Multiple criteria decision making for sustainable energy and transportation systems proceedings of the 19th international conf

Lecture Notes in Economics
and Mathematical Systems

634

Founding Editors:
M. Beckmann
H.P. Künzi
Managing Editors:
Prof. Dr. G. Fandel
Fachbereich Wirtschaftswissenschaften
Fernuniversität Hagen
Feithstr. 140/AVZ II, 58084 Hagen, Germany
Prof. Dr. W. Trockel
Institut für Mathematische Wirtschaftsforschung (IMW)
Universität Bielefeld
Universitätsstr. 25, 33615 Bielefeld, Germany
Editorial Board:
H. Dawid, D. Dimitrov, A. Gerber, C.-J. Haake, C. Hofmann, T. Pfeiffer,
R. Slowińksi, W.H.M. Zijm
_

For further volumes:
/>

Matthias Ehrgott • Boris Naujoks
Theodor J. Stewart • Jyrki Wallenius
Editors

Multiple Criteria Decision
Making for Sustainable

Energy and Transportation
Systems
Proceedings of the 19th International
Conference on Multiple Criteria
Decision Making, Auckland,
New Zealand, 7th - 12th January 2008

ABC


Ass. Prof. Dr. Matthias Ehrgott
The University of Auckland
Department of Engineering Science
Level 3, 70, Symonds Street
Auckland 1010
New Zealand


Boris Naujoks
Login GmbH
Wilhelmstraße 45
58332 Schwelm
Germany


Professor Theodor J. Stewart
University of Cape Town
Department of Statistical Sciences
P D Hahn Building
Rondebosch 7701

South Africa


Professor Jyrki Wallenius
Helsinki School of Economics
Department of Business Technology
Runeberginkatu 22-24
00100 Helsinki
Finland


ISSN 0075-8442
ISBN 978-3-642-04044-3
e-ISBN 978-3-642-04045-0
DOI 10.1007/978-3-642-04045-0
Springer Heidelberg Dordrecht London New York
Library of Congress Control Number: 2009933604
c Springer-Verlag Berlin Heidelberg 2010
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9,
1965, in its current version, and permission for use must always be obtained from Springer. Violations
are liable to prosecution under the German Copyright Law.
The use of general descriptive names, registered names, trademarks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective
laws and regulations and therefore free for general use.
Cover design: SPi Publisher Services
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface

In the twenty-first century the world has entered an age of exponentially
increasing demand for energy and transportation services in a globalised economy.
The evidence for climate change as a consequence of human activity and a growing
realization of limited resources has put the sustainability of energy and transportation systems on the top of the political agenda in many countries around the world.
Economic and technological growth as well as the development of infrastructure
must consider the sustainability of such activity for the future and governments are
establishing policies towards a sustainable, low emissions energy future.
The environmental impacts of human economic activity necessitate the consideration of conflicting goals in decision-making processes to develop sustainable
systems. Any sustainable development has to reconcile conflicting economic and
environmental objectives and criteria. The science of Multiple Criteria DecisionMaking (MCDM) has a lot to offer in addressing this need. Decision-making with
multiple (conflicting) criteria is the topic of research that is at the heart of the
International Society of Multiple Criteria Decision-Making. To provide a forum
for the discussion of current research the Society organised the 19th International
Conference under the theme “MCDM for Sustainable Energy and Transportation
Systems”.
This book is based on selected papers presented at the conference, held at
The University of Auckland, New Zealand, from 7th to 12th January, 2008. The
conference was attended by 137 people from 39 countries on six continents.
125 papers were presented in 39 scientific sessions, including two plenary
addresses by Prof. Anna Nagurney, University of Massachusetts, on “Multicriteria Decision-Making for the Environment: Sustainability and Vulnerability Analysis
of Critical Infrastructure Systems from Transportation Networks to Electric Power
Supply Chains” and Prof. Jim Petrie, University of Sydney and University of Cape
Town, on “Multi Criteria Decision-Making within Energy Networks for Electricity
Production in Emerging Markets”.
The International Society on Multiple Criteria Decision-Making awards prizes to
outstanding researchers in the field. The winners in 2008 were:

MCDM Gold Medal: Prof. Theodor J. Stewart, University of Cape Town
Edgeworth-Pareto Award: Prof. Kalyanmoy Deb, Indian Institute of Technology
Kanpur
Georg Cantor Award: Prof. Valerie Belton, University of Strathclyde.
v


vi

Preface

Fig. 1 The participants of the 19th International Conference on Multiple Criteria DecisionMaking

A total of 58 papers were submitted for publication in this book, 32 of which
have been accepted following a thorough peer review process. Eight of the accepted
papers were included in a special track on evolutionary multi-objective optimization organized by Boris Naujoks. These papers by Srivastava et al., Woehrle et al.,
Mikhailov and Knowles, Klinkenberg et al., Bader et al., Hernandez-Diaz et al.,
Preuss et al. and Tantar et al. were submitted and peer reviewed ahead of the
conference. This volume organized in four parts:
1. Multiple Criteria Decision-Making, Transportation, Energy Systems, and the
Environment
2. Applications of Multiple Criteria Decision-Making in Other Areas
3. Theory and Methodology of Multiple Criteria Decision-Making
4. Multiple Objective Optimization.
Part I contains ten papers applying MCDM methods to problems in energy and
transportation systems and environmental contexts. The applications range from city
electric transport to natural resource management, railway transport, and environmental synergies in supply chain integration. An even wider variety of applications
is covered in the ten papers in Part II. Many different MCDM methods are applied
in risk assessment, banking, manpower planning, wirelesses sensor networks, and
others. Parts III and IV have a theoretical and methodological focus. The five papers

in part III address the analytic hierarchy process, a bibliometric analysis of MCDM
and multiattribute utility theory, conjoint measurement, model predictive control,


