Energy 35 (2010) 2517e2527
Contents lists available at ScienceDirect
Energy
journal homepage: www.elsevier.com/locate/energy
Multicriteria renewable energy planning using an integrated fuzzy
VIKOR & AHP methodology: The case of Istanbul
Tolga Kaya a, *, Cengiz Kahraman b
a
b
Istanbul Technical University, Department of Management Engineering, 34367 Macka, Istanbul, Turkey
Istanbul Technical University, Department of Industrial Engineering, 34367 Macka, Istanbul, Turkey
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 October 2009
Received in revised form
25 February 2010
Accepted 26 February 2010
Available online 1 April 2010
The purpose of this study is twofold: first, it is aimed at determining the best renewable energy alternative for Istanbul by using an integrated VIKOR-AHP methodology. Second, a selection among alternative energy production sites in this city is made using the same approach. In the proposed VIKOR-AHP
methodology, the weights of the selection criteria are determined by pairwise comparison matrices of
AHP. In energy decision making problems, the judgments of decision makers are usually vague. As it is
relatively difficult for decision makers to provide exact values for the criteria, the evaluation data for the
alternative energy policies should be expressed in linguistic terms. In order to model this kind of
uncertainty in human preferences, fuzzy logic is applied very successfully. Thus, both classical VIKOR and
classical AHP procedures are performed under fuzzy environment. The originality of the paper comes
from the application of the proposed integrated VIKOR-AHP methodology to the selection of the best
energy policy and production site. It is found that wind energy is the most appropriate renewable energy
option and Çatalca district is the best area among the alternatives for establishing wind turbines in
Istanbul.
2010 Elsevier Ltd. All rights reserved.
Keywords:
Renewable energy
Fuzzy
Multi-criteria
VIKOR
AHP
Wind energy
1. Introduction
Energy planning endeavor involves finding a set of sources and
conversion devices so as to meet the energy requirements/demands
of all the tasks in an optimal manner [1]. Making an energy planning
decision involves a process of balancing diverse ecological, social,
technical, and economic aspects over space and time. This balance is
critical to the survival of nature and to the prosperity of energy
dependent nations.
Technical and environmental aspects are usually represented in
the form of multiple criteria and indicators that are often expressed
by conflicting objectives. Energy planning using multi-criteria
analysis has attracted the attention of decision makers for a long
time. During the 1970s, dealing with energy problems by single
criterion approaches which aimed at identifying the most efficient
supply options at a low cost was popular. However, in the 1980s,
growing environmental awareness modified the above decision
framework. The need to incorporate environmental and social
considerations in energy planning resulted in the increasing usage
of multi-criteria approaches [2e4].
* Corresponding author. Fax: þ90 212 2407260.
E-mail address: kayatolga@itu.edu.tr (T. Kaya).
0360-5442/$ e see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.energy.2010.02.051
As the complexity of decisions increases, it becomes more
difficult for decision makers to identify an alternative that maximizes all decision criteria. Operationally, energy assessments must
deal with attributes difficult to define and components that may
involve both quantitative and qualitative factors. In terms of scope,
an assessment may cover technical, economic or geographic areas
whose boundaries may not be easily identifiable, and socioeconomic regions that affect various interest groups or stakeholders
each with their own demands and socioeconomic needs. In view of
these difficulties, methods based on fuzzy logic may be quite useful
in undertaking difficult assessment procedures. The fuzzy set
theory was introduced by Zadeh [5] to express the linguistic terms
in decision-making process in order to resolve the vagueness,
ambiguity and subjectivity of human judgment.
VIKOR (VIsekriterijumsko KOmpromisno Rangiranje) is a multiattribute decision making technique which has a simple computation procedure that allows simultaneous consideration of the
closeness to ideal and anti-ideal alternatives. In the literature, there
are many studies which handle VIKOR method in a comparative
manner: Opricovic and Tzeng [6] conducted a comparative analysis
of VIKOR and TOPSIS (Technique for Order Preference by Similarity
to Ideal Solution) methods with a numerical example. Tzeng et al.
[7] also compared the two methodologies to determine the best
compromise solution among alternative fuel modes. Opricovic and
2518
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Table 1
List of evaluation criteria used in MCDM studies conducted on energy issues.
Table 3
Fuzzy evaluation scores for the weights.
Aspects
Criteria
Linguistic terms
Fuzzy score
Technical
Efficiencya
Exergy (rational) efficiencya
Primary energy ratio
Safety
Reliability
Maturity
Others
Economic
Investment costa
Operation and maintenance costa
Fuel cost
Electric cost
Net present value
Payback period
Service life
Equivalent annual cost
Others
Absolutely strong (AS)
Very strong (VS)
Fairly strong (FS)
Slightly strong (SS)
Equal (E)
Slightly weak (SW)
Fairly weak (FW)
Very weak (VW)
Absolutely weak (AW)
(2, 5/2, 3)
(3/2, 2, 5/2)
(1, 3/2, 2)
(1, 1, 3/2)
(1, 1, 1)
(2/3, 1, 1)
(1/2, 2/3, 1)
(2/5, ½, 2/3)
(1/3, 2/5, 1/2)
Environmental
NOx emissiona
CO2 emissiona
CO emission
SO2 emission
Particles emission
Non-methane volatile compounds
Land usea
Noise
Others
Social
Social acceptabilitya
Job creationa
Social benefits
Others
a
Most frequently used.
Tzeng [8] made a comparison of VIKOR with PROMETHEE (Preference Ranking Organization METHod for Enrichment Evaluations),
ELECTRE (ELimination and Choice Expressing REality) and TOPSIS
approaches. Chu et al. [9] provided a comparative analysis of SAW
(Simple Additive Weighting), TOPSIS and VIKOR, which demonstrated the similarities and differences of these methodologies in
achieving group decisions.
In fuzzy VIKOR, linguistic preferences can be converted to fuzzy
numbers [10]. For the determination of the relative importance of
selection criteria, fuzzy Analytic Hierarchy Process (AHP) can be
used since it is based on pairwise comparisons and allows the
utilization of linguistic variables. Although the pairwise comparison approach is the most demanding in terms of solicited input
from the experts, it offers maximum insight, particularly in terms of
assessing consistency of the experts’ judgments. In this context,
this technique is ideal for the closer examination of a selected set of
renewable energy planning and site selection criteria. The technique is also the most accurate when it comes to reflecting the
relative weight of each criterion and indicator [11].
Turkey, which is still extensively dependent on energy imports
from foreign countries, as it was in the past, can only meet about
one third of its energy demand by means of native generation. The
shares depending on foreign supply of oil and natural gas are even
much higher. As fossil fuel energy becomes scarcer, Turkey will
most probably face energy shortages, high energy prices, and
energy insecurity within the next few decades. Moreover, Turkey’s
reliance on fossil fuel consumption will accelerate global warming
and reduce the domestic environmental quality. For these reasons,
the development and usage of renewable energy sources and
technologies in Istanbul, Turkey’s most industrialized and energy
consuming region, are increasingly becoming vital for sustainable
economic development of the country [12,13].
In this study, a modified fuzzy VIKOR methodology is proposed
to make a multi-criteria selection among alternative renewable
energy options and production sites for Istanbul area. In the
proposed methodology, the decision makers’ opinions on the
relative importance of the selection criteria are determined by
a fuzzy AHP procedure.
The rest of the paper is organized as follows: In Section 2,
a literature review on multi-criteria energy decision making is
briefly given. In Section 3, a modified fuzzy VIKOR methodology is
presented. In Section 4, following the determination of the selection criteria and alternatives, the proposed methodology is applied
to a two step renewable energy planning problem for Istanbul. In
Section 5, concluding remarks are given.
2. Literature review
An energy planning process usually consists of a study of demand
and supply, forecasts of the trends of inputeoutput items, based on
economics and technological models, and a list of actions, collecting
several measures voted to fulfill the main objectives of the energy
plan [14]. One of the most common problems of energy planning is to
choose among various alternative energy sources and technologies
to be promoted. Technologies based on solar energy, wind energy,
hydraulic energy, biomass, animal manure, geothermal energy,
Table 2
Cavallaro and Ciraolo’s list of evaluation criteria for production site selection.
Aspects
Criteria
Technical
Energy production capacity
Technological maturity
Economic
Investment cost
Operation and maintenance cost
Fuel cost
Realization times
Environmental
Impact on ecosystems
CO2 emission
Visual impact
Noise
Social
Social acceptability
~ and M
~ .
Fig. 1. The intersection between M
1
2
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Table 4
Fuzzy evaluation scores for the alternatives.
