently, in January 2002, the Bureau of Energy became the Ministry of Energy. A
comparatively small-sized ministry consisting of six departments: The Office of the Ministry,
The Energy Policy and Planning Office, The Department of Mineral Fuels, The Office of the
Permanent Secretary, The Department of Alternative Energy Development and Efficiency,
and The Department of Energy Business (MOEN 2005).
5.1.2. Energy supply and demand analysis in Thailand
Thailand is endowed with diverse energy resources. Natural gas and low-grade lignite have
the greatest potential, while crude oil, condensate and hydro are available to some extent.
Natural gas is proving to be the country's prime energy resource. From 2000 to 2004, it
contributed about 57% of the domestic production of primary energy and about 31% of the
total primary energy supply. Total natural gas reserves approximated were 33,516 Billion
Cubic Feet (BCF) as of 2004, of which 96% are found offshore and 11,911 BCF are proven
reserves. Lignite is one of the main indigenous resources. In 2004, total lignite reserves were
estimated at about 2,870 million tons and almost 60% of them are found in the northern
region. Crude oil and condensate account for a small amount of the domestic reserves. The
total estimation of oil reserves is 516 million bbl and the condensate is approximately 903
million bbl (DEDE 2004).
Among the domestic production of primary energy, hydro and other renewable energy
resources make a great contribution to the energy system. In 2004, Hydro, geothermal, solar
cell and wind power accounted for 4% of the total production of commercial energy.
Biomass, including fuel wood, paddy husk, bagasse, agricultural waste, garbage and biogas
contributed to 30% of the country’s primary energy production.
However, in terms of total primary energy supply, Thailand relies on imported energy, which
accounts for half of the country's total energy supply. Most of the imported energy is crude
oil, which accounts for about 70% of the total energy imported and is followed by coal,
petroleum products, hydro and charcoal, respectively. Figure 5.1 shows a comparison of total
primary energy supply and the domestic production of primary energy. The comparisons
include commercial energy, new and renewable energy, and other energy such as black
Domestic production of primary energy
Total primary energy supply
Figure 5.1. Comparison of total primary energy supply and domestic production of primary
energy (Data source: DEDE 2004).
Regarding the energy demand, the economic growth rate over the period 1987-1996 was
about 10% annually, and the energy demand has grown by an average of 11% annually
(DEDP 2000). During the period 1997-2000, due to the economic crisis in Southeast Asia,
the average annual GDP growth rate was about 1.2% and the energy demand growth rate
was 0.8% annually. The annual final energy demand growth rate was approximately 6%
during 2000 and 2004, caused by the recovery of the economic situation. These relations
can be explained in a form of energy intensity, which is an index to show the relationship
between energy consumption and a value added. As shown in Figure 5.2 the energy
intensity is expressed in ktoe per thousand Baht at 1988 constant prices. Energy intensity
increased from 15.9 ktoe/thousand Baht in 2000 to 16.4 ktoe/thousand Baht in 2004.
ktoe/1000 Baht
Figure 5.2. Energy intensity in Thai economy during 2000-2004
(Data source: DEDE 2004).
Typically, the final energy demand in Thailand is disaggregated into four sectors: agriculture,
industry, residential and commercial, and transport. In the past ten years, transportation has
been the largest consumer of final energy consumption, followed by industry, residential and
commercial, and agriculture sectors, respectively. Figure 5.3 shows the final energy
consumption in Thailand by sectors during the period 2000-2004. In 2004, the total final
energy consumption of both commercial and renewable energy was 60,269 ktoe; of this, the
commercial energy accounted for 84.2%. Petroleum products consisted of 64% of the total
commercial energy consumption, while biomass was the main source of renewable energy
Figure 5.3. Final energy consumption by sectors in Thailand (Data source: DEDE 2004).
5.1.3. Energy use pattern in Bangkok
As mentioned above, the typical energy analysis is disaggregated into four sectors;
therefore, the energy use pattern in Bangkok is described in the same manner. In Bangkok,
the consumption of energy is dominated by electricity and petroleum products. The
proportion of each energy requirement depends upon the energy services within each sector.
Figure 5.4 shows an estimation of energy consumption by sectors in Bangkok.
Res & Com, 18%
Transport, 58%
Industry, 24%
Figure 5.4. Estimation of energy consumption by sectors in Bangkok
(estimated by LEAP model).
The transport sector accounts for half of the energy consumption in Bangkok. The petroleum
products consumed in this sector are mainly comprised of diesel, gasoline, and liquefied
petroleum gas (LPG). A small fraction of electricity is used by sky trains and sub-ways. The
sky train and sub-way have been in operation since 2000 and 2004, respectively. Road
transport takes the biggest proportion. According to the vehicles registered under the Motor
Vehicle Act, motorcycles account for almost 40%, passenger cars approximately 30%, vans
and pickups share about 15%, and the rest is contributed by microbus, motor-tricycle, urban
taxi, fixed route taxi, fixed route bus, bus for hire, private bus and others.
The industry sector can be broadly divided into two categories: manufacturing and
construction. Manufacturing consists of food and beverages, textiles, wood products, paper,
rubber, chemical and petroleum industries, non-metallic, basic metal, and other
manufacturing. The energy demand basically consumes electricity and non-electricity. The
major purpose of the electricity consumption is for the electric equipment such as motors and
pumps. Also, space conditioning shares a big proportion due to the structure of the
manufacturing industry, which is required to control temperatures in their processes, for
instance car assembly, electronic equipment, etc. Non-electricity utilization is for thermal
purposes in boilers and furnaces.
The residential and commercial sector mainly consumes electricity for the purposes of airconditioning and lighting. The non-electricity demand is primarily for cooking devices such as
LPG cooking stoves. The economic development of the Thai economy brings Bangkok into
being an industrialized and service-oriented city. The Bangkok Metropolitan Area (BMA)
urbanization pattern can be divided into inner BMA and outer BMA. The inner BMA is built up
of temples, government offices, educational institutions, commercial offices, and two to four
story row houses that serve for commercial retail as well as living quarters. The outer BMA
was traditionally cultivated land; however, economic development and immigration have
transformed this area into residential and industrial estates.
A small amount of energy consumed in the agriculture sector accounts for approximately
0.05% of the total energy consumption in Bangkok. The energy consumed is based on nonelectricity, comprising almost entirely of diesel. However, the zones for agricultural purposes
have been transformed and have other purposes such as for households and offices.
Moreover, land prices are increasing and this causes agricultural activities in Bangkok to
become less significant in terms of energy consumption.
5.2. Direct and indirect energy consumption
Theoretical IO-analysis and embodied energy analysis has been described in Chapter 3. As
it is greatly data-intensive, data collection has been the central issue, for a model structure
that can be analyzed is largely dependent upon the data available. The models need detailed
accounts of the flow of energies and materials for each sector. The material flow data is
extracted from the Thai IO-tables. The data available makes it difficult to collect the energy
flows from each sector in Bangkok. To avoid this problem, the LEAP model is used to
calculate the energy demand in Bangkok.
5.2.1. Data specification Input-output tables
The IO-tables of Thailand have been compiled every five years since 1970. An IO-table
records a matrix of goods and transactions of services among the sectors of a certain region
within a certain period of time, normally a duration of one year. The values of domestic
production as well as the input components (for goods and services) of the individual sector
may be inferred from the figures, as illustrated in the sectors column of the IO-table.
Moreover, the amount of domestic production sales, as well as the import of the respective
goods and service sectors demanded, may be inferred from the sector row within the tables.
In this study, the IO-tables of Thailand from 1998 (NESDB 2003) were used to estimate the
embodied energy. Due to unavailable data, the IO-table of Bangkok has been transformed to
be based upon the national level. The IO-table for the basic sector classification consists of
180 sectors. For analysis, several intermediate sectors in the national IO-table were
consolidated to convert the matrix into 16 sectors such as food manufacturing, textile
industry, trade, services, etc., according to the economic structure of Bangkok. The IOcoefficient is estimated from the national IO-table and converted into the city level (Bangkok)
by the Cross-Industry Location Quotient Method (CIQ). For more details of this method see
Schaffer (1976). Other relevant data was collected from various statistics and materials for
the estimation of gross consumption. Energy demand data
Compared with the material flow data, the energy consumption data to detail each sector is
much more difficult to obtain; the detailed breakdown of energy demand in Bangkok cannot
be drawn from statistical data. The amount of data recorded in Bangkok is usually restricted
to the total consumption of the various energy forms e.g., diesel oil, gasoline and electricity.