Preface

vii

and classification. Part IV includes seven papers on multiple objective optimization.
These papers present a variety of algorithms for discrete and continuous multiobjective optimization problems, including five of the eight papers presented in the
special track on evolutionary multiple objective optimization of the conference.
Acknowledgements As editors, we wish to thank all the people who made the conference and
this book possible. First of all, our thanks go to the local organizing committee of Matthias Ehrgott
(chair), Ivan Kojadinovic, Richard Lusby, Michael OSullivan, Andrea Raith, Paul Rouse, Lizhen
Shao, Cameron Walker, Judith Wang, Hamish Waterer, and Oliver Weide. Secondly, we acknowledge the contributions of the Executive Committee of the International Society on Multiple Criteria
Decision-Making.
The book, of course depends on the hard work of the authors who have submitted papers and
the referees whose dedication in reviewing papers ensure the quality of this book. We wish to thank
the following individuals who acted as referees:
Lauren Basson, Nicola Beume, Bogusław Bieda, Antonio Boggia, Henri Bonnel, Claude
Bouvy, Dimo Brockhoff, G¨ulc¸in B¨uy¨uk¨ozkan, Herminia I. Calvete, Metin Celik, Eng Choo, Carlos
A. Coello Coello, Kalyanmoy Deb, Xavier Delorme, Hepu Deng, Liz Dooley, Ian Noel Durbach,
Matthias Ehrgott, Michael T.M. Emmerich, Jos´e Luis Esteves dos Santos, L. Paul Fatti, Carlos
M. Fonseca, Eugenia Furems, Lucie Galand, Xavier Gandibleux, Martin Josef Geiger, Evangelos
Grigoroudis, Evan J. Hughes, Masahiro Inuiguchi, Alessio Ishizaka, Rafikul Islam, Yaochu Jin,
Dylan F. Jones, Julien Jorge, Alison Joubert, Birsen Karpak, Joshua D. Knowles, Ivan Kojadinovic,
Murat K¨oksalan, Juha Koski, Elizabeth Lai, Riikka Leskel¨a, Anatoly Levchenkov, Chieh-Yow
ChiangLin, Richard Lusby, Oswald Marinoni, Benedetto Matarazzo, J¨orn Mehnen, Kristo Mela,
Gilberto Montibeller, Jos´e Mar´ıa Moreno-Jim´enez, Sanaz Mostaghim, Anna Nagurney, Boris
Naujoks, Shigeru Obayashi, Tatsuya Okabe, Lu´ıs Paquete, Long Pham, Carlo Poloni, Mike Preuß,

Domenico Quagliarella, Andrea Raith, Piet Rietveld, G¨unter Rudolph, Thomas L. Saaty, Ahti
Salo, Ramiro Sanchez-Lopez, Robert Scheffermann, Thomas Schlechte, Anita Sch¨obel, Yong
Shi, Theodor J. Stewart, Christian Stummer, Jacques Teghem, Jeffrey Teich, J´ozsef Temesi,
Heike Trautmann, Luis G. Vargas, Bego˜na Vitoriano, Raimo Voutilainen, Tobias Wagner, Jyrki
Wallenius, William C. Wedley, Heinz Roland Weistroffer, John F. Wellinton, Fred Wenstop, Lyndon While, Marino Widmer, Diederik Wijnmalen, Jan-Bo Yang, Ming-Miin Yu, Yeboon Yun,
Mahdi Zarghami, Wim Zeiler, Eckart Zitzler, Constantin Zopounidis.

Auckland
Dortmund
Cape Town
Helsinki
June 2009

Matthias Ehrgott
Boris Naujoks
Theodor J. Stewart
Jyrki Wallenius


Contents

Part I Multiple Criteria Decision Making, Transportation,
Energy Systems, and the Environment
On the Potential of Multi-objective Optimization in the Design
of Sustainable Energy Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .
Claude Bouvy, Christoph Kausch, Mike Preuss, and Frank Henrich

3

Evaluation of the Significant Renewable Energy Resources

in India Using Analytical Hierarchy Process.. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 13
Joseph Daniel, Nandigana V. R. Vishal, Bensely Albert,
and Iniyan Selvarsan
Multiple Criteria Decision Support for Heating Systems
in Electric Transport . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 27
Ivars Beinarts and Anatoly Levchenkov
Multi Criteria Decision Support for Conceptual Integral
Design of Flex(eble)(en)ergy Infrastructure.. . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 35
Wim Zeiler, Perica Savanovic, Rinus van Houten, and Gert Boxem
A Multi Criteria Knapsack Solution to Optimise Natural
Resource Management Project Selection .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 47
Oswald Marinoni, Andrew Higgins, and Stefan Hajkowicz
Environmental and Cost Synergy in Supply Chain Network
Integration in Mergers and Acquisitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 57
Anna Nagurney and Trisha Woolley
The Analytic Hierarchy Process in the Transportation Sector .. . . .. . . . . . . . . . . 79
Rafikul Islam and Thomas L. Saaty
RECIFE: A MCDSS for Railway Capacity Evaluation.. . . . . . . . . . . .. . . . . . . . . . . 93
Xavier Gandibleux, Pierre Riteau, and Xavier Delorme

ix


x

Contents

Balancing Efficiency and Robustness – A Bi-criteria
Optimization Approach to Railway Track Allocation . . . . . . . . . . . . . .. . . . . . . . . . .105
Thomas Schlechte and Ralf Bornd¨orfer