Linguistic terms
Fuzzy score
Very poor (VP)
Poor (P)
Medium poor (MP)
Fair (F)
Medium good (MG)
Good (G)
Very good (VG)
(0,
(0,
(1,
(3,
(5,
(7,
(9,
0, 1)
1, 3)
3, 5)
5, 7)
7, 9)
9, 10)
10, 10)
energy saving in residential and industry sectors, wave energy are
among the most popular alternatives [15e18]. Despite environmental drawbacks, nuclear and conventional energy resources like
coal, oil, and natural gas may still be included in the list of alternative
technologies to be promoted [17,19].
There is a vast multi-criteria decision making literature on
energy issues. Keeney et al. [20] structured a hierarchical representation of criteria and then aggregated them into a combined
‘value tree’ in order to evaluate future energy systems of West
Germany. Hamalainen and Karjalainen [21] utilized AHP to determine the relative weights of the evaluation criteria of Finland’s
energy policies. Mirasgedis and Diakoulaki [22] compared the
external costs of power plants which used different energy sources
by a multi-criteria analysis. Mavrotas et al. [23] presented a multiple
objective linear programming model and applied it to the Greek
electricity generation sector. Taking energy resources, environment
capacity, social indicators, and economic indicators into account,
Afgan and Carvalho [24] defined energy indicators which are used in
the assessment of energy systems. Haralambopoulos and Polatidis
[25] used PROMETHEE II to achieve group consensus in renewable
energy projects and applied the decision framework to
a geothermal resource usage case in the island of Chios. Beccali et al.
[14] utilized ELECTRE-III under fuzzy environment to assess an
action plan for the diffusion of renewable energy technologies at
regional scale. Polatidis and Haralambopoulos [26] proposed a new
methodological framework of multi-participatory and multicriteria decision making to evaluate renewable energy options in
Greece. Providing an integrated decision aid framework, Topcu and
Ulengin [27] dealt with the problem of selecting the most suitable
electricity generation alternative for Turkey. Cavallaro and Ciraolo
[28] proposed a multi-criteria method in order to support the
2519
feasibility analysis of installing alternative wind energy turbine
configurations in an island in Italy. Zhou et al.’s [29] literature
review showed that the importance of multiple criteria decisionmaking methods and energy-related environmental studies have
increased substantially since 1995. Begic and Afgan [30] evaluated
the options of energy power systems for Bosnia Herzegovina under
a multi-criteria sustainability assessment framework. Burton and
Hubacek [31] compared the perceived social, economic, and environmental cost of small-scale energy technologies to larger-scale
alternatives. Afgan et al. [32] evaluated the potential natural gas
usage in energy sector. Önüt et al. [33] employed analytic network
process (ANP) to solve an energy resource selection problem for the
manufacturing industry. Patlitzianas et al. [34] developed an information decision support system, which contains an MCDM
subsystem and applied to 13 accession member states of the European Union. Kahraman et al. [12] used axiomatic design (AD) and
AHP for the selection of the best renewable energy alternative under
fuzzy environment.
Wang et al.’s [35] literature review on the application of the
MCDM techniques to the energy issues shows that evaluation
criteria for energy source and site selection problems can be
grouped into four main categories: Technical, economic, environmental, and social. Table 1 gives a list of the criteria utilized in
energy planning oriented MCDM studies up to 2009. Table 1 also
shows that efficiency, exergy (rational) efficiency, investment cost,
operation and maintenance cost, NOx emission, CO2 emission, land
use, social acceptability, and job creation are the most frequently
used evaluation criteria in energy planning, energy management,
and resource allocation studies.
Cavallaro and Ciraolo [28] used a smaller group of evaluation
criteria for an energy production site selection case study on an
Italian island. The eleven criteria used by the authors can be
grouped into technical, economic, environmental, and social categories as in Table 2:
3. An integrated VIKOR & AHP methodology
A modified fuzzy approach to the classical VIKOR is proposed in
this section. Some basic definitions and notations of fuzzy sets are
summarized in Appendix A. The importance weight of each criterion can be obtained by either a direct-assignment based on
experts’ experiences or by using pairwise comparisons of criteria.
Fig. 2. Turkey’s wind speed map.
2520
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
GOAL: Energy
technology selection
Technical
C1
Economic
C2
C3
E1: Geothermal
Environmental
C4
C5
E2: Solar
E3: Wind
C6
Social
C7
C8
E4: Hydraulic
C9
E5: Biomass
Fig. 3. The hierarchical structure for the selection of the renewable energy technology.
Here, it is suggested that the experts use the linguistic variables in
Table 3 to evaluate the importance of the criteria. Wang et al. [36]
calculated the weight of each criterion by summing the assigned
weights by experts and then dividing the sum by the number of
experts as in Eq. (1):
~ ij ¼
w
1 1
~ j ð þ Þw
~ 2j ð þ Þ/ð þ Þw
~ Kj
w
K
(1)
~ Kj is the importance weight of the Kth expert.
where w
Since a comparison matrix divides the problem into sub-problems which can be solved easier, a pairwise comparison matrix in
the AHP method can be considered a good way of determining the
weights of the criteria. Therefore, we propose modifying the
C1 1
C2 E1: E
E2: E
E3: SS
C2
C3
C4
C5
C6
C7
C8
C9
E1: E
E1: SS E1: SS E1: SW E1: FW E1: SW E1: SS E1: SS
E2: FW E2: FW E2: SW E2: FW E2: SW E2: FW E2: AW
E2: E
E3: SW E3: SW E3: E
E3: SS E3: SS E3: E
E3: SW E3: E
1
E1: SS E1: SS E1: SW E1: FW E1: SW E1: SS E1: SS
E2: FW E2: FW E2: SW E2: FW E2: SW E2: FW E2: VW
E3: SS E3: E
E3: E
E3: SS E3: FS E3: FS E3: SS
C3 E1: SW E1: SW 1
E2: FS E2: FS
E3: SW
E3: E
E1: E
E1: FW E1: VW E1: FW E1: E
E2: E
E2: SS E2: E
E2: SS E2: E
E3: SW E3: SW E3: E
E3: SS E3: SS
E1: E
E2: SW
E3: E
C4 E1: SW E1: SW E1: E
E2: FS E2: FS E2: E
E3: SS
E3: SS E3: E
1
E1: SS
E2: SW
E3: SS
C5 E1: SS
E2: SS
E3: SS
E1: SS
E2: SS
E3: E
C6 E1: FS
E2: FS
E3: E
E1: FS E1: VS
E2: FS E2: E
E3: SW E3: E
E1: FW E1: AW E1: FW E1: E
E2: SS E2: E
E2: SS E2: E
E3: E
E3: SS E3: FS E3: FS
E1: FS E1: FS 1
E2: SW E2: SW
E3: SS E3: E
E1: SW E1: E
E2: SW E2: E
E3: SS E3: FS
E1: AS E1: SS 1
E2: E
E2: SS
E3: SW E3: SW
E1: SS
E2: SS
E3: SS
C7 E1: SS E1: SS E1: FS E1: FS E1: E
E1: SW 1
E2: SW
E2: SS E2: SS E2: SW E2: SW E2: E
E3: SW E3: FW E3: SW E3: FW E3: FW E3: SW
E1: FS E1: FS
E2: SW E2: FW
E3: FS E3: SS
E1: VS
E2: E
E3: SS
E1: VS
E2: SW
E3: E
E1: FS E1: FS
E2: SW E2: FW
E3: SW E3: SS
E1: E
E1: FW E1: VW E1: FW 1
C8 E1: SW E1: SW E1: E
E2: FS E2: FS E2: E
E2: E
E2: SS E2: E
E2: SS
E3: SW E3: FW E3: SW E3: FW E3: FW E3: SW E3: SS
E1: E
E2: SW
E3: SW
C9 E1: SW E1: SW E1: E
E2: AS E2: VS E2: SS
E3: SW E3: E
E3: E
1
E1: SW E1: FW E1: VW E1: FW E1: E
E2: SS E2: FS E2: SS E2: FS E2: SS
E3: SW E3: SW E3: E
E3: SW E3: SS
~
Sj ¼
n
X
j¼1
2
~ 54
M
j
m X
n
X
k¼1 j¼1
3
1
~ 5
M
j
(2)
In our case, n ¼ m since a comparison matrix for criteria always
has to be a square matrix.
Table 5
Pairwise comparisons of renewable energy source evaluation criteria.
C1
classical weighting procedure of VIKOR methodology by using
fuzzy pairwise comparison matrices. Chang’s [37] extent analysis
will be utilized for this purpose.
The stages of extent analysis approach can be summarized as
follows: Letting Cj ¼ {C1,C2,.,Cn} be a criteria set, extent analysis
values for each criterion can be obtained as follows: Let
~ ðj ¼ 1; 2; 3; .; nÞ be TFNs.