In order to overcome this obstacle and to obtain the required data, the LEAP model has been
used to estimate the energy demand within the various sectors corresponding to the IOtables (Phdungsilp 2005). The estimated breakdown of direct energy consumption in
Bangkok from 1998 is shown in Table 5.1.
Since the central concept of the LEAP model is based on end-use analysis, the energy
demand is first disaggregated into various economic activities corresponding to the
consolidated IO-table. The energy intensity values in each device are collected from both
local and international studies. Other input parameters are related to a number of known or
predictable data. These parameters can then be classified into the following categories: (a)
structural characteristics e.g., GDP, population, number of dwellings, etc. (b) technological
characteristics e.g., the type and efficiency of the end-use devices, and (c) behavioral
characteristics e.g., mode of technologies used.
5.2.2. Direct and indirect energy consumption in Bangkok
The estimated breakdown of direct and indirect energy demand in Bangkok for the year 1998
is presented in Table 5.1. It shows that the transport sector consumed the majority of energy
in Bangkok. This aspect was consistent with the country's consumption. From the national
point of view, between 1991 and 2000 transportation was the largest consumer of energy
with a growth rate of 5.8% within the total energy consumption. It accounted for
approximately 40% of the total energy demand in Thailand.
The indirect energy demands of Bangkok were estimated and compared with the direct
energy demands. Table 5.1 shows the comparison between the estimated breakdown
between direct and indirect energy demands in Bangkok for 1998. The total direct and
indirect energy demand is shown in Figure 5.5. It should be noted that household
consumption is only accounted in the direct energy demand due to its direct payment from
residents and to avoid double counting within the calculation. According to Figure 5.5,
indirect energy demand is more significant than direct energy demand in Bangkok. This can
be explained by the fact that Bangkok has a great reliance upon the outside in terms of
energy demand.
Table 5.1. Comparison estimated breakdown direct and indirect energy demand in Bangkok
in 1998.
Food manufacturing
Textile industry
Saw mills & wood products
Paper industries & printing
Rubber, chemical & petroleum
Non-metallic products
Metal, metal product &
Other manufacturing
Public utilities
Transportation and
Consumption of households
Direct energy demand
Indirect energy demand
An interesting finding from the estimated direct and indirect energy demand is the deviation
in the service and trade sectors. The indirect energy consumption in these sectors is about
five times higher than the direct energy consumption. Therefore, measures (e.g., sustainable
consumption) attempting to reduce the consumption of goods and services would show a
great reduction in energy consumption. However, this point needs further research to
examine the affect.
Direct energy
Indirect energy
Figure 5.5. The estimated total direct and indirect energy demand in Bangkok for 1998.
The essence of the embodied energy analysis is an indirect reflection on the behavior that
follows after the direct energy consumption. Indirect energy consumption illustrates how a
city relies upon the outside. Certainly, the direct energy demand can be taken as an indicator
on how a city relies upon factors from the outside. However, indirect energy demand can
provide additional information for further studies as well as policy measures.
5.3. Bangkok energy modeling
The aim of this section is to develop some scenarios for Bangkok’s energy future. The
simulation model (LEAP) has been used to simulate how energy might develop in Bangkok
over the twenty-five years from 2000 to 2025. The planning period starts in 2005 and ends in
2025, with 2000 being taken as the base year. These developments are driven not only by
the nature of the energy sector itself, but also by broader factors, notably population and
economic growth, and other factors that vary with each sector. A range of policy interventions
have been selected, and how these would change energy development in Bangkok is
examined and compared to a reference case. Interventions were evaluated by various
criteria, including implementation costs, technical feasibility, environmental priority, and
political will. This method will be explained in the MCDM to evaluated policies session.
Different policies can be grouped by their sectors: commercial, industrial, residential and
transportation, respectively. These scenarios should be understood as a series of “what if”
questions, as described earlier.
5.3.1. Model of energy demand and supply
The model of the energy demand for the different sectors is formulated and structured upon
the framework of the LEAP mode. The historical data were collected from several reports of
the governmental and non-governmental agencies, which include the Ministry of Energy
(MOE), the Department of Energy Development and Promotion (DEDP), the recently named
Department of Alternative Energy Development and Efficiency (DEDE), the Metropolitan
Electricity Authority (MEA), the National Economic and Social Development Board (NESDB),
the Bangkok Metropolitan Administration (BMA), the National Statistical Office (NSO), the
Bank of Thailand (BOT), the Mass Rapid Transit Authority of Thailand (MRTA), the Bangkok
Mass Transit System (BTS), and others.
As mentioned earlier, a detailed breakdown of the energy demand in Bangkok cannot be had
from statistical data (which is primarily data of the various energy forms). Also, there has not
been any research carried out on city energy modeling in Thailand (Phdungsilp 2005).
Therefore, this study is probably the first study of energy modeling at a city level in Thailand.
It also makes an attempt to analyze the policies of intervention. Due to the sparse availability
of data, the greatest challenge in constructing the energy modeling is to set up a system that
can be efficiently analyzed. Commercial sector
In this thesis, the assumption concerning commercial buildings is based upon “Designated
Buildings” in different sub-sectors. This is due to the difficulty of obtaining the number of
buildings in Bangkok. In 1992, the Royal Thai Government promulgated an Energy
Conservation Promotion Act (ECP Act) and these buildings covered by the Act were entitled
“Designated Buildings”, defined as those that are not royal buildings or palaces, embassies
or consulates, offices of international organizations, any that are established by an
agreement between Thai and foreign governments, ancient places, temples or buildings for
religious purposes, along with the buildings that meet the following conditions (NEPO 1999):
A building or buildings under the same address that are allowed by any energy
distributor to install an electricity metering device, or to install one or more
transformers whose combined capacity is 1,000 kW or 1,175 kVA and above.
A building or buildings under the same address that consume commercial energy,
including electricity and steam as of January 1 to December 31 of the previous year,
in total volume energy of 20*106 MJ or more of electrical energy equivalent.
Figure 5.6 summarizes the status of Bangkok’s commercial sector in the year 2000, the base
year of the energy model. According to DEDP, the designated buildings are divided into six
categories: office, hotel, hospital, department store, academic institute and miscellaneous.
There are about 881 designated buildings situated in Bangkok (AIT 2002).
The energy demand estimation in commercial buildings is made up of six building types and
three end-uses, including HVAC systems, lighting and others. The consumption is formulated
as a function of the number of buildings, proportion of energy services (HVAC, lighting, and
others), device efficiency and the energy intensity on a building basis for each end-use.
8.8 ktoe, 12%
Education Inst.
5.6 ktoe, 8%
31.3 ktoe, 42%
Department store
11.3 ktoe, 15%
7.6 ktoe, 10%
10 ktoe, 13%
Figure 5.6. Total commercial energy used by building types in 2000 (AIT 2002).
The projected number of designated buildings is shown in Table 5.2. It is also assumed that
during the study period, the office growth rate will be 2% per year during the years 2000 to
2015 and will drop down to 1.5% per year from 2016 to 2025. With hotels, the growth rate will
be 1.5% per year during the years from 2000 to 2015 and drop to 1% per year from 2016 to
2025. The growth for hospitals is assumed to be constant throughout the planned period.
The department store growth rate from 2000 to 2015 is assumed to be 1% per year and then
drops to 0.5% per year from 2016 to 2025. With academic institutions it is assumed that the
growth rate will be 1% per year from 2000 to 2010. With the miscellaneous category a growth
rate of 1.5% per year is assumed to be constant from 2000 to 2025.
Table 5.2. The number of projected designated buildings in Bangkok
(compiled from AIT 2002).
Department store
Academic institution
The end-use considered is electricity, and the proportions of energy services and the
breakdown of the annual energy consumption are derived from AIT (2002) and
Chirarattananon & Taweekun (2003). They are also assumed to be constant for the planned
period. These parameters are presented in Table 5.3.