Tolling Analysis with Bi-objective Traffic Assignment . . . . . . . . . . . . . .. . . . . . . . . . .117
Judith Y.T. Wang, Andrea Raith, and Matthias Ehrgott
Part II Applications of Multiple Criteria Decison Making
in Other Areas
National Risk Assessment in The Netherlands. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .133
Erik Pruyt and Diederik Wijnmalen
Evaluation of Green Suppliers Considering
Decision Criteria Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .145
Orhan Feyzio˜glu and G¨ulc¸in B¨uy¨uk¨ozkan
A Multiobjective Bilevel Program for Production-Distribution
Planning in a Supply Chain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .155
Herminia I. Calvete and Carmen Gal´e
An Ordinal Regression Method for Multicriteria Analysis
of Customer Satisfaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .167
Isabel M. Jo˜ao, Carlos A. Bana e Costa, and Jos´e Rui Figueira
Discrete Time-Cost Tradeoff with a Novel Hybrid
Meta-Heuristic.. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .177
Kamal Srivastava, Sanjay Srivastava, Bhupendra K. Pathak,
and Kalyanmoy Deb
Goal Programming Models and DSS for Manpower Planning
of Airport Baggage Service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .189
Sydney C.K. Chu, Minyue Zhu, and Liang Zhu
A MCDM Tool to Evaluate Government Websites
in a Fuzzy Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .201
G¨ulc¸in B¨uy¨uk¨ozkan
Investigating Coverage and Connectivity Trade-offs
in Wireless Sensor Networks: The Benefits of MOEAs . . . . . . . . . . . . .. . . . . . . . . . .211
Matthias Woehrle, Dimo Brockhoff, Tim Hohm, and Stefan Bleuler
AHP as an Early Warning System:
An Application in Commercial Banks in Turkey . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .223

Ilyas Akhisar and Birsen Karpak


Contents

xi

A Multi-Criteria Evaluation of Factors Affecting Internet
Banking in Turkey . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .235
Sezi Cevik Onar, Emel Aktas, and Y. Ilker Topcu
Part III Theory and Methodology of Multiple Criteria
Decision Making
Priority Elicitation in the AHP by a Pareto Envelope-Based
Selection Algorithm .. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .249
Ludmil Mikhailov and Joshua Knowles
Bibliometric Analysis of Multiple Criteria Decision
Making/Multiattribute Utility Theory .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .259
Johanna Bragge, Pekka Korhonen, Hannele Wallenius,
and Jyrki Wallenius
Ordinal Qualitative Scales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .269
Salvatore Greco, Benedetto Matarazzo, and Roman Słowi´nski
Multi-objective Model Predictive Control .. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .277
Hirotaka Nakayama, Yeboon Yun, and Masakazu Shirakawa
Multiple Criteria Nonlinear Programming Classification with
the Non-additive Measure .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .289
Nian Yan, Yong Shi, and Zhengxin Chen
Part IV

Multiple Objective Optimization

A Reduced-Cost SMS-EMOA Using Kriging, Self-Adaptation,
and Parallelization . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .301
Jan-Willem Klinkenberg, Michael T. M. Emmerich, Andr´e H.
Deutz, Ofer M. Shir, and Thomas B¨ack
Faster Hypervolume-Based Search Using Monte Carlo
Sampling . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .313
Johannes Bader, Kalyanmoy Deb, and Eckart Zitzler
Using a Gradient Based Method to Seed an EMO Algorithm . . . . .. . . . . . . . . . .327
Alfredo G. Hernandez-Diaz, Carlos A. Coello, Fatima Perez,
Rafael Caballero, and Julian Molina
Nadir Point Estimation Using Evolutionary Approaches:
Better Accuracy and Computational Speed Through
Focused Search . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .339
Kalyanmoy Deb and Kaisa Miettinen


xii

Contents

A Branch and Bound Algorithm for Choquet Optimization
in Multicriteria Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .355
Lucie Galand, Patrice Perny, and Olivier Spanjaard
Decision Space Diversity Can Be Essential for Solving
Multiobjective Real-World Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .367
Mike Preuss, Christoph Kausch, Claude Bouvy, and Frank Henrich
Computing and Selecting "-Efficient Solutions
of f0,1g-Knapsack Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .379
Emilia Tantar, Oliver Sch¨utze, Jos´e Rui Figueira, Carlos A. Coello
Coello, and El-Ghazali Talbi

Contributors

Ilyas Akhisar School of Banking and Insurance, Marmara University, Istanbul,
Turkey,
Emel Aktas Istanbul Technical University, Management Faculty, Macka 34367,
Istanbul, Turkey,
Bensely Albert Department of Mechanical Engineering, College of Engineering,
Guindy, Anna University, Chennai 600025, India,
Johannes Bader Computer Engineering and Networks Lab, ETH Zurich, 8092
Zurich, Switzerland,
Thomas B¨ack Leiden Institute for Advanced Computer Science (LIACS), Leiden
University, Niels Bohrweg 1, 2333-CA Leiden, The Netherlands,
Carlos A. Bana e Costa Centre for Management Studies of Instituto Superior
T´ecnico, Technical University of Lisbon, Av. Rovisco Pais, 1049-001 Lisbon,
Portugal,
Ivars Beinarts Riga Technical University, Kronvalda blvd. 1-202, Riga, Latvia,