M
j
The value of fuzzy synthetic extent for the degree of possibility
~ is defined, respectively, as
~ M
of M
1
2
h
i
~
~ M
¼ sup min mM
V M
~ 1 ðxÞ; mM
~ 2 ðyÞ
1
2
(3)
xy
~
When (x,y) exists such that x y and mM
~ 1 ¼ mM
~ 2 ¼ 1; VðM 1
~
~
~
M2 Þ ¼ 1 is obtained. Since M 1 and M2 are convex fuzzy numbers,
the following principle of the comparison of fuzzy numbers is
applied:
~
~ M
¼ 1
V M
1
2
iff
(4)
m1 m2
and
~ XM
~
~
~ M
¼ hgt M
¼ mðdÞ
V M
1
2
1
2
(5)
where d is the ordinate of the highest intersection point D between
~ ¼ ðl ; m u Þ and M
~ ¼ ðl ; m ; u Þ, the
mM~ 1 and mM~ 2 . If M
1
1
1; 1
2
2
2 2
following equation for the ordinate of the point D is given (see
Fig. 1);
8
>
> 0; if m2 m1
<
>
~ M
~
~
~ XM
1; if l1 u2
ð6Þ
V M
¼
hgt
M
¼
2
1
1
2
>
>
l1 u2
>
;
otherwise
: ðm u Þ ðm l Þ
2
2
1
1
~ M
~ Þ and VðM
~ M
~ Þ are required for
The values of VðM
1
2
2
1
~ . The degree of possibility for a convex fuzzy
~ and M
comparing M
1
2
number to be greater than p convex fuzzy numbers
~ ; j ¼ 1;2;3;.;nÞ is defined as
ðM
j
2521
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Table 6
Fuzzy evaluation matrix for the weights.
C1
C2
C3
C4
C5
C6
C7
C8
C9
C1
C2
C3
(1, 1, 1)
(1, 1, 1.17)
(0.89, 1.17, 1.33)
(0.89, 1.17, 1.5)
(1, 1, 1.5)
(1, 1.33, 1.67)
(0.89, 1, 0.89)
(0.78, 1.17, 1.33)
(1.22, 1.5, 1.67)
(0.89, 1, 1)
(1, 1, 1)
(0.78, 1.17,
(0.89, 1.17,
(1, 1, 1.33)
(0.89, 1.33,
(0.83, 0.89,
(0.72, 1.06,
(0.94, 1.33,
(0.83, 0.89,
(0.83, 0.89,
(1, 1, 1)
(1, 1, 1.17)
(0.89, 1.17,
(1.17, 1.33,
(0.78, 1.17,
(0.89, 1, 1)
(1, 1, 1.17)
1.33)
1.33)
1.67)
1.33)
1.33)
1.5)
C4
1.17) (0.72, 0.89, 1.17)
1.33) (0.83, 0.89, 1.17)
(0.89, 1, 1)
(1, 1, 1)
1.5) (0.89, 1.17, 1.33)
1.5) (1.22, 1.5, 1.67)
1.33) (0.72, 1.06, 1.33)
(0.83, 0.89, 1)
(0.78, 1, 1.17)
C5
C6
C7
C8
C9
(0.67, 1, 1)
(0.78, 1, 1)
(0.72, 0.89, 1.17)
(0.83, 0.89, 1.17)
(1, 1, 1)
(0.89, 1, 1.33)
(0.83, 0.89, 1)
(0.67, 0.78, 1.17)
(0.72, 1.06, 1.33)
(0.67, 0.78, 1)
(0.67, 0.78, 1.17)
(0.8, 0.83, 0.89)
(0.78, 0.8, 1)
(0.78, 1, 1.17)
(1, 1, 1)
(0.67, 1, 1)
(0.69, 0.83, 0.89)
(0.8, 0.83, 1.06)
(0.78, 1, 1.17)
(0.78, 1.17, 1.33)
(0.83, 0.89, 1.33)
(0.83, 1.06, 1.5)
(1, 1.17, 1.33)
(1, 1, 1.5)
(1, 1, 1)
(0.83, 0.89, 1.33)
(0.72, 1.06, 1.33)
(0.83, 0.89, 1.33)
(0.83, 1.06, 1.5)
(1, 1, 1.17)
(1, 1.17, 1.33)
(0.89, 1.33, 1.67)
(1.17, 1.33, 1.67)
(0.78, 1.17, 1.33)
(1, 1, 1)
(1, 1, 1.33)
(0.78, 0.8, 1)
(0.8, 0.83, 1.22)
(0.89, 1, 1)
(0.89, 1, 1.33)
(0.83, 1.06, 1.5)
(1.06, 1.33, 1.5)
(0.83, 1.06, 1.5)
(0.78, 1, 1)
(1, 1, 1)
CR for the defuzzified version of this matrix is 0.027<0.10.
h
~p
~ ;M
~ ; .; M
~
~
~
~p M
¼ V M
V M
1
2
p 1 ; M pþ1 ; .; M n
i
~p M
~p M
~
~
~n
and M
and.and M
M
1
2
~p M
~
¼ d Cj ; jsp
¼ min V M
j
2
(7)
Consequently, the weight vector W 0 ¼ ðd0 ðC1 Þ; d0 ðC2 Þ; .; d0
ðCn ÞÞT ; j ¼ 1; 2; 3; .; n is obtained. Finally, via normalization, the
following normalized weight vector is obtained:
W ¼ ðdðC1 Þ; dðC2 Þ; .; dðCn ÞÞT
(8)
Obtaining the weight vector via the extent analysis, we can
continue implementing the steps of VIKOR. VIKOR method is based
on the compromise programming of MCDM. The concepts of
compromise solutions were first demonstrated by Yu [38] and
Zeleny [39]. The methodology simply works on the principle that
each alternative can be evaluated by each criterion function; the
compromise ranking will be realized by comparing the degrees of
closeness to the ideal alternative. In fuzzy VIKOR, it is suggested
that decision makers use linguistic variables to evaluate the ratings
of alternatives with respect to criteria. Table 4 gives the linguistic
scale for the evaluation of alternatives. Assuming that a decision
group has K people, the ratings of alternatives with respect to each
criterion can be calculated as in Eq. (9) [36];
~
xij ¼
1 1
~
xij ð þ Þ~
x2ij ð þ Þ/ð þ Þ~
xKij
K
(9)
where ~
xKij is the rating of the Kth expert for ith alternative with
respect to jth criterion.
After obtaining the weights of criteria and fuzzy ratings of
alternatives with respect to each criterion, we can now express the
fuzzy multi-criteria decision-making problem in matrix format as,
~
~ ¼ ðl ; m ; u Þ
Sj ¼ M
j
j
j; j
Wj0 ¼ d0 ðCj ÞT
Wj ¼ dðCj ÞT
(0.072,
(0.075,
(0.078,
(0.081,
(0.083,
(0.094,
(0.074,
(0.072,
(0.082,
0.564
0.663
0.656
0.743
0.841
1.000
0.713
0.613
0.807
0.086
0.101
0.100
0.113
0.128
0.152
0.109
0.093
0.123
0.099,
0.103,
0.107,
0.111,
0.119,
0.134,
0.111,
0.103,
0.117,
0.139)
0.154)
0.145)
0.16)
0.174)
0.191)
0.152)
0.142)
0.163)
~
x12
/
«
~
xm2
/
~
x2n
/
/
~
x1n
3
7
7
7
« 5
(10)
~
xmn
W ¼ ½w1 ; w2 ; .; wn ; j ¼ 1; 2; .; n
where ~
xij is the rating of Alternative Ai with respect to Criterion j (i.
e. Cj) and wj denotes the importance weight of Cj.
*
Next step is to determine the fuzzy best value ðFBV; ~f j Þ and the
~
fuzzy worst value ðFWV; f j Þ of each criterion function.
~f * ¼ max ~
xij ; j˛B;
j
~f ¼ min ~
xij ; j˛C
j
i
*
~ j ð~f j
Then, the values w
order to obtain
~
Si ¼
n
X
j¼1
*
~ j ~f j
w
~
xij
*
~ ¼ max w
~ j ~f j
R
i
*
~
xij Þ=ð~f j
~f *
j
~
xij
j
(11)
i
~f Þ;
j
~f
j
~f *
j
~
~ are computed in
Si and R
i
~f
j
(12)
(13)
where ~
Si refers to the separation measure of Ai from the fuzzy best
~ to the separation measure of Ai from the fuzzy worst
value, and R
i
value.