Table 5.3. Average breakdown of annual energy consumption and percentage share of
energy services per building
(compiled from AIT 2002 and Chirarattananon & Taweekun 2003).
A/C system (kWh)
Average % share
Lighting (kWh)
Average % share
Others (kWh)
Average % share
20 Industry sector
The industry sector can be broadly classified as nine manufacturing sub-sectors, including
food and beverages, textile, wood and furniture, paper, chemical, non-metallic, basic metal,
fabricated metal, and others (unspecified), according to DEDP and NSO. The model
calculates the energy demand as a function of gross provincial product (GPP), and as a
proportion of utilized energy, device efficiency and useful energy intensity. Figure 5.7
summarizes the status of Bangkok's industry sector in 2000.
322 ktoe, 8%
Food & Beverages
1329.1 ktoe, 34%
Fabricated metal
281 ktoe, 7%
Basic metal
351.1 ktoe, 9%
76.8 ktoe, 2%
W ood & Furniture
20.9 ktoe, 1%
1076.9 ktoe 28%
255.4 ktoe, 7%
154.1 ktoe, 4%
Figure 5.7. Energy consumption of industrial sub-sectors in 2000.
Data of GPP is taken from the NESDB and is used to project the future GPP by using the
average annual growth rate of 3%. The proportions of GPP in the industrial sub-sectors are
assumed to be constant for the duration of the time horizon. The proportion of energy utilized
at end-use and device levels are derived from the energy consumption in the year 2000
(DEDP 2000). It is also assumed to be constant throughout the study period. The proportion
of energy utilization in industrial sub-sectors is shown in Table 5.4.
Table 5.4. Proportion of energy utilization in industrial sub-sectors (DEDP 2000).
Fuel types
Fuel oil
Fuel wood
Rice husk
Proportion of energy utilization (%)
LPG stands for Liquefied Petroleum Gas
The average device efficiencies are taken from the energy audit reports of King Mongkut's
University of Technology Thonburi (KMUTT), Thailand. The average useful energy intensity
in each sub-sector is presented in Table 5.5. However, it is too difficult to classify different
types of devices with a specific energy because of the variety of devices and also the
limitation of data within the industrial sector. It is assumed that each type of energy in each
sub-sector is used as a specific device. The useful energy intensity is estimated by the
following equation:
UEI j = ∑
( EnU i, j ∗ ηi, j )
Eq. (5.1)
where UEI j is the useful energy intensity in the industrial sub-sector j (ktoe/106 Baht),
EnU i, j is the energy type i utilized in the industrial sub-sector j (ktoe), ηi, j is the device
efficiency using fuel type i utilized in the industrial sub-sector j , and GPPj is Gross
Provincial Products of the industrial sub-sector j , (106 Baht).
Table 5.5. Average useful energy intensities in industrial sub-sectors (Tanatvanit et al. 2003).
Industrial sub-sectors
Food and beverages
Wood and furniture
Basic metal
Fabricated metal
Average useful energy intensity (ktoe/106 Baht)
0.0155091 Government sector
There is limited data in this sector; therefore, only street lighting in Bangkok is counted.
However, large government buildings are subject to the requirements of the ECP Act, so they
are considered in the commercial sector. The annual street lighting consumption data is
obtained from the Bangkok Metropolitan Administration (BMA) and the Metropolitan
Electricity Authority (MEA). Energy demand is calculated from the parameter of activity and
fuel intensity per unit of activity. Residential sector
The household sector in Bangkok is divided into two areas: the inner BMA and outer BMA.
Figure 5.8 presents the status of Bangkok's residential sector in the year 2000. The number
of households is formulated as a function of the population and the size of households as the
Ht =
Eq. (5.2)
where H t is the number of households in year t , Pt is the number of the populace in year t ,
and S t is the size of the household in year t .
Outer BMA
417 ktoe, 42%
Inner BMA
584.2 ktoe, 58%
Figure 5.8. Energy consumption of residential sector in 2000.
The size of the household of the projected population is estimated by using the log-limit
model (Tanatvanit et al. 2003). The lower limit of household size applied in this study is three
persons per household. The estimated number of households in Bangkok is shown in Table
The end-uses are organized into six categories, including cooking, lighting, space
conditioning (air-conditioning and fan), refrigeration, water heating, and others (iron, TV,
radio, etc.). The proportion of end-use devices is disaggregated into existing appliances and
high efficiency appliances, which are either in use or are likely to be introduced in the near
future. The energy intensity of each appliance can be formulated as a function of the number
of appliances, capacity of appliances, and the average usage hours. The average number of
appliances per household and usage hours per day is obtained from statistical data and
reports of the DEDP. The average capacity of conventional and high efficiency appliances is
taken from the public documents of the DEDE and market surveys, as shown in Table 5.7.
The average intensity of each end-use appliance can be estimated by the following equation:
EIi = hri ∗ Capi ∗ Noi
Eq. (5.3)
where EIi is the energy intensity of appliances, (kWh/year.HH), hri is the usage hours per
year of end-use device i , (h/year), Capi is the average capacity of end-use device i , (kW),
Noi is the average number of end-use devices i per household (unit/HH), and HH is the
Table 5.6. The number of projected households in Bangkok Metropolitan
(compiled from Tanatvanit et al. 2003).
Inner Bangkok
Outer Bangkok
Table 5.7. Energy intensity of the appliances used in the residential sector
(DEDP 1997; Tanatvanit et al. 2003).
Type of appliance
Non-electricity appliance
1 Wood stove
2 Charcoal stove
3 LPG stove
Electricity appliance (kWh/HH.yr)
1 Electric cooking stove
2 Rice cooker
3 Incandescent lamp
4 Fluorescent lamp
5 Compact fluorescent lamp
6 Refrigerator
Inner Bangkok
Outer Bangkok
7 Air-conditioner
8 Fan
9 Iron
10 Clothes washer
11 Water heater
12 VDO
13 Television
14 Radio
Figures are the average energy intensity of efficient appliances.
Energy consumption of a given technology is calculated as the number of households in
each area, the saturation of the end-use within the household, the technology share of the
end-use, the efficiency of the end-use, and the unit energy consumption of the given
technology. Total energy consumption is the sum of the different technology categories. Transport sector
In the transport sector, only road transport is considered due to its high proportion of the total
consumed energy in the transport sector and the available data. Passenger travel includes
the vehicle types of car, microbus and pickup, van and pickup, motor-tricycle, urban taxi,
fixed route taxi, motor-tricycle taxi, motorcycle, fixed route bus, bus for hire, private bus, nonfixed route truck, private truck, tractor, mass transit system, and others. Figure 5.9
summarizes the status of Bangkok’s transport sector by vehicle types for the year 2000. It is
shown that in the transport sector almost all the energy is in passenger travel. There was a
small amount of energy (electricity) used by the BTS or the sky train, which was in operation
in the year 2000.
Typically, the energy demand for transportation is estimated by using the volume of traffic in
terms of vehicle-kilometer. The projected number of vehicles is multiplied with the average
distance per year to appraise the vehicle-kilometers per year. The energy demand is
formulated as a function of the number of vehicles, average distance, proportion of vehicle
types, and fuel economy or fuel efficiency of vehicle. In the LEAP model, total passengertravel demand is expressed in terms of passenger-kilometers (pass-km). Therefore, total
travel demand can be estimated as follows:
TDt = ∑Vi,t * Dt * LFi
Eq. (5.4)
where TDt is the total travel demand in year t expressed in pass-km, Vi,t is the number of
vehicles i registered in year t , Dt is the average distance of vehicle i (km), and LFi is the
load factor of vehicle i or the average number of occupants (passenger/vehicle).
Private truck
551.7 ktoe, 7.97%
Non-fixed routh truck
123.1 ktoe, 1.78%
72.8 ktoe, 1.05%
4.6 ktoe, 0.07%
Private bus
10.7 ktoe, 0.15%
Mass transit system
3.3 ktoe, 0.05%
Bus f or hire
33.1 ktoe, 0.48%
Fixed route bus
80.2 ktoe, 1.16%
Passenger car
2718.5 ktoe, 39.25%
356.8 ktoe, 5.15%
Motorcycle taxi
28.4 ktoe, 0.41%
Fixed route taxi
9.6 ktoe 0.14%
Urban taxi
476.6 ktoe, 6.88%
1.8 ktoe, 0.03%
Microbus & pickup
800.5 ktoe, 11.56%
V an
1653.9 ktoe, 23.88%
Figure 5.9. Energy consumption of transport sector by vehicle types in 2000.