Stefan Bleuler Computer Engineering and Networks Lab, ETH Zurich, 8092
Zurich, Switzerland,
Ralf Bornd¨orfer Konrad-Zuse-Zentrum f¨ur Informationstechnik Berlin (ZIB),
Takustr 7, Berlin-Dahlem 14195, Germany,
Claude Bouvy Forschungsgesellschaft Kraftfahrwesen mbH Aachen,
Steinbachstraß e7, 52074 Aachen, Germany,
Gert Boxem Faculty of Architecture, Building and Planning, Technische
Universiteit Eindhoven, The Netherlands,
Johanna Bragge Helsinki School of Economics, Department of Business
Technology, P.O. Box 1210, Helsinki 00101, Finland,
Dimo Brockhoff Computer Engineering and Networks Lab, ETH Zurich, 8092

Zurich, Switzerland,

xiii


xiv

Contributors

Gulc
¨ ¸ in Buy
¨ uk¨
¨ ozkan Department of Industrial Engineering, Galatasaray
University, C¸ıra˘gan Caddesi No. 36 Ortak¨oy, ˙Istanbul, Turkey,

Rafael Caballero Department of Applied Economics (Mathematics), University
of Malaga, Malaga, Spain, r
Herminia I. Calvete Dpto. de M´etodos Estad´ısticos, IUMA, Universidad
de Zaragoza, Pedro Cerbuna 12, Zaragoza 50009, Spain,
Sezi Cevik Onar Istanbul Technical University, Management Faculty, Macka,
Istanbul 34367, Turkey,
Zhengxin Chen College of Information Science and Technology, University
of Nebraska, Omaha, NE 68182, USA,
Sydney C.K. Chu Department of Mathematics, University of Hong Kong,
Pokfulam Road, Hong Kong, China,
Carlos A. Coello Coello Centro de Investigacion y de Estudios Avanzados,
CINVESTAVIPN, Department of Computer Science, M´exico D.F., Mexico,

Joseph Daniel Department of Mechanical Engineering, College of Engineering,
Anna University, Guindy, Chennai 600025, India,

Kalyanmoy Deb Department of Mechanical Engineering, Indian Institute
of Technology, Kanpur 208016, India,
Xavier Delorme Centre G´enie Industriel et Informatique, Ecole des Mines
de Saint-Etienne, 158 cours Fauriel, F-42023 Saint-Etienne cedex 2, France,

Andr´e H. Deutz Leiden Institute for Advanced Computer Science (LIACS),
Leiden University, Niels Bohrweg 1, Leiden 2333-CA, The Netherlands,

Matthias Ehrgott Department of Engineering Science, The University of
Auckland, Private Bag 92019, Auckland 1142, New Zealand, m.ehrgott@auckland.
ac.nz
Michael T. M. Emmerich Leiden Institute for Advanced Computer Science
(LIACS), Leiden University, Niels Bohrweg 1, Leiden 2333-CA, The Netherlands,

Orhan Feyzio˘glu Department of Industrial Engineering, Galatasaray University,
C¸ıra˘gan Caddesi No: 36 Ortak¨oy, ˙Istanbul, Turkey,
Jos´e Rui Figueira Centre for Management Studies of Instituto Superior T´ecnico,
Technical University of Lisbon, Tagus Park, Av. Cavaco Silva, Porto Salvo, Lisbon
2780-990, Portugal,
Lucie Galand LIP6-UPMC, 104 av. du Pr´esident Kennedy, Paris 75016, France,



Contributors

xv

Carmen Gal´e Dpto. de M´etodos Estad´ısticos, IUMA, Universidad de Zaragoza,
Mar´ıa de Luna 3, Zaragoza 50018, Spain,
Xavier Gandibleux Laboratoire d’Informatique de Nantes Atlantique UMR

CNRS, 6241, Universit´e de Nantes, 2, rue de la Houssini`ere BP 92208, F-44322
Nantes cedex 03, France,
Salvatore Greco Faculty of Economics, University of Catania, Corso Italia, 55,
Catania 95129, Italy,
Stefan Hajkowicz CSIRO Sustainable Ecosystems, St Lucia Qld 4067, Australia,

Frank Henrich Siemens AG, Energy Sector, Wolfgang-Reuter-Platz 4, Duisburg
47053, Germany,
Alfredo G. Hernandez-Diaz Department of Economics, Quantitative Methods
and Economic History, Pablo de Olavide University, Seville, Spain,
Andrew Higgins CSIRO Sustainable Ecosystems, St Lucia Qld 4067, Australia,

Tim Hohm Computer Engineering and Networks Lab, ETH Zurich, Zurich 8092,
Switzerland,
Rafikul Islam Department of Business Administration, International Islamic
University, Malaysia, P.O. Box 10, Kuala Lumpur 50728, Malaysia,

Isabel M. Jo˜ao Department of Chemical Engineering, Instituto Superior
de Engenharia de Lisboa, Polytechnic Institute of Lisbon, Rua Conselheiro Em´ıdio
Navarro, Lisbon 1957-007, Portugal,
Birsen Karpak Management Department, Youngstown State University, USA,

Christoph Kausch Chair of Technical Thermodynamics, RWTH Aachen
University, Schinkelstr 8, Aachen 52062, Germany,
Jan-Willem Klinkenberg Leiden Institute for Advanced Computer Science
(LIACS), Leiden University, Niels Bohrweg 1, Leiden 2333-CA, The Netherlands,