*
~ values are calculated:
~* ; R
~ ; and Q
S ;R
In the next step, ~
S ;~
i
*
~
S ¼ maxi ~
S ¼ mini ~
Si ; ~
Si
*
~ ¼ min R
~; R
~
~ ¼ max R
R
i
i
i i
(14)
Table 8
Evaluation scores of the renewable energy alternatives.
C1
C2
C3
C4
C5
C6
C7
C8
C9
E1 E1: G
E1: F
E1: F
E1: F
E1: MP E1: MP E1: G
E1 : F
E1: F
E2: MG E2: MG E2: F
E2: F
E2: MG E2: MG E2: MG
E2: MG E2: F
E3: MG E3: MG E3: F
E3: F
E3: MP E3: G
E3 : G
E3: G
E3: G
E2 E1: MG E1: VG E1: G
E1: MG E1: VG E1: G
E1: MG E1: MG E1: MP
E2: MG E2: MG E2: F
E2: F
E2: G
E2: MG E2: MG E2: F
E2: F
E3: MG E3: G
E3: G
E3: F
E3: MG E3: G
E3: VG E3: VG
E3: F
Table 7
Results of the fuzzy AHP extent analysis procedure for the weights.
C1
C2
C3
C4
C5
C6
C7
C8
C9
~
x
6 ~11
6 x21
~
D ¼ 6
4 «
~
xm1
E3 E1: G
E1: P
E1: G
E1: G
E1: F
E2: F
E2: G
E2: MG E2: MP E2: G
E3: MG E3: MG E3: MG E3: G
E3: G
E1: G
E1: G
E1 : G
E2: VG E2: VG E2: G
E3: G
E3: VG E3: G
E4 E1: F
E2: F
E3: F
E1: F
E2: G
E3: G
E1: MG E1: VP E1: VG E1: G
E2: G
E2: VP E2: VP E2: F
E3: MG E3: VG E3: VG E3: G
E1: G
E2: G
E3: G
E1: G
E1 : P
E1: P
E2: G
E2 : P
E2: VP
E3: MP E3: VG E3: G
E5 E1: MG E1: G
E1: G
E1: G
E1: MP E1: F
E1: F
E2: MG E2: F
E2: MG E2: F
E2: F
E2: F
E2: G
E3: MP E3: F
E3: MG E3: MG E3: MG E3: MP E3: F
E1 : F
E1: MG
E2: MG E2: G
E3: MG E3: G
2522
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Table 9
Fuzzy evaluation matrix for the alternatives.
C2
C1
E1
E2
E3
E4
E5
(6.33, 8.33,
(3.67, 5.67,
(6.33, 8.33,
(3, 5, 7)
(5.67, 7.67,
~ ¼ v ~
Si
Q
i
C3
9.67) (3.67, 5.67, 7.67)
7.67) (6.33, 8, 9.33)
9.67) (2, 3.67, 5.67)
(5.67, 7.67, 9.33)
9.33) (5.67, 7.67, 9.33)
*
~
S
~
S
*
~
S
C4
C5
þ ð1
~
vÞ R
i
~*
R
~
R
~*
R
(15)
~ are related to a maximum
The indices min ~
Si and min R
i
i
i
majority rule, and a minimum individual regret of an opponent
strategy, respectively. As well, v is introduced as the weight of the
strategy of the maximum group utility. v is usually assumed to be 0.5.
Next task is the defuzzification of the triangular fuzzy number
~ . Various defuzzifi~ and ranking the alternatives by the index Q
Q
i
i
cation strategies have been suggested in the literature. In this
paper, the graded mean integration approach is used [40].
According to the graded mean integration approach, for triangular
~ ¼ ðc ; c ; c Þ can be transformed
fuzzy numbers, a fuzzy number C
1 2 3
into a crisp number by employing the below equation:
~ ¼ C ¼ c1 þ 4c2 þ c3
P C
6
(16)
Finally, the best alternative with the minimum of Qi is
determined.
To summarize the methodology, the steps of the modified fuzzy
VIKOR approach are given in the following:
Step 1: A group of decision-makers identifies the evaluation
criteria.
Step 2: Appropriate linguistic variables for the weights of the
criteria and alternatives are chosen.
Step 3: A pairwise comparison matrix for the criteria is constructed
and experts’ linguistic evaluations are aggregated to get
a mean value for each pairwise comparison.
Step 4: Extent analysis approach is used to obtain the weights of
the criteria.
Step 5: Linguistic evaluations of the experts are aggregated to get
the fuzzy ratings of the alternatives with respect to each
criterion.
Step 5: Fuzzy decision matrix is constructed for the implementation of VIKOR.
*
Step 6: Fuzzy best value ðFBV; ~f j Þ and fuzzy worst value ðFWV; ~f j Þ
of each criterion function are determined.
~ Þ are calculated.
Step 7: Separation measures ð~
Si and R
i
~ values are calculated.
Step 8: Q
i
~ values are defuzzified and the alternatives are ranked by
Step 9: Q
i
the index Qi.
Step10: The best alternative with the minimum of Qi is
determined.
Table 10
Separation measures of Ai from the fuzzy best and fuzzy worst values.
E1
E2
E3
E4
E5
C6
(4.33, 6.33, 8.33) (3.67, 5.67, 7.67) (2.33, 4.33,
(6.33, 8.33, 9.67) (5, 7, 8.67)
(5, 6.67, 8)
(6.33, 8.33, 9.67) (5, 7, 8.67)
(5.67, 7.67,
(3, 3.33, 4)
(6, 6.67, 7)
(5.67, 7.67,
(5, 7, 8.67)
(4.33, 6.33, 8)
(2.33, 4.33,
~
Si
~
R
i
(18.09, 31.15, 61.08)
(17.2, 27.19, 46.41)
(16.31, 25.95, 45.95)
(24.34, 37.2, 61.1)
(17.41, 29.03, 54.42)
(2.98, 8.23, 12.6)
(2.9, 6.66, 11.57)
(2.64, 6.66, 12.67)
(4.78, 7, 14.14)
(2.86, 7.36, 11.35)
6.33) (1.67,
(6.33,
9)
(7.67,
9)
(5.67,
6.33) (2.33,
C7
3.67,
8.33,
9.33,
7.67,
4.33,
5.67)
9.67)
10)
9)
6.33)
C8
C9
(6.33, 8.33, 9.67) (5, 7, 8.67)
(5.67, 7.67, 9.33) (6.33, 8, 9.33)
(8.33, 9.67, 10)
(7, 9, 10)
(5, 7, 8.33)
(3, 4, 5.33)
(3, 5, 7)
(4.33, 6.33, 8.33)
(5, 7, 8.67)
(4.33, 6, 7.33)
(7, 9, 10)
(2.33, 3.33, 4.67)
(6.33, 8.33, 9.67)
4. An application: renewable energy planning for Istanbul
For a country, energy is one of the key indicators to show
economic and social development and improved quality of life.
Turkish energy consumption has risen dramatically over the past 20
years due to the combined demands of industrialization and
urbanization. It has increased from 32 mtoe (million tons of oil
equivalent) in 1980 to 74 mtoe in 1998. According to the planning
studies, Turkey’s final consumption of primary energy is estimated to
be 171 mtoe in 2010 and 298 mtoe in 2020. In other words, domestic
energy production will probably meet 28% of the total primary
energy demand in 2010 and 24% in 2020. The level of Turkey’s energy
consumption is still low relative to similar sized countries [41].
As fossil fuel energy becomes scarcer, Turkey and its most
populated and industrialized city, Istanbul, will face energy shortages, increasing energy prices, and energy insecurity within the
next few decades. Therefore, the development and use of renewable energy sources and technologies are increasingly becoming
vital for the sustainable economic development of Turkey.
In recent years some authors have made important contributions on the renewable energy literature on Turkey such as Ediger
and Kentel [42], Hepbaşlı et al. [43], Demirbaş [44,45], Kaygusuz
[46e48], Kaygusuz and Sarı [49], Evrendilek and Ertekin [50], Balat
[51,52], Demirbaş and Bakış [53], Hepbaşlı and Ozgener [54,55].
Among the different forms of renewable energy, biomass energy is
one of the major resources in Turkey. Turkey’s domestic energy
consumption accounts for about 37% of the total energy
consumption. Of this, about 52% is from biomass-based fuels.
Turkey’s first biomass power project is under development in
Adana province, at an installed capacity of 45 MW. Two others, at
a total capacity of 30 MW, are at the feasibility study stage in Mersin
and Tarsus provinces.