The population per vehicle tends to decrease when a country develops. The estimation of the
number of vehicles from the person-per-vehicle point of view should be of a lower limit of car
ownership to be incorporated into the estimated number of vehicles (Tinakorn &
Sussangkarn 1996). The estimation can be done with the log-limit equation where the limit is
incorporated into the GDP and the population. The number of vehicles can be estimated by
the following equation:
Vi,t = e( a i + bi *GDPt + c i * POPt ) + LVi
Eq. (5.5)
where ai , bi and ci are the coefficients in the model of the vehicle i , GDPt is the Gross
Domestic Product in the year t , POPt is the number of the populace in the year t , and LVi
is the lower limit of vehicle type i . The lower limits of each vehicle are used in this thesis, a
comparison of model-based forecasting, and the actual number of registered vehicles in the
year 2000 are presented in Tables 5.8 and 5.9, respectively.
The travel demand for road transportation is estimated on the basis of Eqs. (5.4) and (5.5),
shown in Table 5.10. In the LEAP model, sub-sector and end-use levels, the shares of total
passenger travel demand, and the different vehicles are estimated on the basis of Eq. (5.4).
Each vehicle is disaggregated into existing technologies and alternative technologies or
fuels. The energy intensity of vehicle technologies under device levels are calculated on the
basis of fuel economy and occupancy, and re expressed in terms of liters per pass-km
(l/pass-km) as follows:
EIVi =
( FE i * LFi )
Eq. (5.6)
where EIVi is the energy intensity of vehicle i (l/pass-km) and FE i is the fuel economy of
vehicle i expressed in vehicle-kilometer per litre (veh-km/l). The estimated energy intensity
of the vehicle type is presented in Table 5.11. In a mass transit system the estimation of
transport demand is based upon the load factor, average distance and operating hours,
which are taken from MRTA 2005 and BTS 2005.
Hence, the energy demand in the transport sector is calculated as a function of the number
of vehicles, average distance, proportion of vehicle types and the fuel economy or fuel
efficiency of vehicle. Model of electricity generation
Most of the analysis in this study is focused upon the demand-side, but the demand has to
be met by the supply-side, which is transformation in the LEAP model. Electricity generation
is included in the transformation module. Apart from ensuring that the energy system
balances, there are two major environmental issue concerns: emission of GHGs contributing
to global climate change and local air pollution. This characteristic allows the analysis of both
Electricity generation calculations depend upon demand sector calculations. This sector
specifies the technology attributes, merit order, and annual system load curve shape of
electricity generation sector (SEI-Boston 2005). Actual electricity generation in any given
model year depends on the level of electricity consumption generated by the five demand
sectors and the level of imports (national grid). The model structure characterizes the
generation sector, and then based on the level of electricity required to meet the annual
demand requirements, it dispatches technologies to generate the needed electricity.
The electricity generation modeled in the Bangkok energy modeling is significant in
examining the annual magnitude and composition of power generation. The Bangkok model
offers a simple and transparent framework for exploring both the magnitude and the
composition of the generation of electricity for the duration of the study period. This feature is
useful for the implementation of the generation of electricity from renewable energy sources.
Table 5.8. Lower limits of persons per vehicle.
Vehicle types
Actual no. of
persons per vehicle
in 2000
Other studies
Study Aa
1. Passenger car
2. Microbus and pickup
3. Van and pickup
4. Motortricycle
5. Urban taxi
6. Fixed route taxi
7. Motortricycle taxi
8. Motorcycle
9. Fixed route bus
10. Bus for hire
11. Private bus
12. Non-fixed route
13. Private truck
14. Tractor
15. Small rural bus
16. Others
Study A is obtained from Tinakorn & Sussangkarn 1996.
Study B is obtained from Tanatvanit et al. 2003.
This study
Study Bb
Table 5.9. Comparison of model-based forecast and actual number of registered vehicle in
2000 (extracted from Tanatvanit et al. 2003).
Vehicle types
1. Passenger car
2. Microbus and pickup
3. Van and pickup
4. Motortricycle
5. Urban taxi
6. Fixed route taxi
7. Motortricycle taxi
8. Motorcycle
9. Fixed route bus
10. Bus for hire
11. Private bus
12. Non-fixed route truck
13. Private truck
14. Tractor
15. Others
Whole country
Table 5.10. Estimated travel demand for road transportation in Bangkok
(partially extracted from Tanatvanit et al. 2003).
Vehicle types
1. Passenger car
2. Microbus and pickup
3. Van and pickup
4. Motortricycle
5. Urban taxi
6. Fixed route taxi
7. Motortricycle taxi
8. Motorcycle
9. Fixed route bus
10. Bus for hire
11. Private bus
12. Non-fixed route truck
13. Private truck
14. Tractor
15. Others
16. Mass transit system
Estimated travel demand (106 pass-km)
Table 5.11. Fuel economy of the automotive technologies by fuel types in Bangkok
Metropolitan Area (Chanchaona et al. 1997).
Vehicle types
1. Passenger car
2. Microbus and pickup
3. Van and pickup
4. Motortricycle
5. Urban taxi
6. Fixed route taxi
7. Motortricycle taxi
8. Motorcycle
9. Fixed route bus
10. Bus for hire
11. Private bus
12. Non-fixed route truck
13. Private truck
14. Tractor
15. Others
Average fuel economy (l/km)
5.3.2. Energy policies and scenarios
The LEAP model is implemented in four sectors, commercial building, industrial, residential
and transport, to illustrate the effects of alternative strategies on energy utilization and
emissions. Furthermore, the supply-side policy considered meets a target of 10% electricity
generation from renewable resources in a Bangkok energy system by 2025.
Individual policies have been created for the sectors (except the government sector), then
combined in sectoral combinations, and finally brought together in an all policy scenario. For
example, in the residential sector policies simulating the promotion of high efficiency
appliances, passive design, and daylighting application are outlined separately. Residential
policies are in turn combined with commercial, industrial and transport policies. Examining
these policies together avoids double counting in the energy system. The policy question
addresses the implications of what if a policy was implemented according to the assumptions
stated? The question is “what if the energy policy in Bangkok changed the development path
away from business-as-usual?” – what would the policy impacts be?
The study focuses on three key implications, including energy savings, local pollutants, and
the avoidance of GHG emissions. It should be noted that local pollutants are analyzed at
source, for instance where the combustion of the fuel leading to the pollution occurs. The air
pollution problem in Bangkok exemplifies this problem, with local transport emissions in
particular contributing to poor air quality.
The approach taken with global pollution is different. Emissions of GHGs (notably CO2)
contribute to global climate change, and it does not matter where the emissions occur. Most
of the GHG emissions are due to the use of fossil fuels and actually occur at power stations.
However, LEAP allows these emissions to be analyzed either at the source or to be
attributed to the demand sector. Business-as-usual scenario
The reference case in LEAP represents a base case without policy interventions. It is often
referred to as the “business-as-usual” (BAU) scenario. In other words, it is a projection of
what would happen in the absence of any specific energy policies and strategies.
The BAU scenario is constructed upon the current trends of parameters in each sector and
assumed to be continuously increasing. The commercial sector is assumed to grow with the
number of buildings. The industry grows with the GPP in the manufacturing sector at 3%
annually. The government sector is assumed to be constant during the study period. The
residential sector will grow by the number of households in each area. It is also assumed that
the present efficiency of any appliances and technologies, and the pattern of energy
utilization for different appliances and technologies, is unchanged in the future. The transport
sector is assumed to grow with the estimated travel demand and number of vehicles. Alternative policies and interventions
Alternative scenarios aim to illustrate the effect of policies and interventions on energy
utilization and emissions. As mentioned earlier, the study focuses on energy savings, local
pollutants and the avoidance of GHG emissions. Therefore, the selection of policies and
interventions to be explored in LEAP was based upon those three factors. In addition, the
use of renewable energy technologies is of interest in the Bangkok energy model. A total of
sixteen policies have been selected for simulation. Table 5.12 summarizes the policies and
interventions within the modeling.