Joshua Knowles School of Computer Science, University of Manchester, Oxford
Road, Kilburn building, Manchester M13 9PL, UK,
Pekka Korhonen Helsinki School of Economics, Department of Business

Technology, P.O. Box 1210, Helsinki 00101, Finland,
Anatoly Levchenkov Riga Technical University, Kronvalda blvd., Riga 1-202,
Latvia,
Oswald Marinoni CSIRO Sustainable Ecosystems, St Lucia Qld 4067, Australia,



xvi

Contributors

Benedetto Matarazzo Faculty of Economics, University of Catania, Corso Italia,
55, Catania 95129, Italy,
Kaisa Miettinen Department of Mathematical Information Technology, P.O. Box
35 (Agora), University of Jyv¨askyl¨a, FI-40014, Finland,
Ludmil Mikhailov Manchester Business School, University of Manchester, Booth
Street East, Manchester M15 6PB, UK,
Julian Molina Department of Applied Economics (Mathematics), University
of Malaga, Malaga, Spain,
Anna Nagurney Department of Finance and Operations Management, Isenberg
School of Management, University of Massachusetts Amherst, Massachusetts
01003, USA,
Hirotaka Nakayama Konan University, 8-9-1 Okamoto, Higashinada, Kobe
658-8501, Japan,
Bhupendra K. Pathak Department of Mathematics, Dayalbagh Educational
Institute, Dayalbagh, Agra 282110, India,
Fatima Perez Department of Applied Economics (Mathematics), University
of Malaga, Malaga, Spain,f
Patrice Perny LIP6-UPMC, 104 av. du Pr´esident Kennedy, Paris 75016, France,

Mike Preuss Chair of Algorithm Engineering, TU Dortmund University,
Otto-Hahn-Str. 14, Dortmund 44227, Germany,
Erik Pruyt Faculty of Technology, Policy and Management, Delft University of
Technology, P.O. Box 5015, GA Delft 2600, The Netherlands,
Andrea Raith Department of Engineering Science, The University of Auckland,
Private Bag 92019, Auckland 1142, New Zealand,
Pierre Riteau Laboratoire d’Informatique de Nantes Atlantique UMR CNRS
6241, Universit´e de Nantes, 2, rue de la Houssini´ere BP 92208, F-44322 Nantes
cedex 03, France,
Thomas L. Saaty Joseph Katz Graduate School of Business, University
of Pittsburgh, 322 Mervis Hall, Pittsburgh, PA 15260, USA,
Perica Savanovic Faculty of Architecture, Building and Planning, Technische
Universiteit Eindhoven, The Netherlands,
Thomas Schlechte Konrad-Zuse-Zentrum f¨ur Informationstechnik Berlin (ZIB),
Takustr. 7, Berlin-Dahlem 14195, Germany,
Oliver Schutze
¨
CINVESTAV-IPN, Computer Science Department, M´exico D.F.
07360, Mexico,
Iniyan Selvarsan Department of Mechanical Engineering, College of Engineering,
Guindy, Anna University, Chennai 600025, India,


Contributors

xvii

Yong Shi College of Information Science and Technology, University of Nebraska,
Omaha, NE 68118, USA,
and

Chinese Academy of Sciences Research Center on Fictitious Economy And Data
Science, Graduate University of Chinese Academy of Sciences, Beijing 100080,
China,
Masakazu Shirakawa Toshiba Corporation, 2-4 Suehirocho, Tsurumi, Yokohama
230-0045, Japan,
Roman Słow´ınski Institute of Computing Science, Pozna´n University of
Technology, 60-965 Poznan, and Systems Research Institute, Polish Academy
of Sciences, Warsaw 01-447, Poland,
Olivier Spanjaard LIP6-UPMC, 104 av. du Pr´esident Kennedy, Paris 75016,
France,
Kamal Srivastava Department of Mathematics, Dayalbagh Educational Institute,
Dayalbagh, Agra 282110, India,
Sanjay Srivastava Department of Mechanical Engineering, Dayalbagh
Educational Institute, Dayalbagh, Agra 282110, India,
El-Ghazaali Talbi INRIA Lille-Nord Europe, LIFL (UMR USTL/CNRS 8022),
Parc Scientifique de la Haute Borne 40, avenue Halley Bˆat.A, Park Plaza,
Villeneuve d’Ascq C´edex 59650, France,
Emilia Tantar INRIA Lille-Nord Europe, LIFL (UMR USTL/CNRS 8022), Parc
Scientifique de la Haute Borne 40, avenue Halley Bˆat.A, Park Plaza, Villeneuve
d’Ascq C´edex 59650, France,
Y. Ilker Topcu Istanbul Technical University, Management Faculty, Macka,
Istanbul 34367, Turkey,
Rinus van Houten Faculty of Architecture, Building and Planning, Technische
Universiteit Eindhoven, The Netherlands,
Nandigana V.R. Vishal Department of Mechanical Engineering, College
of Engineering, Guindy, Anna University, Chennai 600025, India
Hannele Wallenius Helsinki University of Technology, Department of Industrial
Engineering and Management, P.O. Box 5500, TKK 02015, Finland, hannele.