According to the studies on the determination of Turkey’s wind
energy potential, Turkey’s gross wind energy potential has been
estimated as 400 billion kW h/year and technical potential has been
estimated as 120 billion kW h/year which is equal to the 1.2 times of
the current annual electricity production of Turkey. The most
attractive sites are the Marmara Sea region, Mediterranean Coast,
Aegean Sea Coast and Anatolia inland. In Turkey, electricity
production through wind energy for general usage purposes was
first realized at the Ceşme Altınyunus Resort in 1986 by using a 55
kW nominal powered wind turbine. Fig. 2 shows Turkey’s average
wind speed (m/s) map.
4.1. Selection of the best energy source alternative for Istanbul
In this study, as given in Table 1, the most frequently used
technical, economic, environmental and social criteria among the
Table 11
*
~
~ * ; and R
~ values.
S ;~
S ;R
*
~
S
~
S
~*
R
~
R
(16.31, 25.95, 45.95)
(24.34, 37.2, 61.1)
(2.64, 6.66, 11.35)
(4.78, 8.23, 14.14)
2523
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
CO2 emission (C6): CO2 is a colorless, odorless and tasteless gas
that contributes to the greenhouse effect. It is mainly released
through conventional energy systems. CO2 leads to the global
warming, which is focused by many governments, academicians,
and researchers. Naturally, CO2 emission of an energy system is
certainly a criterion to evaluate its sustainability.
Land use (C7): as the environment and landscape are directly
affected by the energy systems, the land required by each plant is
a matter of great concern for their evaluation. Different energy
systems may occupy different land while the products are same.
Thus, land use must be considered by energy experts.
Social acceptability (C8): social acceptability expresses the overview of opinions related to the energy systems by the local population. It is extremely important since the opinion of the population
and pressure groups may heavily influence the amount of time
needed to complete an energy project. It should be noted that social
acceptance is not a directly measurable figure. It is a qualitative one.
Job creation (C9): in the decision making process of local
governments, job creations of energy systems are indispensably
considered and selected to evaluate their contributions.
After determining the evaluation criteria and the alternatives,
the steps of the integrated fuzzy VIKOR-AHP algorithm are implemented. In order to determine the importance of each criterion, the
experts employed a nine point scale given in Table 3. While evaluating the alternatives, the experts assumed that all the criteria
were benefit criteria. For example, if an energy source is evaluated
as ‘very good’ in terms of ‘CO2 emission’, this means that CO2
emission level of that renewable energy alternative is very low. On
the other hand, if an energy alternative is evaluated as ‘very good’ in
terms of ‘technical efficiency’, this means that technical efficiency
of that energy alternative is very high. Each linguistic term is
associated with a triangular fuzzy number. Table 5 gives the results
of the pairwise comparisons of the evaluation criteria made by
three renewable energy planning experts.
In the next step, using Table 3 and Table 5, the fuzzy evaluation
matrix for the criteria weights is obtained as in Table 6. Next, in
order to check the consistency ratio (CR) of the evaluation matrix,
the graded mean integration approach (Eq. (16)) is utilized for
defuzzification. CR for the evaluation matrix is computed as 0.027
and it is less than 0.10. Therefore, the comparison results can be
considered consistent.
Next, synthetic extent values ð~
Si Þ for the evaluation criteria are
produced under fuzzy environment employing Eq. (2). After
obtaining the synthetic extent values, Eqs. (3e7) are utilized for
calculating the vector for the criteria weights. Finally, the normalized weight vector is obtained as in Table 7.
Next step is the determination of the best renewable energy
source alternative with the proposed modified fuzzy VIKOR
Table 12
Integrated fuzzy VIKOR-AHP analysis results.
~
Q
i
E1
E2
E3
E4
E5
(
(
(
(
(
2.42,
2.46,
2.54,
1.89,
2.47,
0.73, 5.11)
0.05, 3.96)
0, 4.19)
0.61, 5.48)
0.36, 4.41)
Qi
Rank order
0.94
0.29
0.27
1.00
0.56
4
2
1
5
3
list of the energy technology selection criteria are utilized in evaluating renewable energy alternatives for Istanbul. The energy
alternatives considered are: Geothermal energy (E1), solar energy
(E2), wind energy (E3), hydraulic energy (E4), and biomass energy
(E5). The structure of the renewable energy planning decision
making problem formulated in this study is presented in Fig. 3.
The criteria used in this study are briefly explained in the
following [35]:
Technical efficiency (C1): efficiency measures how much useful
energy can be obtained from an energy source. The efficiency
coefficient, which is one of the most frequently used measures of
efficiency, is defined as the ratio of the output energy to the input
energy. Efficient energy usage is essential to slowing the energy
demand growth. It is the most used technical criterion to evaluate
energy systems.
Exergy efficiency (C2): exergy efficiency (rational efficiency)
computes the efficiency of a process taking the second law of
thermodynamics into consideration. There is always an exergy loss
when a process involves a temperature change. Exergy is the net
energy that is left to be used. The CHP systems are frequently
evaluated with this criterion.
Investment cost efficiency (C3): the components of investment
costs are the purchase of mechanical equipment, technological
installations, construction of roads and connections to the national
web, engineering services, drilling and other incidental construction work. The investors must consider the costs and the benefits of
investments. Investment cost is the most used economic criterion
to evaluate energy systems.
Operation and maintenance cost efficiency (C4): operation cost
includes the wages and funds spent for the energy, products and
services. Maintenance cost consists of the funds spent for maintenance. The operation and maintenance costs are also divided into
two subcategories: fixed and variable costs.
NOx emission (C5): NOx comprises a group of molecules that can
contribute to air pollution, acid deposition and climate change.
Reacting with organic chemicals, NOx can also form a wide variety
of toxic products which may damage health and cause mutations.
Therefore, NOx emission of an energy system is considered an
important criterion according to most experts.
GOAL: Wind energy
production site selection
C1
L1: Silivri
C3
L2: Çatalca
C4
C8
L3: GOP
C10
L4: Sarıyer
C11
C12
L5: Tuzla
Fig. 4. The hierarchical structure for the selection of the wind energy production site.
L6: ile
2524
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Fig. 5. Geographical distribution of the alternative wind energy production sites in Istanbul.
Then, using Eqs. (11)e(13), separation measures from the fuzzy
~ are computed as in Table 10:
best value ~
Si and the fuzzy worst value R
i
*
~ * and R
~ fuzzy values
S ; R;
S ;~
In the next step, using Eq. (14), ~
are calculated (Table 11).
~ values are computed. In the calculations,
Then, using Eq. (15), Q
i
the weight of the strategy of the maximum group utility (v) is
~ values are defuzzified via graded
assumed to be 0.5. Finally, Q
i
mean integration method (Eq. (16)) and ranked according to Qi
index values. Table 12 gives the results of the integrated fuzzy
VIKOR-AHP analysis results.
Based on the crisp Qi index values, the ranking of the alternatives in descending order is determined as E3,E2, E5, E1 , and E4. The
best alternative is found to be E3 (wind energy). The rank order of
the rest is solar, biomass, geothermal, and hydraulic.
procedure. To do this, three experts evaluated the five renewable
energy alternatives with respect to each criterion using Table 4.
Evaluation results are given in Table 8:
After calculating the arithmetic means of the associated fuzzy
evaluation scores, fuzzy evaluation matrix is obtained as in Table 9:
Table 13
Pairwise comparisons of renewable energy source evaluation criteria.
C1
C3
C4
C8
C10
C11
C12
C1
1
E1: SS
E2: SW
E3: SS
E1: FS
E2: E
E3: E
E1: VS
E2: E
E3: SS
E1: FS
E2: E
E3: SS
E1: FS
E2: SS
E3: SS
E1: SS
E2: E
E3: SW
C3
E1: SW
E2: SS
E3: SW
1
E1: SS
E2: SS
E3: SW
E1: SS
E2: SS
E3: E
E1: SS
E2: SS
E3: E
E1: SS
E2: FS
E3: E
E1: E
E2: E
E3: FW
C4
E1: FW
E2: E
E3: E
E1: SW
E2: SW
E3: SS
1
E1: E
E2: E
E3: SS
E1: E
E2: E
E3: SS
E1: E
E2: SS
E3: SS
E1: SW
E2: E
E3: SW
C8
E1: VW
E2: E
E3: SW
E1: SW
E2: SW
E3: E
E1: E
E2: E
E3: SW
1
E1: E
E2: E
E3: E
E1: E
E2: SS
E3: E
E1: SW
E2: E
E3: FW
C10
E1: FW
E2: E
E3: SW
E1: SW
E2: SW
E3: E
E1: E
E2: E
E3: SW
E1: E
E2: E
E3: E
1
E1: E
E2: SW
E3: E
E1: SW
E2: E
E3: FW
C11
E1: FW
E2: SW
E3: SW
E1: SW
E2: SW
E3: E
E1: E
E2: SW
E3: SW
E1: E
E2: SW
E3: E
E1: E
E2: SS
E3: E
1
E1: SW
E2: SW
E3: FW
C12
E1: SW
E2: E
E3: SS
E1: E
E2: E
E3: FS
E1: SS
E2: E
E3: SS
E1: SS
E2: E
E3: FS
E1: SS
E2: E
E3: FS
E1: SS
E2: SS
E3: FS
1
4.2. Selection of the best wind energy plantation site for Istanbul
In this study, in order to evaluate the alternative wind turbine
plantation sites in Istanbul, a combination of the most frequently
used evaluation criteria by Cavallaro and Ciraolo’s [28] and Wang’s
[35] will be used. Other than technical efficiency (C1), investment cost
efficiency (C3), operational and maintenance cost efficiency (C4), and
social acceptability (C8), three new criteria which were proposed by
Cavallaro and Ciraolo [28] for the evaluation of alternative wind
turbine projects are employed: Visual impact (C10), acoustic noise
(C11), and impact on ecosystems (C12). Below are the definitions of
these criteria.