Table 5.12. Description of energy policies and interventions for sustainable energy
development in Bangkok.
Policy or measure
Promoting high efficiency appliances (RHEFF:
Res. High Eff. Appliances)
Passive design and daylighting application
(RPASS: Res. Passive design & daylighting)
Conventional refrigerators, air-conditioners, and fans are
replaced by high efficiency ones by 2025.
New households in outer the Bangkok Metropolitan Area (BMA)
will be built on passive design and improve lighting by daylighting
application, consequently reducing cooling load (10%) and
energy consumption for lighting (10%), starting 2005.
Transport (T)
Industry (I)
building (C)
Efficient HVAC system (CHAVC: Com. HVAC)
Improving efficiency of HVAC 10% by 2005 in every building
Utilization of daylighting in lighting system Utilization of daylighting can improve the lighting systems.
(CDAY: Com. Daylighting)
Assume 5% savings in lighting by 2005 and increasing this to
10% in 2010, and 15% in 2020, respectively.
Behavioural change in HVAC and lighting 10% savings in HVAC systems by 2025, and 10% savings in
systems (CBEH: Com. Behavioural change)
lighting systems by 2025, driven by changes in user behaviour.
Energy efficiency (IEFF: Ind. High Efficiency Increasing industrial energy efficiency target of 10% by 2010.
The improvements are from lighting, compressed air, motors, as
well as improved boiler and steam system efficiency.
Switching to natural gas (ING: Ind. NG The Thai government has the policy to promote the use of
natural gas (NG), which is a domestic resource and the country’s
major source of energy. Thermal energy supplied by nonrenewable resources, LPG and electricity in industries switches
fuels to NG by 2010.
Introducing NGV to gasoline and diesel Introducing NGV to gasoline and diesel vehicles: passenger car,
vehicles (TNGV: Tran. NGV)
microbus & passenger pickup, van & pickup, fixed routed taxi,
fixed route bus, bus for hire, private bus, non-fixed route truck,
private truck, and other. Penetration rates are about 5% by 2005,
rising to 10% in 2010 and 20% in 2025.
Switching to gasohol in gasoline vehicles Thai government has a policy to switch all gasoline vehicles to
(TGSH: Tran. Gasohol)
gasohol (gasoline + ethanol) starting in 2007. This measure will
be applied to passenger car, microbus & passenger pickup, van
& pickup, motortricycle, urban taxi, fixed routed taxi, motortricycle
taxi, and motorcycle.
Introducing biodiesel in diesel vehicles (TBID: Biodiesel grows to a market share of 20% by 2025. This
Tran. Biodiesel)
measure will be applied to all diesel vehicles.
Modal shift from private passenger to mass Increasing the share of mass transit of 40% by 2015, rising to
transit systems (TSHIF: Tran. Modal shift).
60% by 2025. Considering for Passenger car, microbus &
pickup, van & pickup, and urban taxi
A target of 10% electricity generation from renewable resources by 2025. It is assumed that in order to meet the
target electricity will generate from biogas PV will be installed in households and buildings, municipal solid waste,
and solar thermal electricity.
Biogas Generation (BIO: Biogas generation)
Electricity produced from biogas. Assume 100 MW of plant with
capacity factor of 0.7 by 2025.
PV installed in households & buildings (PV: PV Assuming 500 MW of installed PV in households and buildings
in Household and Building)
by 2025.
Municipal Solid Waste (MSW: Municipal Solid Assuming 120 MW of MSW plant will be installed by 2025.
Solar Thermal Electricity (SOEE: Solar Assuming 400 MW of Solar Thermal Electricity plant by 2025.
Thermal Electricity)
Renewable Electricity (EE: EE Renewable)
Including all scenarios from supple-side.
5.3.3. Results and discussion of the scenarios The BAU scenario
The key drivers of the energy development and the assumptions used in the reference case
or BAU scenario have already been described in section The BAU scenario in the
model shows that the total energy demand is estimated to be about 14,320 ktoe and 22,915
ktoe in 2005 and 2025, respectively. Table 5.13 shows the final energy demand for each
sector. It can be seen that the transport sector, which focused only on road transportation,
takes the biggest share of energy consumption in Bangkok. The energy demand is estimated
at 8,692 ktoe in 2005 and 13,445 in 2025. Table 5.14 shows the energy demand for different
fuels in the BAU scenario. It shows that diesel and gasoline are the major energy sources in
Bangkok, which has an approximately 60% share of the total energy consumption. Detailed
estimations of the energy demand for the different sectors under the BAU scenario are
shown in Appendix A.
Table 5.13. Final energy demand by sectors in BAU scenario (ktoe).
Table 5.14. The energy demand by fuel types in BAU scenario (ktoe).
Fuel oil
Natural gas
22,915.1 Policies for the commercial sector
The main energy use in commercial buildings is electricity; therefore, the selected policies
are intended to reduce electricity use by promoting high efficiency appliances, introducing the
use of renewable energy technologies, as well as a behavioral change regarding the users.
The energy savings that can be gained from implementing efficient HVAC systems and the
utilization of daylighting as well as changing user behavior in commercial buildings are
shown in Figure 5.10. The efficient HVAC systems policy shows the largest potential savings,
while the user's behavioral policy shows a greater potential for reduction than the daylighting
policy. The deviation of the daylighting policy to reference the case might seem small;
however, the total savings in 2025 are estimated to be 16.2 GWh. The energy savings in
commercial buildings are in the form of electricity. Reduced electricity consumption lowers
demand and ultimately decreases the amount of CO2 emissions from power plants. Figure
5.11 shows the avoided CO2 emissions under different policies in the commercial sector.
Com. Daylighting
Com. Behaviour change
Figure 5.10. Energy savings in commercial buildings.
Thousand tons CO2 equivalent
Com. Daylighting
Com. Behaviour change
Figure 5.11. Avoided CO2 emissions from commercial buildings.
127 Policies for the industrial sector
There are two scenarios in this sector. In the first scenario an increase in energy efficiency is
assumed regarding industrial applications. The efficiency improvements occur in electrical
devices as well as with suppliers of thermal heat. The improvements are likely to come from
lighting, compressed air, motors, and improved boiler and steam system efficiency. The
second scenario follows the Thai government to promote the use of natural gas. It is
assumed that natural gas is available and a switch in the type of fuel used takes place within
the thermal energy that is supplied by non-renewable resources.
In regards to the energy efficiency scenario, industry in Bangkok meets a 10% improvement
in energy efficiency by 2010. The fuel switching scenario starts to take effect by 2010 as well.
The reduction in the final energy demand in the industrial sector that occurs as a result of
improved energy efficiency and fuel switching are shown in Figure 5.12. The potential energy
savings in the energy efficiency scenario rises from 73.5 ktoe in 2005 to 736.1 ktoe in 2025,
while the fuel switching scenario expects energy savings from 40.7 ktoe in 2005 to 441.1
ktoe in 2025. The reductions in CO2 emissions from both scenarios are presented in Figure
5.13. It shows that the fuel switching scenario has great potential for reducing the amount of
CO2 emissions. Also, it will improve the level of the local air pollutants by reducing sulfur
dioxide (SO2). However, the main source of SO2 in Bangkok comes from transportation, so
this sector does not show the local pollutant reduction under both scenarios.
9 000,00
8 000,00
7 000,00
6 000,00
5 000,00
4 000,00
3 000,00
2 000,00
1 000,00
Ind. NG Switching
Ind. High Efficiency 10%
Figure 5.12. Final energy demand of the industrial sector under different scenarios.
7 000,00
Thousand tons CO2 equivalent
6 000,00
5 000,00
4 000,00
3 000,00
2 000,00
1 000,00
Ind. NG Switching
Ind. High Efficiency 10%
Figure 5.13. CO2 reductions in the industrial under industrial policies. Policies for the residential sector
Two scenarios have been examined in the residential sector, including the promotion of high
efficiency appliances, passive design and daylighting scenarios. The fist scenario aims to
simulate the effect of replacing conventional refrigerators, air-conditioners, and fans by high
efficiency ones. The latter scenario assumes that the new households in the outer BMA will
be built with passive design and improved lighting systems with daylighting application.