Jyrki Wallenius Helsinki School of Economics, Department of Business

Technology, P.O. Box 1210, Helsinki 00101, Finland,
Judith Y.T. Wang The Energy Centre, The University of Auckland Business
School, Private Bag 92019, Auckland 1142, New Zealand,
Diederik Wijnmalen Strategic Choices Department, TNO Organisation
for Applied Research, P.O. Box 96864, 2509 JG, The Hague, The Netherlands,



xviii

Contributors

Matthias Woehrle Computer Engineering and Networks Lab, ETH Zurich,
Zurich 8092, Switzerland,
Trisha Woolley Department of Finance and Operations Management, Isenberg
School of Management, University of Massachusetts Amherst, Massachusetts
01003, USA,
Nian Yan College of Information Science and Technology, University of Nebraska,
Omaha, NE 68182, USA,
Yeboon Yun Kagawa University, 2217-20 Hayashicho, Takamatsu 761-0396,
Japan,
Wim Zeiler Faculty of Architecture, Building and Planning, Technische
Universiteit, Eindhoven, The Netherlands,
Liang Zhu Department of Mathematics, Fudan University, Shanghai, China,
godloveme
Minyue Zhu Department of Mathematics, University of Hong Kong, Pokfulam,
Road, Hong Kong, China,
Eckart Zitzler Computer Engineering and Networks Lab, ETH Zurich, 8092
Zurich, Switzerland,

Part I

Multiple Criteria Decision Making,
Transportation, Energy Systems,
and the Environment


On the Potential of Multi-objective
Optimization in the Design of Sustainable
Energy Systems
Claude Bouvy, Christoph Kausch, Mike Preuss, and Frank Henrich

Abstract A new multi-criterial methodology is introduced for the combined structural and operational optimization of energy supply systems and production processes. The methodology combines a multi-criterial evolutionary optimizer for
structural optimization with a code for the operational optimization and simulation. The genotype of the individuals is interpreted with a superstructure. The
methodology is applied to three real world case studies: one communal and one
industrial energy supply system, one distillation plant. The resulting Pareto fronts
and potentials for cost reduction and ecological savings are discussed.
Keywords Communal energy supply concepts Distillation plants Evolutionary
algorithms Industrial energy supply systems Multi-objective optimization

1 Introduction
Due to the finite resources of fossil fuels, their increasing costs and the rising awareness concerning the impact of CO2 emissions on the climate, the design of highly
efficient energy supply systems and manufacturing processes is essential for a sustainable energy supply in the future. Like for most engineering tasks several decision
making criteria (i.e., objectives), mostly contradictory, are relevant for this task. In
the design phase of energy systems economic factors (e.g., investment sum, overall
yearly costs) are opposed to ecological (e.g., yearly CO2 emissions, yearly primary
energy consumption) and reliability (e.g., availability of a given technology, supply security) criteria. The sustainability of an energy supply system will be given
with minimal ecological impact and maximal availability (as no back-up system
based on fossil fuels will be needed). However the economic range will in general

C. Bouvy (B)
Forschungsgesellschaft Kraftfahrwesen mbH Aachen, Steinbachstraß e7,
52074 Aachen, Germany,
e-mail:
M. Ehrgott et al., Multiple Criteria Decision Making for Sustainable Energy and
Transportation Systems, Lecture Notes in Economics and Mathematical Systems 634,
DOI 10.1007/978-3-642-04045-0 1,
c Springer Physica-Verlag Berlin Heidelberg 2010

3


4

C. Bouvy et al.

be restricted for the realization of such systems. Thus an optimal configuration with
respect to all criteria needs to be found.
The increasing number of available energy conversion units (e.g., micro-turbines,
thermal heat pumps, mechanical heat pumps) makes the design phase of energy
supply systems even more difficult. Furthermore the behavior of modern energy
conversion units is more and more complex (e.g., temperature sensitivities of heat
pumps). It is clear that for such a complex task computer based tools to support the
planning engineer are desirable and will be of increased practical relevance in the
near future.
The design task formulated above is an overall (i.e., operational and structural)
optimization problem. Such optimization tools were developed at the Chair of Technical Thermodynamics of RWTH Aachen University (cf. Bouvy 2007 and Bouvy
and Lucas 2007) for energy supply systems and in co-operation with the Chair of
Algorithm Engineering of the University of Dortmund for distillation plant layout
(cf. Henrich et al. 2008 and Preuss et al. 2008).

2 Methodology
It is clear that for highly multi-objective optimization tasks, as the overall optimization of energy supply systems and manufacturing processes, an a priori decision
making concept (e.g., a priori weighting of decision criteria) is not convenient,
because it will not take into account the complex topology of the solution space.
Thus in this work a n-dimensional Pareto concept is used to support the planning
engineer in the design phase. The multi-objective optimization methodology presented in this work combines a multi-objective structural optimization tool, based
on evolutionary strategy with an operational optimizer and simulator. Evolutionary
strategies (specific methodology of evolutionary algorithms) are bionic, probabilistic optimization method belonging to the category of metaheuristics (cf. e.g.,
Schwefel 1995 and Eiben and Smith 2003). Evolutionary algorithms perform a
direct search, i.e., no gradient information is required. The chosen evolutionary
optimizers are a modified . C /-evolutionary strategy for the two first examples in Sect. 3 (cf. Bouvy 2007 and Bouvy and Lucas 2007) and the SMS-EMOA
(cf. Emmerich et al. 2005, Henrich et al. 2008, and Preuss et al. 2008) for the third
example. The used methodology is outlined in Fig. 1.
As evolutionary algorithms in general and evolutionary strategies in particular
need an initialization at least one individual (i.e., a precise energy supply system
or distillation plant) is manually entered (“Initial individual(s)” in Fig. 1). The evolutionary optimizer then generates an initial population by method of a stochastic
algorithm, which is designed to get a good distribution of the initial population over
the solution space.
The generated solutions and later on the newly generated individuals (by the evolutionary optimizer) are a vector of real, integer and listed values (i.e., the genotype).
In order to interpret this set of values their interactions have to be defined, which is


EMO in Sustainable Energy Systems Design

5

Fig. 1 Scheme of the used
methodology

realized in this work by means of a superstructure. A superstructure is a structure
which includes all (or at least a large number of) reasonable possibilities of combinations of the considered units (e.g., co-generation units, district heating pipe) and
has thus to be adapted to every optimization task.
In order to compare the fitness of the different solutions generated by the evolutionary optimizer, all n decision making criteria are computed by a simulator. For
the design of energy systems the operational optimizer and simulator “eSim” of the
toolbox “TOP-Energy” (cf. Augenstein et al. 2004 and Augenstein and Herbergs
2005) was used, whereas ASPEN PLUSTM was used for the layout of distillation
plants.
Based on the determined fitness values the evolutionary optimizer applies the
operators “mutation” and “recombination” (cf. e.g., Schwefel 1995) to generate a
new set of solutions, which will again be interpreted and computed by the simulator.
At each run of the closed cycle shown in Fig. 1 only the fittest individuals survive.
Similar to the evolutionary process in nature, this methodology will result in the
improvement of the living population.
If a predefined termination criterion is reached, the optimization runs result
in n-dimensional Pareto sets. When comparing any individual to an individual i
according to the Pareto criterion all individuals with all fitness values larger than
those of i (for minimization) are inferior to i . In the same way all individuals with
all fitness values smaller than those of i are clearly better than i . For all other individuals no statement can be made because some criteria are better and other ones
worse. On the one hand these Pareto sets will identify promising structural alternatives, reducing for example the CO2 emissions considerably with simultaneous low
costs. On the other hand the ecological criteria are a good indicator of the stability of
the chosen structural solutions towards rising energy costs. Thus this methodology
yields actual potentials for reducing ecological impacts as well as information about
the stability towards changing energy supply costs whilst fulfilling all boundary
conditions.


6

C. Bouvy et al.

3 Real-world Applications
In this section the application of the methodology described above to three different
“real world” case studies is discussed.

3.1 Communal Energy System
The communal area considered in this section consists of seven residential districts,
which represent the consumers. The task is to determine optimal structural solutions
for a communal energy supplier. The goal of this optimization problem is to reduce
overall costs of the system as well as the ecological impact. Thus the overall yearly
costs (i.e., sum of the yearly capital costs and the yearly operational costs) and the
yearly primary energy consumption were chosen as decision making criteria.
On the demand side only electricity and space heating are taken into account. The
electrical peak load for this real world problem is 100 MW whereas a thermal peak
load of 176:6 MW was calculated. It is important to note that the time dependencies
of the heat and the electricity demand are considered for the optimization runs.
A detailed description of the demands can be found in Bouvy and Lucas (2007)
or Bouvy (2007). For cost calculation all relevant shares were considered and current
supply costs for electricity (6 ct=kWh) and natural gas (2 ct=kWh) were assumed
(costs for a regional supplier).
A standard supply system for the considered districts would consist of an electricity grid, fed by distant condensation power plants, and a natural gas rail for space
heat production by boilers in the various buildings. This solution was chosen as start
individual for the optimization run (cf. Fig. 2).
The structural margin for the optimization, coded in the superstructure, included
heat production on the house level with boilers or heat pumps, co-generation on a
centralized level for each district (supported by peak-load boilers), heat distribution
by several possible district heating networks and electricity generation at a powerplant level (condensation and co-generation plants). A detailed description is found
in Bouvy and Lucas (2007) and Bouvy (2007). Due to the high number of possible structural alternatives (nominal powers and crosslinking) the resulting solution
space is highly complex.
The progress of the optimization and the resulting pareto fronts at different

generations are given in Fig. 2.
As the evaluation in “eSim” of one solution (i.e., one individual) takes about
1 min (8640 h=a) the optimization time for this task was very high (about 720 h
on a Pentium 4 desktop computer with a 3:2 GHz CPU for the results shown in
Fig. 2) and had to be prematurely stopped. Thus the evolutionary optimizer did not
yet reach the vicinity of the global optimum. The large computing times are a well
known phenomenon of the chosen self adapting evolutionary strategies. The main
reason for this were superfluous components (e.g., district heating networks that
were not used during the operation over the considered year) causing higher yearly


Yearly Primary Energy Demand in GWh/a

EMO in Sustainable Energy Systems Design

7

Initial individual

Yearly costs in 106 €/a

Fig. 2 Results and progress of the optimization run for the communal energy system

costs. As the optimization runs were stopped early these costs had to be corrected
manually. Figure 2 shows the corrected individuals marked as red dots ( D 500 ).
These corrected individuals show, that both primary energy and yearly costs can be
saved compared to a non-integrated energy supply system based on heat supply with
boilers and electricity supply with a condensation combined gas and steam power
plant. This savings are mainly based on the extensive use of electricity driven heat
pumps on a decentralized level due to the reasonable average power to heat ratio

of the heat pumps ("
5), resulting from the assumed low temperature heating
systems (supply temperatures of 55ı C). For the considered boundary conditions a
reduction in primary energy demand of approximatively 25% compared to a nonintegrated energy supply system can be reached with a simultaneous reduction of
yearly overall costs.

3.2 Industrial Energy System
The second case study considered in this work is the optimization of an industrial energy system. Contrary to the communal application, presented above, the
energy demand consists of electricity (peak load 720 kW), low (peak load 1260 kW)
and high temperature heat (steam, peak load 910 kW). The superstructure for this
case study considers steam production with high temperature co-generation units
(micro-turbines) or steam boilers. Low temperature heat can be supplied by either
boilers, mechanical heat pumps, motoric co-generation units or by heat exchange
from the steam grid. Besides the considered co-generation units electrical energy


C. Bouvy et al.
Yearly Primary Energy Demand in MWh/a

8

Yearly costs / €/a

Fig. 3 Results of the optimization run for the industrial energy system

can be bought from a supplier. Again all time dependencies over a year were considered for all three energies. A detailed description of the demand profiles and
their interactions is given in Bouvy (2007). As decision making criteria again the
total yearly costs and the yearly primary energy demand were chosen. For cost calculation current supply costs for electricity (12 ct=kWh) and natural gas (5 ct=kWh)
were assumed (costs for an industrial customer).
As start individual again a non-integrated energy supply system, based on external electricity supply, and heat production with a low temperature and a steam

boiler.
The results of the optimization run for this industrial energy system are shown in
Fig. 3.
As only three demand profiles (i.e., consumers) had to be covered, the complexity
and thus the solution space of this optimization task is smaller than for the communal energy system. Furthermore the evaluation in “eSim” of one precise solution is
about 30 times quicker. Better results are thus reached in less calculation time. The
results shown in Fig. 3 were reached after 120 h on a Pentium 4 desktop computer
with a 3:2 GHz CPU. They should be situated very near to the global optimum as
nearly no superfluous units were found in the individuals of the Pareto front shown
in Fig. 3.
For this optimization run numerous solutions clearly dominating (i.e., with both
lower costs and lower primary energy demand) the start individual were found,
thus revealing an important potential for highly integrated energy systems. This
is mainly due to the higher energy supply costs compared to the first case study
presented. The start individual was not found in the final Pareto front. It should be
mentioned that all Pareto optimal solutions had at least one 80 kWel micro-turbine
for steam production. This technology is very stable for the considered case study as


EMO in Sustainable Energy Systems Design

9

the recuperator bypass allows an adaption of the exhaust gas temperature and thus
an optimal covering of the demand profile.
Figure 3 reveals also another interesting fact. Two regions for saving primary
energy can be identified. In the first region (from 11750 to 1260 MWh=a) costs for
saving primary energy can be estimated to about 3:75 ct=kWh (i.e., the slope of the
blue line). It is to mention that these saving costs are even less than the supply costs
for natural gas. If a further saving in primary energy is intended the costs raise to

19:63 ct=kWh (i.e., the slope of the red line). This supports the idea of real multi
criterial decision making (i.e., not by a priori weighting) as important potentials can
only be revealed when the knowledge of the solution space is included.

3.3 Distillation Plant
Another problem which has been investigated is the layout and operation of a general distillation sequence for the separation of multi-component feed streams into
multi-component products using non-sharp splits. Different objective functions are
analyzed including economic criteria like the total annual cost, investment cost
and the return on investment as well as ecological criteria like the exergy loss
(cf. Henrich et al. 2008 and Preuss et al. 2008). The structural alternatives included
in the superstructure are stream bypassing and blending, the number and sequence
of splits as shown in Fig. 4.
Together with the operational variables of the distillation columns this results in
a highly complex decision space involving non-convexities and multi-modalities.
Each proposed structure and operation is modeled with ASPEN PLUSTM to ensure
that all boundary conditions are met and the solution is thermodynamically sound.
The exergy loss and investment cost Pareto front for the case described in detail
in Preuss et al. (2008) is shown in Fig. 5. In practice generally the solutions of interest would be those forming “corners” in the front (e.g., the points at 720 tUSD
and 340 kW, at 790 tUSD and 325 kW, at 890 tUSD and 315 kW), whereas solutions promising a small gain in the one objective at the cost of a large loss in the
other objective would not normally be of interest (e.g., 1060 tUSD and 312 kW).

A
A,B, (C)

F

Fig. 4 Structural alternatives
for separation of
3-component feed into multi
component products with

non-sharp splits

2

P4
(A), B

1

P5

(B), C
A,B,C


10

C. Bouvy et al.

Fig. 5 Exergy loss and investment cost Pareto fronts of the best 5 runs

Additionally there are 3 fronts which at first glance seem similar but the solutions
of the fronts stem from very different areas in the variable space thus if points are of
interest where the fronts overlap or are near identical, then an additional criteria can
be taken into account which is of course vastly beneficial to the design and planning
process.

4 Conclusions and Potential
The applications of the introduced methodology have shown that the introduced
concept is adequate for supporting the planning engineer and the decision maker

in the design phase of energy supply systems and complex production processes.
Very promising solutions were found for each real world case. Especially the Pareto
concept is very important for multi-criterial decision making as it takes into consideration the correlations of the different criteria. Thus real potentials for both energy
and ecological saving can be estimated and a decision can be taken for example with
the information of relative primary energy saving costs. Methodologies that optimize only by means of a single criterion or that perform an a priori fixed weighting
of the decision making criteria will in general be computed quicker but on the other
hand will only result in a single solution and not a set.
However the very high computing times for highly complex optimization tasks
show that the methodology has to be improved. Several possibilities were given in
Bouvy (2007). On the one hand the parallelization of computers will reduce the
computational time by distributing the evaluation of the different individuals. On