Visual impact (C10): this criterion reflects the visual nuisance that
may be created by the establishment of a wind turbine in a specific
area. The landscape of the different sites, the distance from the
Table 14
Fuzzy evaluation matrix for the site selection criteria weights.
C1
C3
C4
C8
C10
C11
C12
C1
C3
C4
C8
C10
(1, 1, 1)
(0.78, 1, 1.17)
(0.83, 0.89, 1)
(0.69, 0.83, 0.89)
(0.72, 0.89, 1)
(0.61, 0.89, 1)
(0.89, 1, 0.89)
(0.89, 1, 1.33)
(1, 1, 1)
(0.78, 1, 1.17)
(0.78, 1, 1)
(0.78, 1, 1)
(0.72, 0.89, 1)
(1, 1.17, 1.33)
(1, 1.17, 1.33)
(0.89, 1, 1.33)
(1, 1, 1)
(0.89, 1, 1)
(0.89, 1, 1)
(0.78, 1, 1)
(1, 1, 1.33)
(1.17, 1.33, 1.67)
(1, 1, 1.33)
(1, 1, 1.17)
(1, 1, 1)
(1, 1, 1)
(0.89, 1, 1)
(1, 1.17, 1.5)
(1,
(1,
(1,
(1,
(1,
(1,
(1,
CR for the defuzzified version of this matrix is 0.012 < 0.10.
1.17, 1.5)
1, 1.33)
1, 1.17)
1, 1)
1, 1)
1, 1.17)
1.17, 1.5)
C11
C12
(1, 1.17, 1.67)
(1, 1.17, 1.5)
(1, 1, 1.33)
(1, 1, 1.17)
(0.89, 1, 1)
(1, 1, 1)
(1, 1.17, 1.67)
(0.89, 1, 1.17)
(0.83, 0.89, 1)
(0.78, 1, 1)
(0.72, 0.89, 1)
(0.72, 0.89, 1)
(0.61, 0.89, 1)
(1, 1, 1)
2525
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Table 15
Results of the fuzzy AHP procedure for the site selection criteria weights.
C1
C3
C4
C8
C10
C11
C12
Table 18
Separation measures of Ai from the fuzzy best and fuzzy worst values.
~
~ ¼ ðl ; m ; u Þ
Sj ¼ M
j
j
j;
j
Wj0 ¼ d0 ðCj ÞT
Wj ¼ dðCj ÞT
(0.123,
(0.115,
(0.113,
(0.107,
(0.106,
(0.099,
(0.122,
1
0.822
0.738
0.618
0.622
0.622
0.962
0.186
0.153
0.137
0.115
0.115
0.116
0.179
0.158,
0.142,
0.139,
0.135,
0.137,
0.134,
0.155,
0.218)
0.195)
0.176)
0.159)
0.158)
0.161)
0.208)
L1
L2
L3
L4
L5
L6
Table 16
Evaluation scores of the wind energy production site alternatives.
C1
C3
C4
C8
C10
C11
C12
E1: MG
E2: MG
E3: F
E1: F
E2: F
E3: F
E1: MG
E2: G
E3: G
E1: G
E2: G
E3: G
E1: G
E2: MP
E3: MG
E1: G
E2: G
E3: F
E1: P
E2: MP
E3: F
L2
E1: G
E2: MG
E3: MG
E1: MG
E2: F
E3: G
E1: G
E2: MG
E3: MG
E1: M
E2: G
E3: G
E1: F
E2: F
E3: G
E1: MG
E2: G
E3: MG
E1: F
E2: F
E3: G
L3
E1: VG
E2: VG
E3: G
E1: G
E2: G
E3: G
E1: F
E2: G
E3: G
E1: F
E2: MG
E3: MG
E1: VP
E2: G
E3: F
E1: F
E2: MG
E3: F
E1: F
E2: F
E3: G
L4
E1: P
E2: P
E3: F
E1: F
E2: F
E3: F
E1: MG
E2: G
E3: G
E1: F
E2: MG
E3: G
E1: F
E2: P
E3: G
E1: F
E2: P
E3: G
E1: G
E2: VG
E3: F
L5
E1: P
E2: VP
E3: F
E1: MP
E2: P
E3: G
E1: F
E2: VP
E3: VG
E1: P
E2: P
E3: G
E1: P
E2: G
E3: G
E1: VP
E2: P
E3: G
E1: VG
E2: P
E3: MG
L6
E1: P
E2: VP
E3: F
E1: F
E2: MG
E3: MG
E1: F
E2: G
E3: MG
E1: P
E2: F
E3: G
E1: F
E2: MG
E3: MG
E1: MP
E2: F
E3: F
E1: VP
E2: G
E3: MP
~
R
i
(12.7, 18.08, 33.85)
(11.81, 15.79, 24.87)
(13.61, 18.54, 29.1)
(14.28, 20.97, 38.28)
(17.55, 25.81, 45.42)
(14.69, 22.29, 41.76)
(3.79,
(3.79,
(5.05,
(4.55,
(3.95,
(3.64,
5.2, 9.44)
5.2, 7.14)
7.05, 9.28)
6.58, 9.28)
5.2, 8.5)
5.2, 8.5)
average wind speed and total number of turbines constructible on
the site and represents the total energy which can be generated by
the project.
The considered location alternatives are: Silivri (L1), Çatalca (L2),
Gaziosmanpaşa (GOP) (L3), Sarıyer (L4), Tuzla (L5), and Şile (L6).
Fig. 4 gives the hierarchical structure of the wind turbine site
selection problem.
Fig. 5 gives the geographical distribution of the alternative wind
energy production sites in Istanbul.
After determining the evaluation criteria and the alternatives,
the steps of the AHP extent analysis are performed. In order to
determine the importance of each turbine site selection criterion,
the experts employed the nine point scale given in Table 3. Table 13
gives the results of the pairwise comparisons of the evaluation
criteria made by three experts.
The fuzzy evaluation matrix for the weights of the turbine site
selection criteria is obtained as in Table 14. CR for the evaluation
matrix is computed as 0.012 and it is less than 0.10. The comparison
results can be considered consistent.
Next, synthetic extent values ð~
Si Þ for the production site evaluation criteria are computed. After obtaining the synthetic extent
values, the vector for the criteria weights and its normalized
version are obtained as in Table 15.
Final step is the determination of the best wind energy
production site with the proposed fuzzy VIKOR approach. To do
this, first, three experts evaluated the six wind turbine plantation
areas with respect to each site selection criterion using Table 4.
Evaluation results are given in Table 16:
After calculating the arithmetic means of the associated fuzzy
evaluation scores, the fuzzy evaluation matrix is obtained as in
Table 17:
Then, separation measures from the fuzzy best value ~
Si and the
~ are computed as in Table 18:
fuzzy worst value R
i
*
~ * and R
~ fuzzy values are
S ;R
S ;~
In the next step, using Eq. (14) ~
calculated (Table 19).
~ values are computed. In the calculations, v is assumed
Then, Q
i
~ values are defuzzified and ranked according to
to be 0.5. Finally, Q
i
Qi index values. Table 20 gives the results of the integrated fuzzy
VIKOR-AHP analysis results.
Based on the crisp Qi index values, the ranking of the alternatives in descending order are L2,L1, L6, L3, L5, and L4. The best
alternative is found to be L2 (Çatalca). The second best alternative is
L1 (Silivri). The rank order of the rest is Şile, GOP, Tuzla, and Sarıyer.
nearest observer, the type and size of plants to be installed and the
possibility to integrate them with their surroundings must all be
considered when evaluating the alternatives. This criterion is
evaluated in qualitative terms.