These two scenarios will result in reduced electricity consumption as well as a reduction in
The potential energy savings in the residential sector under both scenarios can be seen in
Figure 5.14. The high efficiency appliances scenario shows a greater possibility of reducing
the consumption of electricity than the passive design and daylighting scenario. The energy
savings are expected to be 111 ktoe and 21.6 ktoe by 2025 under the high efficiency
appliances, and the passive design and daylighting scenario, respectively. In the residential
sector, the policies primarily reduce electricity consumption, and hence have an impact on
CO2 emissions. Figure 5.15 illustrates the avoided CO2 emissions caused by residential
policies. The reduction of CO2 in the high efficiency appliances scenario is consistent with the
potential for electricity savings. The amount of reduced CO2 is expected to be 33.6 thousand
tons CO2 in 2005 and 721.3 thousand tons CO2 in 2025. In the passive design and
daylighting scenario, reduction in CO2 is almost constant throughout the study period.
1 400,00
1 200,00
1 000,00
Res. Passive design & daylighting
Res. High Eff appliances
Figure 5.14. Trend of energy savings in residential sector.
Thousand tons CO2 equivalent
Res. Passive design & daylighting
Res. High Eff appliances
Figure 5.15. Avoided CO2 emissions from residential policies. Policies for the transport sector
As mentioned earlier, the transport energy policies focus on land-based transportation only.
Air travel is not included in the model due to the difficulties in obtaining data and
differentiating the energy use and environmental impacts within the boundaries of Bangkok,
since most of the travel would be outside of the boundary. Air travel would have an influence
on the more economic aspects rather than just the energy and environmental aspects.
Policy interventions in the Bangkok transport sector involve introducing cleaner fuel, such as
introducing NGV, gasohol and biodiesel along with behavioral changes e.g., shifting transport
modes from a private to a mass transit system. There are no energy savings for the
introduction of cleaner fuel policies because it assumes that the cleaner fuels that replace
existing fuels have the equivalent energy per liter. However, the introduction of cleaner fuel
scenarios can decrease emissions and improve the level of local pollutants. The modal shift
from a private transport system to a mass transit system results in large energy savings, as
shown in Figure 5.16.
Due to the assumption in the modeling that the penetration rate of the introduction of cleaner
fuels would be low, the reductions in the CO2 emissions results change only a little compared
to the reference case. Hence, only CO2 reduction from the modal shift scenario is shown in
Figure 5.17. Local emissions of (SO2) are reduced in all transport scenarios over the time
horizon, but they are more so within the modal shift policy. However, introducing the NGV
policy has a very small change compared to the reference case. Figure 5.18 shows the
overall reduction of local pollution regarding transport under the modal shift, gasohol and
biodiesel scenarios. In the modal shift scenario, the reduced local pollutants show a
significant decrease due to a shift to a much more efficient means of transport.
14 000,00
12 000,00
10 000,00
8 000,00
6 000,00
4 000,00
2 000,00
Tran. Modal shift
Figure 5.16. Trend of energy savings under the transport modal shift scenario.
35 000,00
Thousand tons CO2 equivalent
30 000,00
25 000,00
20 000,00
15 000,00
10 000,00
5 000,00
Tran. Modal shift
Figure 5.17. The transport CO2 reductions under the modal shift scenario.
Thousand tons SO2
Tran. Modal shift
Tran. Gasohol
Tran. Biodiesel
Figure 5.18. Local pollutant reduction in the transport sector. Electricity generation from renewable energy technologies
The model is assumed to have a target of 10% concerning the electricity generated from
renewable energy technologies by 2025. These requirements would be met by electricity
being produced by biogas, PV installed in households and buildings, municipal solid waste
(MSW) plants, and solar thermal electricity. In the model, process shares are set for
individual policy, for example the PV installed has a process share of 3%, but most of the
electricity is imported from the national grid. In the EE Renewable scenario (see Figure 5.19),
it is assumed that all the scenarios where electricity generation is from renewable energy
technologies are implemented together.
The results of the scenarios simulation show that the renewable electricity scenario has a
significant reduction in CO2 emissions. The expected CO2 emissions would be 1,757
thousand tons CO2-equivalent by 2025. Figure 5.19 shows the CO2 reduction under different
scenarios in the supply-side model. Among the different policies within the electricity
generation scenario, the installation of PV in households and buildings shows the largest
avoidance of CO2 emissions, which would reach about 831 thousand tons CO2-equivalent,
while the solar thermal electricity, MSW, and biogas policies are expected to be, 671, 207,
and 173 thousand tons CO2-equivalent by 2025, respectively.
2 000,00
Thousand tons CO2 equivalent
1 800,00
1 600,00
1 400,00
1 200,00
1 000,00
Solar thermal EE
EE Renewable
Figure 5.19. Avoided CO2 emissions with renewable energy target.
The LEAP model represents a powerful tool for the analysis of energy planning with its
associated environmental impacts. It can provide disaggregated energy demand in different
sectors in particular areas. The Bangkok energy modeling examined in this study would be
helpful to energy planners as well as to the city's administration in developing an energy
plan, which regards sustainability.
Regarding the policy interventions in this study, they are considered with the broader policy
goals of Bangkok in its effort to develop in a sustainable way. Some policies are consistent
with government policies, some policies have cost savings, while others meet social or
environmental goals. It is also important to consider whether or not policies can be
implemented practically. All this balanced with some policies that would cause long-term
change within Bangkok energy systems. Nevertheless, together this basket of policies could
reduce local pollutants as well as help address global climate change. If all strategies on the
demand side are simultaneously implemented the highest potential of energy savings in
2025 is expected to be 7,947 ktoe.
The modeling analysis outlines the implications of different policy scenarios. Because
policies are not implemented as packages, the study provides analysis for each policy.
However, the LEAP framework ensures that the total energy supply and demand match,
avoiding the double-counting that sometimes occurs with policy-by-policy analysis (Winkler et
al. 2005).
5.4. Energy decision with multi-criteria decision-making
MCDM and computer-based decision support systems can provide ways to systematically
structure and analyze complex energy decision problems. The MCDM approach can help
decision-makers organize and synthesize information in a way that leads them to feel
comfortable about making a decision. With these systems, individuals can evaluate and
compare policy alternatives (Mustajoki et al. 2004). The following section shows the results
of the MCDM method as described in Chapter 4, along with a discussion of the scenarios
evaluation and lessons learned from this approach.
Sustainability is an interdisciplinary task; therefore, teamwork between different but related
disciplines makes it possible to face the challenge of solving the problem. In order to
evaluate criteria and alternatives in this study, members of the group consisted of Ph.D.
students at the Royal Institute of Technology, Stockholm, Sweden. Each member in the test
group played the role of real members in a decision-making team, including politician,
engineer, economist and environmentalist.
5.4.1. Weighting of criteria
The performance prediction for the alternatives was done using estimations based upon
rules of thumb and the experience and knowledge of the team members. Computer
simulation was carried out for the predicted energy demand. The criteria and sub-criteria
were weighted directly. The weights were produced through individual opinion corresponding
to their role-play. Data from members were entered into Web-HIPRE. The weighted main
criteria are shown in Figure. 5.20. This chart shows the relative importance of weights for the
criteria that were used in creating the total weighted scores for the different alternatives.
Resource use
Financial &
Figure 5.20. Results of main criteria weighting.
The residential sector was done with two policies: promoting high efficiency appliances (Alt1:
RHEFF), and passive design and daylighting applications (Alt2: RPASS). Figure 5.21 shows
the total weighted scores for the alternatives with the contributions of the different criteria
indicated. Star diagrams showing the score for different policies in the residential sector are
shown in Figure 5.22. A star diagram shows an overall picture of how well the alternative
scores in regards to the different criteria. The center of the star represents the lowest point
on the scale. Thus, the better the score, the more the star will be filled or shaded. Based on
the figure and the star diagrams, the preliminary findings are: alternative 1 (Alt1: RHEFF)
gets a total weighted score of 0.852 and alternative 2 (Alt2: RPASS) has a total score of
0.848. Both alternatives seem to get similar values. These alternatives have superior scores
on resource use.
Financial & economic
Environment loading
Resource use
Figure 5.21. Total weighted scores for the alternatives in the residential sector.
Resource use
Resource use
Environment loading
Environment loading
Financial & economic
Financial & economic
Figure 5.22. Star diagrams for residential energy policies.
The commercial sector consists of three policies: efficient HVAC system (Alt1: CHVAC),
utilization of daylighting in lighting systems (Alt2: CDAY), and behavioral change in HVAC
and lighting systems (Alt3: CBEH). Figures 5.23 and 5.24 show the total weighted scores for
the alternatives and the star diagrams for the commercial energy policies. Alternative 2 (Alt2:
CDAY) seems to be the best scheme with a total score of 0.801, while alternative 1 (Alt1:
CHVAC) and alternative 3 (Alt3: CBEH) get approximately equal values of 0.797 and 0.762,
Financial & economic
Environment loading
Resource use
Alt2: CDAY
Alt3: CBEH
Figure 5.23. Total weighted scores for the alternatives in the commercial sector.
Resource use
Resource use
Environment loading
Environment loading
Financial & economic
Financial & economic
Alt2: CDAY
Resource use
Environment loading
Financial & economic
Alt3: CBEH
Figure 5.24. Star diagrams for commercial energy policies.
Two policies have been examined in the industrial sector. The total weighted scores of these
alternatives are shown in Figure 5.25. The energy efficiency policy (Alt1: IEFF) receives a
higher score (0.836), while the switching to natural gas (Alt2: ING) gets a total weighted
score of 0.8. Star diagrams for industrial energy policies are shown in Figure 5.26.
Alternative 1 (Alt1: IEFF) has a superior score on resource use and environment loading.
Financial & economic
Environment loading
Resource use
Alt1: IEFF
Alt2: ING
Figure 5.25. Total weighted scores for the alternatives in the industrial sector.
Resource use
Resource use
Environment loading
Environment loading
Financial & economic
Alt1: IEFF
Financial & economic
Alt2: ING
Figure 5.26. Star diagrams for the industrial energy policies.
The transport sector is comprised of four policy interventions; the evaluations of these
policies are shown in Figure 5.27. Alternative 4 (Alt4: TSHIF), the modal shift from private
passenger to mass transit system, receives the highest score of 0.876. It has a superior
score on resource use, environment loading as well as on social aspects (see star diagrams
in Figure 5.28). Introducing biodiesel policy (Alt3: TBID) gets a total weighted score of 0.802,
while the introduction of NGV (Alt1: TNGV) and switching to gasohol (Alt2: TGSH) policies
get approximately equal values of 0.773 and 0.775, respectively.
Financial & economic
Environment loading
Resource use
Figure 5.27. Total weighted scores for the alternatives in the transport sector.
Resource use
Resource use
Environment loading
Financial & economic
Financial & economic
Alt1: TNGV
Alt2: TGSH
Resource use
Resource use
Environment loading
Environment loading
Environment loading
Financial & economic
Alt3: TBID
Financial & economic
Figure 5.28. Star diagrams for the transportation energy policies.
A similar assessment was carried out for the supply-side. A star diagram for the target of
10% of the electricity being generated from renewable sources is shown in Figure 5.29. This
alternative receives a total weighted score of 0.844, with the highest contribution to resource
use (0.189) follow by environment loading (0.179), social (0.167), financial and economic
(0.163), and practical (0.146), respectively.
Resource use
Environment loading
Financial & economic
Figure 5.29. A star diagram for the alternative on the supply-side.
5.4.2. Lessons learned
Based on the evaluation of the exercises, the test group found that the hierarchical way of
structuring the criteria was a useful tool in enabling them to have an overview, while
simultaneously focusing on what is most important. The star diagrams were found to be
valuable in comparing individual attribute performances across alternatives, but they could
not be used to compare overall performances unless the weights were equaled. A star
diagram is one of the useful formats in the evaluation. Such a diagram is produced for each
scenario. By inspecting these diagrams, the team could get a quick understanding of the big
picture. The diagram of the selected scenario might even make a good graphic to display in
the energy system scenario, showing how it performs according to all relevant criteria. By
visually inspecting these diagrams the team could attain an immediate perception of the
overall situation.
In general, the MCDM method provides a methodology that helps decision-makers
systematically identify management objectives and priorities and then to develop alternatives
to achieve those objectives. It has proved to be a useful tool for decision-making regarding
energy. Additionally, the test group felt that the approach was a useful way to document the
results of the evaluations. They felt more comfortable that the best solution had been found
when this structured approach was used. Furthermore, the method encourages team
members to gain a deeper understanding of each other's field of expertise, for example the
energy engineer learned more about politics and the environmentalist learned more about
However, some members suggested that the method should incorporate uncertainty and risk
analysis. This is due to the fact that the future costs and benefits that the alternatives may
bring include degrees of uncertainty. For example, it is hard to predict exactly how energy
prices will develop during the next 25 years. The ability to predict and evaluate the future
consequences of alternatives may be crucial to the success of policy making. How then,
should the uncertainties best be accounted for in the evaluation of the alternatives?
In the literature, some researchers argue that risky procedures are useful because they
make the decision-maker think about the decision within the context of prevailing
uncertainties, and therefore they can help to clarify preferences and values (Keeney 1980).
This, however, may be adequately achieved through sensitivity analysis.
5.5. References
AIT (2002), The Adjustments to the Building Energy Code. Asian Institute of Technology,
Bangkok, Thailand.
BTS (2005), Bangkok Mass Transit System Public Company Limited. Available at, as accessed 10.4.2005.
Chanchaona S., Suwantragul B., Sasivimolphan S., Jugjai S., Chuntasiriwan S.A. (1997),
Study of strategies for energy conservation in vehicles. Department of Mechanical
Engineering, King Mongkut’s University of Technology, Thonburi (KMUTT), Thailand.
Chirarattananon S., Taweekun J. (2003), A technical review of energy conservation
programs for commercial and government buildings in Thailand. Energy Conversion and
Management 44 (5), pp.743-762.
DEDE (2004), Thailand energy situation 2004. Department of Energy Development and
Promotion, Ministry of Energy, Thailand.
DEDP (1997), The residential energy utilization in rural area. Department of Energy
Development and Promotion, Ministry of Science, Technology and Environment, Thailand.
DEDP (2000), Thailand energy situation 2000. Department of Energy Development and
Promotion, Ministry of Science, Technology and Environment, Thailand.
Keeney R.L. (1980), Siting energy facilities. Academic press, New York.
MOEN (2005), Ministry of Energy, Thailand. Available at, as accessed
MRTA (2005), Mass Rapid Transit Authority of Thailand. Available at, as
accessed 10.4.2005.
Mustajoki J., Hämäläinen R.P. (2004), Participatory multicriteria decision analysis with
Web-HIPRE: a case of lake regulation policy. Environmental Modelling & Software 19,
NEPO (1999), Thailand Energy Strategy and Policy. National Energy Policy Office,
Bangkok, Thailand.
NESDB (2003), Input-Output Tables of Thailand. National Economic and Social
Development Board (NESDB), Thailand. Available at, as accessed
Phdungsilp A. (2005), Towards Sustainable Urban Energy Use in Cities: A Metabolism
Approach. In Proceedings of the 2005 World Sustainable Building Conference, Tokyo,
September 27-29, 2005, Japan.
Schaffer W.A. (1976), On the use of input-output models for regional planning. Martinus
Nijhoff, Leiden.
SEI-Boston (2005), Long-range Energy Alternative Planning (LEAP) System Version
2005. Stockholm Environment Institute – Boston, USA. Available from:
Tanatvanit S., Limmeechokchai B., Chungpaibulpatana S. (2003), Sustainable energy
development strategies: implications of energy demand management and renewable
energy in Thailand. Renewable and Sustainable Energy Review 7, pp.367-395.
Tinakorn P., Sussangkarn C. (1996), Analysis and forecast of registered motor vehicles
and of car ownership in Thailand. Thailand Development Research Institute (TDRI)
Winkler H., Borchers M., Hughes A., Visage E., Heinrich G. (2005), Cape Town energy
futures: polices and scenarios for sustainable city energy development. Energy Research
Centre, University of Cape Town, South Africa.