Acoustic noise (C11): noise can generally be classified according
to its two main sources: aerodynamic and mechanical. Aerodynamic noise is produced when the turbine blades interact with
the atmospheric turbulence. Mechanical noise is generated by
machineries such as gearboxes and generators. Noise could be
reduced by better-designed turbine blade geometry and careful
choice of operating conditions. This criterion takes into the noise
levels generated by the turbines and the distance of residential
areas into account.
Impact on ecosystems (C12): this subjective criterion refers to the
potential risk to ecosystems caused by production of the various
projects included in the strategies and is evaluated in qualitative
terms. The potential disturbance to fauna caused by wind turbines
do present some problems for predatory species of migrating birds
which pass from Istanbul.
It should also be noted that, when it is evaluated in a wind
turbine establishment context, efficiency (C1) criterion takes
L1
~
Si
Table 17
Fuzzy evaluation matrix for the alternative production sites.
L1
L2
L3
L4
L5
L6
C1
C3
C4
(4.33, 6.33, 8.33)
(5.67, 7.67, 9.33)
(8.33, 9.67, 10)
(1, 2.33, 4.33)
(1, 2, 3.67)
(1, 2, 3.67)
(3, 5, 7)
(5, 7, 8.67)
(5.67, 7.67, 9)
(3, 5, 7)
(2.67, 4.33, 6)
(4.33, 6.33, 8.33)
(6.33, 8.33,
(5.67, 7.67,
(5.67, 7.67,
(6.33, 8.33,
(4, 5, 6)
(5, 7, 8.67)
9.67)
9.33)
9)
9.67)
C8
C10
(7, 9, 10)
(6.33, 8.33, 9.67)
(4.33, 6.33, 8.33)
(5, 7, 8.67)
(2.33, 3.67, 5.33)
(3.33, 5, 6.67)
(4.33,
(4.33,
(3.33,
(3.33,
(4.67,
(4.33,
C11
6.33, 8)
6.33, 8)
4.67, 6)
5, 6.67)
6.33, 7.67)
6.33, 8.33)
(5.67,
(5.67,
(3.67,
(3.33,
(2.33,
(2.33,
C12
7.67, 9)
7.67, 9.33)
5.67, 7.67)
5, 6.67)
3.33, 4.67)
4.33, 6.33)
(1.33,
(4.33,
(4.33,
(6.33,
(4.67,
(2.67,
3, 5)
6.33, 8)
6.33, 8)
8, 9)
6, 7.33)
4, 5.33)
2526
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
Table 19
*
~
~ * ; and R
~ values for the site selection problem
S ;~
S ;R
*
~
S
~
S
~*
R
~
R
(11.81, 15.79, 24.87)
(17.55, 25.81, 45.42)
(3.64, 5.2, 7.14)
(5.05, 7.05, 9.44)
A fuzzy number is a fuzzy subset in the universe of discourse X
that is both convex and normal. shows a fuzzy number ~s of the
universe of discourse X which is both convex and normal [57].
Table 20
Integrated fuzzy VIKOR-AHP analysis results for site selection.
~
Q
i
L1
L2
L3
L4
L5
L6
(
(
(
(
(
(
0.85, 0.16, 3.92)
0.88, 0, 2.33)
0.65, 0.49, 3.3)
0.7, 0.59, 4.42)
0.66, 0.7, 5.13)
0.8, 0.45, 4.68)
Qi
Rank order
0.57
0.22
0.89
1.00
0.96
0.77
2
1
4
6
5
3
5. Conclusions
Renewable energy is generated from natural resources such as
sunlight, wind, rain, tides and geothermal heat. In many developing
countries energy projects have demonstrated that renewable
energy can make significant contributions to the economy by
providing the energy needed for creating new businesses and
employment. Moreover, renewable energy technologies make
indirect contributions like providing energy for education, cooking,
space heating, and lighting. Many countries and states have
implemented incentives like government tax subsidies, partial
payment schemes and rebates over purchase of renewables in
order to encourage consumers to shift to renewable energy sources.
Turkey is a developing country which is extensively dependent
on energy imports. Istanbul is the most populated and energy
consuming city of Turkey. Besides, Istanbul is a rich region for the
purposes of renewable energy generation. Considering the future
needs of the region, our study focused on the selection of the most
appropriate renewable energy investment and its location in
Istanbul. A selection among the renewable energy alternatives has
been made using an integrated VIKOR & AHP methodology. Then,
employing the same methodology, a selection among the wind
energy plantation sites has been made.
In the first step nine evaluation criteria were taken into
consideration. The results of the multi-criteria decision analysis
suggest that the wind energy is the best renewable energy alternative for the region. The ranking of the other alternatives in
descending order is determined as solar, biomass, geothermal, and
hydraulic. Secondly, using seven criteria, the site selection problem
is solved for wind energy production. The rank order of alternative
areas from the best to the worst is obtained as Çatalca, Şile, GOP,
Tuzla, and Sarıyer. The proposed methodology has been successfully applied for both decision problems.
In the future research, similar studies can be conducted based
on different multi-criteria decision-making techniques such as
fuzzy PROMETHEE, fuzzy ELECTRE or fuzzy TOPSIS for comparative
purposes.
Appendix A
A linguistic variable is a variable whose values are linguistic
terms [56]. The concept of linguistic variable is very useful in
dealing with situations which are too complex or ill-defined to be
reasonably described in conventional quantitative expressions. The
linguistic values can be represented by fuzzy numbers.
The a-cut of a fuzzy number ~s is defined
~sa ¼
n
o
xi : m~s ðxi Þ a; xi ˛X
(A1)
where l˛½0; 1.
~s is a non-empty bounded closed
interval contained in X and it
i
can be denoted by ~sa ¼ ½sal ; sau ; sal and sau are the lower and
upper bounds of the closed interval, respectively. Fig. 2 shows
a fuzzy number ~s with a-cuts, where
i
h
~sa1 ¼ sal 1 ; sau1 ;
i
h
~sa2 ¼ sal 2 ; sau2 :
(A2)
From ,we can see that if a2 a1 , then sal 2 sal 1 and sau1 sau2 .
A triangular fuzzy number (TFN) ~s can be defined by a triplet
ðs1 ; s2 ; s3 Þ shown in .The membership function m~s ðxÞ is defined as in
Eq. (A3):
8
0;
>
>
< x s1 ;
m~s ðxÞ ¼ sx2 ss31
>
;
>
: s2 s3
0;
x1 s1
s1 x s2
s2 x s3
x s3
(A3)
If ~s is a fuzzy number and sal > 0 for a˛½0; 1, then ~s is called
r; ~s
a positive fuzzy number. Given any two positive fuzzy numbers ~
and a positive
real
number
r,
the
a
-cut
of
two
fuzzy
numbers
are
i
~
ra ¼ ½ral ; ram and ~sa ¼ ½sal ; sam ; ða˛½0; 1Þ, respectively. Some
~ and ~s can be
main operations of positive fuzzy numbers r
expressed as follows [58]:
a
~
rðþÞ~s ¼ ral þ sal ; rau þ sau ;
(A4)
T. Kaya, C. Kahraman / Energy 35 (2010) 2517e2527
a
r~ð Þ~s ¼ ral
r~ð$Þ~s
a
r~ð:Þ~s ¼
~a
a
r
¼ ral $sal ; rau $sau ;
1
sau ; rau
¼
(A6)
(A7)
#
1 1
a; a ;
(A8)
ru rl
a
¼ ral $r; rau $r ;
a
¼
r~ð:Þr
(A5)
#
r~ð$Þr
ral rau
;
;
sau sal
"
"
sal ;
a a
rl ru
;
;
r r
(A9)
(A10)
~ is a triangular fuzzy number and sal > 0; sau 1 for a˛½0; 1,
If n
then ~s is called a normalized positive triangular fuzzy number [59].
References
[1] Hiremath RB, Shikha S, Ravindranath NH. Decentralized energy planning;
modeling and applicationda review. Renewable and Sustainable Energy
Reviews 2007;11:729e52.
[2] Samouilidis J, Mitropoulos C. Energy economy models e a survey. European
Journal of Operational Research 1982;25:200e15.
[3] Meirer P, Mubayi V. Modeling energy-economic interactions in developing
countries-a linear programming approach. European Journal of Operational
Research 1983;13:41e59.
[4] Pohekar SD, Ramachandran M. Application of multi-criteria decision making
to sustainable energy planning e a review. Renewable and Sustainable Energy
Reviews 2004;8:365e81.
[5] Zadeh L. Fuzzy sets. Information Control 1965;8:338e53.
[6] Opricovic S, Tzeng GH. Compromise solution by MCDM methods: a comparative analysis of VIKOR and TOPSIS. European Journal of Operational Research
2004;156(2):445e55.
[7] Tzeng G, Lin C, Opricovic S. Multi-criteria analysis of alternative-fuel buses for
public transportation. Energy Policy 2005;33:1373e83.
[8] Opricovic S, Tzeng GH. Extended VIKOR method in comparison with outranking
methods. European Journal of Operational Research 2007;178:514e29.
[9] Chu MT, Shyu J, Tzeng GH, Khosla R. Comparison among three analytical
methods for knowledge communities group-decision analysis. Expert Systems
with Applications 2007;33:1011e24.
[10] Cevikcan C, Cebi S, Kaya I. Fuzzy VIKOR and fuzzy axiomatic design versus to
fuzzy TOPSIS: an application of candidate assessment. Journal of Multiple
Valued Logic & Soft Computing 2009;15:181e208.
[11] Mendoza GA, Prabhu R. Multiple criteria decision making approaches to
assessing forest sustainability using criteria and indicators: a case study.
Forest Ecology and Management 2000;131:107e26.
_ Cebi S. A comparative analysis for multiattribute selection
[12] Kahraman C, Kaya I,
among renewable energy alternatives using fuzzy axiomatic design and fuzzy
analytic hierarchy process. Energy 2009;34:1603e16.
[13] Özgür MA. Review of Turkey’s renewable energy potential. Renewable Energy
2008;33:2345e56.
[14] Beccali M, Cellura M, Mistretta M. Decision-making in energy planning:
application of the electre method at regional level for the diffusion of
renewable energy technology. Renewable Energy 2003;28:2063e87.
[15] Beccali M, Cellura M, Ardente D. Decision making in energy planning: the
ELECTRE multicriteria analysis approach compared to a fuzzy-sets methodology. Energy Conversion Management 1998;39(16e18):1869e81.
[16] Krukanont P, Tezuka T. Implications of capacity expansion under uncertainty
and value of information: the near-term energy planning of Japan. Energy
2007;32:1809e24.
[17] Dicorato M, Forte G, Trovato M. Environmental-constrained energy planning
using energy-efficiency and distributed-generation facilities. Renewable
Energy 2008;33:1297e313.
[18] Tsoutsos T, Drandaki M, Frantzeskaki N, Iosifidis E, Kiosses I. Sustainable
energy planning by using multi-criteria analysis application in the island of
Crete. Energy Policy 2009;37:1587e600.
[19] Tan RR, Foo DCY. Pinch analysis approach to carbon-constrained energy sector
planning. Energy 2007;32:1422e9.
[20] Keeney RL, Renn O, Winterfeldt DV. Structuring West Germany’s energy
objectives. Energy Policy 1987;15:352e62.
[21] Hamalainen RP, Karjalainen R. Decision support for risk analysis in energy
policy. European Journal of Operational Research 1992;56:172e83.
2527
[22] Mirasgedis S, Diakoulaki D. Multicriteria analysis vs. externalities assessment
for the comparative evaluation of electricity generation systems. European
Journal of Operational Research 1997;102:364e79.
[23] Mavrotas G, Diakoulaki D, Papayannakis L. An energy planning approach
based on mixed 0e1 multiple objective linear programming. International
Transactions in Operational Research 1999;6:231e44.
[24] Afgan NH, Carvalho MG. Multi-criteria assessment of new and renewable
energy power plants. Energy 2002;27:739e55.
[25] Haralambopoulos DA, Polatidis H. Renewable energy projects: structuring
a multicriteria group decision-making framework. Renewable Energy
2003;28:961e73.
[26] Polatidis H, Haralambopoulos DA. Local renewable energy planning: a participatory multi-criteria approach. Energy Sources 2004;26:1253e64.
[27] Topcu YI, Ulengin F. Energy for the future: an integrated decision aid for the
case of Turkey. Energy 2004;29:137e54.
[28] Cavallaro F, Ciraolo L. A multicriteria approach to evaluate wind energy plants
on an Italian island. Energy Policy 2005;33:235e44.
[29] Zhou P, Ang BW, Poh KL. Decision analysis in energy and environmental
modeling: an update. Energy 2006;31(14):2604e22.
[30] Begic F, Afgan NH. Sustainability assessment tool for the decision making in
selection of energy system e Bosnian case. Energy 2007;32:1979e85.
[31] Burton J, Hubacek K. Is small beautiful? A multicriteria assessment of smallscale energy technology applications in local governments. Energy Policy
2007;35:6402e12.
[32] Afgan NH, Pilavachi PA, Carvalho MG. Multi-criteria evaluation of natural gas
resources. Energy Policy 2007;35:704e13.
[33] Önüt S, Tuzkaya UR, Saadet N. Multiple criteria evaluation of current energy
resources for Turkish manufacturing industry. Energy Conversion and
Management 2008;49(6):1480e92.
[34] Patlitzianas KD, Pappa A, Psarras J. An information decision support system
towards the formulation of a modern energy companies’ environment.
Renewable and Sustainable Energy Reviews 2008;12:790e806.
[35] Wang J, Jing Y, Zhang C, Zhao J. Review on multi-criteria decision analysis aid
in sustainable energy decision-making. Renewable and Sustainable Energy
Reviews 2009;13:2263e78.
[36] Wang TC, Liang LJ, Ho CY. Multi-criteria decision analysis by using fuzzy
VIKOR. Proceedings of International Conference on Service Systems and
Service Management 2006;2:901e6.
[37] Chang DY. Applications of the extent analysis method on fuzzy AHP. European
Journal of Operational Research 1996;95:649e55.
[38] Yu PL. A class of solutions for group decision problems. Management Science
1973;19(8):936e46.
[39] Zeleny M. Multiple criteria decision making. New York: McGraw-Hill; 1982.
[40] Yong D. Plant location selection based on fuzzy TOPSIS. International Journal
of Advanced Manufacturing Technologies 2006;28:839e44.
[41] Hepbaşlı A, Ozalp N. Development of energy efficiency and management
implementation in the Turkish industrial sector. Energy Conversion and
Management 2003;44:231e49.
[42] Ediger VS, Kentel E. Renewable energy potential as an alternative to fossil
fuels in Turkey. Energy Conversion and Management 1999;40:743e55.
[43] Hepbaşlı A, Ozdamar A, Ozalp N. Present status and potential of renewable
energy sources in Turkey. Energy Sources 2001;23:631e48.
[44] Demirbaş A. Biomass and the other renewable and sustainable energy options
for Turkey in the twenty-first century. Energy Sources 2001;23:177e87.
[45] Demirbaş A. Importance of biomass energy sources for Turkey. Energy Policy
2008;36:834e42.
[46] Kaygusuz K. Hydropower and biomass as renewable energy sources in Turkey.
Energy Sources 2001;23:775e99.
[47] Kaygusuz K. Environmental impacts of energy utilisation and renewable
energy policies in Turkey. Energy Policy 2002a;30:689e98.
[48] Kaygusuz K. Renewable and sustainable energy use in Turkey: a review.
Renewable and Sustainable Energy Reviews 2002b;6:339e66.
[49] Kaygusuz K. Sarı A. Renewable energy potential and utilization in Turkey.
Energy Conversion and Management 2003;44:459e78.
[50] Evrendilek F, Ertekin C. Assessing the potential of renewable energy sources in
Turkey. Renewable Energy 2003;28:2303e15.
[51] Balat M. The use of renewable energy sources for energy in Turkey and
potential trends. Energy Exploration & Exploitation 2004;22(4):235e51.
[52] Balat M. Use of biomass sources for energy in Turkey and a view to biomass
potential. Biomass and Bioenergy 2005;29:32e41.
[53] Demirbaş A, Bakış R. Energy from renewable sources in Turkey: status and
future direction. Energy Sources 2004;26:473e84.
[54] Hepbaşlı A, Ozgener O. Turkey’s renewable energy sources: part 1. Historical
development. Energy Sources 2004a;26:961e9.
[55] Hepbaşlı A, Ozgener O. Turkey’s renewable energy sources: part 2. Potential
and utilization. Energy Sources 2004b;26:971e82.
[56] Zadeh LA. The concept of a linguistic variable and its application to approximate reasoning. Information Sciences 1975;8:199e249 (I), pp. 301e357 (II).
[57] Chen C. Extensions of the TOPSIS for group decision-making under fuzzy
environment. Fuzzy Sets and Systems 2000;114:1e9.
[58] Kaufmann A, Gupta MM. Introduction to fuzzy arithmetic: theory and applications. New York: Van Nostrand Reinhold; 1985.
[59] Zimmermann HJ. Fuzzy set theory and its applications. 2nd ed. Boston/Dordrech/London: Kluwer Academic Publishers; 1991.