6.1. Conclusions
In this thesis an integrated approach to studying energy utilization in cities has been
developed and applied. An integrated approach seeks to incorporate all significant aspects
into a holistic synthesis. The study begins with understanding energy flows and analyses
their flows by the urban metabolism method, and follows with the application of the proposed
energy modeling and decision support in energy planning. These applications can be
considered to be helpful tools for energy planners as well as city administrations to develop
their energy plans in regard to sustainability issues.
The study describes a number of factors that influence energy use in cities. It shows that the
major factors driving energy consumption include urban settlements, patterns of
urbanization, income growth and social change, the structure of economic activities, and
lifestyle. Income and lifestyle changes are particularly significant in Asian Mega-cities such
as Bangkok, Beijing and Shanghai. Improvements in energy intensity due to positive
technological change and higher productivity of energy have played an important role in
reducing energy use. The improvement of fuel quality and fuel switching would also help to
reduce energy consumption as well as emissions. Energy efficient technologies, industrial
processes and building technologies have a significant effect upon energy use. The urban
effects on energy use are also examined; the urban heat island is worsening in most Asian
mega-cities and air pollution is affecting residents in terms of sickness, loss of productivity,
and in other human development opportunities.
While direct energy consumption is discussed explicitly, the energy embedded in
consumable goods and services are often neglected in the analysis of energy. The true
energy use of a city needs to be clarified before alternative urban development paths are
explored. This requires detailed analyses of the consumption activities of urban dwellers. The
urban metabolism based upon embodied energy analysis can provide some sense of the
size of the total energy consumption. IO-table analysis estimates of indirect energy demand
show interesting results for Bangkok, Beijing, Shanghai and Tokyo. In Bangkok, Shanghai
and Tokyo, indirect energy demands are more significant than direct energy demands. The
volume of indirect energy demand in some Asian mega-cities, such as Bangkok and Tokyo,
could be over two times that of the direct energy consumption.
To minimize the contributions of cities to national as well as global environmental change,
reductions in only direct energy consumption are not enough. The consumption of large
amounts of material goods by cities has an indirect effect on places outside cities where
manufacturing and resource extraction take place. This is because cities do not only
consume material goods but also supply them in the form of exports. The relationship
between the direct and indirect energy consumption for which a city is responsible differs
from city to city depending on its scale of industrialization and the type of industries. This
provides the message that policy-makers should at least regard indirect energy consumption
as an issue worthy of their attention.
The case study has confirmed that the LEAP model as a tool in energy planning can provide
valuable insights into and to contribute to decision-making. Energy modeling can thus be an
effective tool to assist policy-makers in evaluating different scenarios and their environmental
consequences. In the study of Bangkok, the results reveal that the most significant energy
savings are in the transport sector relative to the BAU scenario. The scenarios analysis
shows that the implementation of modal shift from private passengers to mass transit
systems has a great potential to reduce energy demand, CO2 emissions, and improve the
level of local air pollutants. This scenario has the potential of an energy saving of 6,614 ktoe
in 2025. In the industrial sector, the improvement of energy efficiency has shown savings of
736 ktoe in 2025. The electricity savings under the promotion of the high efficiency
appliances scenario in the residential sector and the potential of savings under the efficient
HVAC systems scenario in commercial building are expected to be 111 ktoe and 12.3 ktoe in
2025, respectively. These energy savings are important to Bangkok, since it depends upon
imports of both electricity and fuels. However, if all strategies on the demand-side are
simultaneously implemented, the highest potential of energy savings in 2025 is expected to
be 7,947 ktoe, while on the supply-side the renewable electricity scenario shows a significant
reduction in CO2 emissions. There would be a reduction of 1,757 thousand tons CO2equivalent by 2025.
The policy interventions are considered within the broader policy goals of Bangkok, and in an
effort to develop in a sustainable manner some policies are consistent with government
policies. Some policies are viable in terms of investment costs, social benefits, and the
environment. The behavioral change in HVAC and the lighting systems in commercial
buildings stand out as policies that have benefits from every angle. Bangkok has long
suffered from the problem of traffic jams and high levels of air pollution. The modal shift
policy within the transportation sector plays a key role in reducing both traffic problems and
local air pollution. Hence, these scenarios should be implemented immediately.
As we have seen sustainability is an interdisciplinary question and it is cross-discipline
teamwork that makes it possible to answer the question. The MCDM process has proved to
be a powerful tool in energy decision-making. This system provides a methodology to help
decision-makers systematically identify management objectives and priorities, and to develop
alternatives to achieve those objectives. It has proved to be a useful tool in regards to the
Bangkok case study. The method was applied in order to assess the outcome of the
implementation of different policies within individual sectors; a similar assessment was also
carried out for a supply-side target of 10% electricity being generated from renewable
6.2. Future work
This study has been on an integrated approach, one that includes urban metabolism, energy
modeling and an MCDM approach. Regarding the metabolism of a city, a full urban
metabolism with respect to energy, materials, water, and nutrients should be quantified.
There are practical reasons for understanding urban metabolism. The vitality of cities
depends upon its spatial relationships with its surrounding and global resources. Metabolic
processes that threaten the sustainability of cities are identified. Urban policy-makers should
be encouraged to understand the urban metabolism of their cities. It is useful for them to
know if they are using energy, water, materials and nutrients efficiently, and how this
efficiency compares to other cities.
The metabolism approach to cities is considered a purely biological view, but cities are much
more than a mechanism for processing resources and producing wastes: they are about
creating human opportunity. Therefore, a basic metabolism concept should include a city's
livability such as health, employment, income, education, leisure activities, etc. Therefore,
this kind of study will involve the economic and social aspects of sustainability, together with
the environmental aspect. This extended metabolism can be applied at a range of levels and
to a range of different human activities, for instance city comparisons, industrial areas,
demonstration of urban projects and individual businesses.
Further study in the LEAP model, however, should include cost comparison among policies
as well as an energy supply mix in Bangkok. The cost savings would be an index for
decision-makers to consider in choosing subsequent actions and policies to be implemented.
The results in terms of cost savings, especially in the residential sector, can convince people
to take action. The question of the energy supply mix is; how can Bangkok move away from
its high dependence on petroleum products and still meet the demand? With this in mind,
future work should be in developing renewable energy supply scenarios and strategies with a
focus on fully renewable energy technologies.
The MCDM process provides a methodology to help decision-makers. The use of this
decision aid can clarify the issues for individual decision-makers and provide a deeper
understanding of the alternatives. The future perspective of energy decision-making with
MCDM should include a sensitivity analysis. This could allow for the identification of the most
robust solutions, which are less influenced by the weights attributed criteria. Also, it would be
of interest to assess the role of the uncertainty value within the final preference scores in a
more definite fashion.
Appendix 1. Estimated Bangkok energy demand by economic sectors under BAU
Table A.1.1. Estimated energy demand in commercial sector.
Commercial sector
- Others
- Lighting
- Others
- Lighting
- Others
- Lighting
Department store
- Others
- Lighting
Academic institution
- Others
- Lighting
- Others
- Lighting
Energy demand (ktoe)
Table A.1.2. Estimated energy demand in industrial by sub-sectors.
Food and beverages
Wood and furniture
Basic metal
Fabricated metal
Energy demand (ktoe)
Table A.1.2. Estimated energy demand in industry sector by fuel types.
Fuel oil
Natural gas
Coal and coke
Rice husk
Energy demand (ktoe)
Table A.1.3. Estimated energy demand in residential sector.
Residential sector
Outer Bangkok Metropolitan
- Water heating
- Space conditioning
- Refrigeration
- Lighting
- Cooking
- Others
Inner Bangkok Metropolitan
- Water heating
- Space conditioning
- Refrigeration
- Lighting
- Cooking
- Others
Energy demand (ktoe)
Table A.1.4. Estimated energy demand in residential sector by fuel types.
Residential sector
Energy demand (ktoe)
Table A.1.5. Estimated energy demand in transport sector.
Transport sector
Passenger car
Microbus and pickup
Van and pickup
Urban taxi
Fixed route taxi
Motortricyle taxi
Fixed route bus
Bus for hire
Private bus
Non-fixed route truck
Private truck
Mass transit system
Energy demand (ktoe)
10,507.0 11,195.5 11,884.0
Tryck: Universitetssvervice US AB
Stockholm 2006
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF