Energy and Power En gi neering, 2011, 3, 557-564
doi:10.4236/epe.2011.34069 Published Online September 2011 (
Copyright © 2011 SciRes. EPE
Multiple Criteria Analysis for Energy Storage Selection
Alexandre Barin1*, Luciane Neves Canha1, Alzenira Da Rosa Abaide1,
Karine Faverzani Magnago1, Breno Wottrich2, Ricardo Quadros Machado3
1Federal University of Santa Maria, Post-Graduate Electrical Engineering Program,
Center for Energy and Environment Studies, Santa Maria, Brazil
2Comillas Pontifical University, ICAI School of Engineering, Madrid, Spain
3University of São Paulo, Department of Electrical Engineering, São Carlos, Brazil
E-mail: *
Received May 5, 2011; revised June 7, 2011; accepted June 20, 2011
In view of the current and predictable energy shortage and environmental concerns, the exploitation of re-
newable energy sources offers great potential to meet increasing energy demands and to decrease depend-
ence on fossil fuels. However, introducing these sources will be more attractive provided they operate in
conjunction with energy storage systems (ESS). Furthermore, effective energy storage management is essen-
tial to achieve a balance between power quality, efficiency, costs and environmental constraints. This paper
presents a method based on the analytic hierarchy process and fuzzy multi-rules and multi-sets. By exploit-
ing a multiple criteria analysis, the proposed methods evaluate the operation of storage energy systems such
as: pumped hydro and compressed air energy storage, H2, flywheel, super-capacitors and lithium-ion storage
as well as NaS advanced batteries and VRB flow battery. The main objective of the study is to find the most
appropriate ESS consistent with a power quality priority. Several parameters are used for the investigation:
efficiency, load management, technical maturity, costs, environmental impact and power quality.
Keywords: Multiple Criteria Analysis, Fuzzy Sets, AHP Method, Energy Storage Systems, Power Quality
1. Introduction
In response to the energy crisis and related pollution prob-
lems, it is necessary to adopt new means of generating en-
ergy that use renewable sources and storage technologies in
an efficient and environmentally friendly manner. Renew-
able hybrid systems receive governmental incentives for
their development in many countries. One remarkable ex-
ample of this tendency occurs in Brazil. The Federal Gov-
ernment recently approved Decree 10438/02, which creates
the Incentive Program for Alternative Sources of Energy -
Therefore, considering the future scarcity of fossil fuels
and the environmental damage caused by them, it is incon-
testable that the use of hybrid renewable sources of energy
is the best choice for providing electricity to end-users [2].
However, an effective method for energy storage manage-
ment is essential to guarantee the expansion of the use of
hybrid energy sources [3]. This method must be able to
deal with some trade-offs betw een power quality, costs and
environmental constraints. Accordingly, it is important to
select a multiple criteria method that best satisfies the
management needs [4]. Today, several authors have
achieved exce llent results by using a large number of mul-
tiple criteria analyses in energy management. ELECTRE
[5], PROMETHEE [6], MACBETH [7], AHP [8] and also
Fuzzy sets [9] are some familiar tools. This paper presents
a method based on the routine developed by Saaty in the
AHP method and on multi-rules-based decisions and
multi-set considerations applied with fuzzy logic. To ana-
lyze the ESS operation, this paper uses AHP and fuzzy
logic and then compares and evaluates the final results.
The study focuses on the operational evaluation of
some ESS—compressed air energy storage (CA), pump-
ed hydro energy storage (PH), H2 storage, flywheel and
super-capacitor (SUP), lithium-ion and NaS advanced
batteries and VRB f low battery storage. Several cr iter ia are
used in this analysis: efficiency, load management, techni-
cal maturity, co sts, environmen tal impac ts and power qua l-
ity. In addition, the analysis of these criter ia also considers
transit and end-use ride-through, load leveling, load fol-
lowing, spinning reserve and transmission and distribution
stabilization. Thus, the paper has a focal point on power
quality management needs, but also evaluates costs and
environmental concerns. The paper is organized as follows:
Section 2 introduces the application’s main characteristics
and the purpose of the methods presented for ESS selection.
Section 3 presents the previous classification defined by the
selected criteria, according to the scen arios under analysis.
Sections 4 and 5 introduce the main concepts of AHP and
fuzzy logic, respectively. Subsequently, the application of
AHP and fuzzy logic is outlined in Section 6, in addition to
a comparison of the final results for both methods, con-
cerning the EES selection. Concluding remarks are given
in Section 7.
2. General Aspects of the Proposed Analysis
To develop the proposed method, eight types of ESS are
analyzed according to six criteria. The main objective of
this study is to find the most appropriate type of ESS to
be used in the power quality scenario. Accordingly, the
availability for use of any of the selected EES is consid-
ered for the specific region under analysis. A brief de-
scription of the ESS evaluated in this paper [10-12] is
presented below.
The CA storage system works by compressing air (CA
operates like a gas turb ine at high temperatures). CA has
a significant technical feasibility and it is considered on e
of the most efficient engines for converting heat into
electrical power, presenting an economically attractive
system for load management. In CA, the energy is stored
by compressing air via electrical compressors in huge
storage reservoirs that usually already exist, such as un-
derground caves, abandoned hard-rock mines, or natural
aquifers. CA offers good efficiency and has a high tech-
nical maturity. On the other hand, the need for a fuel in
the combustion process (usually natural gas) increases
greenhouse gas em i ssi ons.
Pumped hydro energy storage (PH) is an ESS based on
conventional hydroelectric technology. PH uses the po-
tential energy of water, transferred by pumps (charging
mode, during off-peaks) and turbines (discharge mode,
during peaks) between two reservoirs located at different
altitudes. The amount of stored energy is proportional to
the height difference between the upper and lower reser-
voirs and the volume of water stored. PH may be consid-
ered extremely expensive in terms of initial costs (install-
lation), it also requires suitable sites and there are long
lead-times for construction. The efficiency is usually
determined by the efficiency of the pumps and turbines
Although H2 storage is an immature technology, it is
seen as a promising means of electrochemical storage,
attracting huge interest. Since hydrogen is not a primary
energy source, its energy storage is based on an electro-
lyzer to split water into hydrogen and oxygen. Most as-
pects in the hydrogen-related technology, including gen-
eration, storage and utilization in fuel cells, still need
further development to be employed on a large scale.
A flywheel can accumulate and store mechanical en-
ergy in kinetic form. The stored energy depends on the
inertia and speed of the rotating mass (rotor). The fly-
wheel is a rotor placed inside a vacuum to eliminate fric-
tion-loss from the air and mounted on bearings for a sta-
bile operatio n. A f lywheel offer s high dens ity energ y and
high efficiency.
Super-Capacitors store energy by means of separating
the charge into two facing plates. They use polarized
liquid layers at the interface between a conducting ionic
electrolyte and a conducting electrode, increasing the
capacitance. Super-Capacitors offer high efficiency and
high costs.
Batteries are the most common devices used to store
electrical energy. Traditionally, they have been used
mainly for small scale applications. However, due to the
liberalization of electricity markets, battery manufactur-
ers have begun to recognize some potential applications
for larger scale energy storage applications. This paper
evaluates the sodium-sulfur (NaS) and lithium-ion ad-
vanced battery techn ol o gies.
Flow Batteries, also known as Regenerative Fuel Cells
or Redox Flow Systems, are a new class of battery that
has been achieving substantial progress - technically and
commercially. Flow Batteries present some features that
make them especially attractive for utility-scale applica-
tions. The operational principle differs from classical
batteries. The latter store energy both in the electrolyte
and the electrodes, while flow batteries store and release
energy using a reversible reaction between two electro-
lyte solutions, separated by an ion-permeable membrane.
Both electrolytes are stored separately in bulk storage
tanks, the size of which defines the energy capacity of
the storage system. The power rating is determined by
the cell stack. Therefore, the power and energy rating are
decoupled, which provides the system designer with an
extra degree of freedom when structuring the system.
This paper evaluates the vanadium redox (VRB) flow
A description of the criteria evaluated by the proposed
method is described below. It is important to observe that
the definition and evaluation of both quantitative and
qualitative criteria must take into accou nt an actual data-
base and management needs for each specific case. After
analyzing these aspects, it is possible to arrange the
method for the ESS selectio n.
The qualitative criteria are expressed through weights
stipulated by the decision maker—a group of researchers
from The Federal University of Santa Maria—in the in-
tervals from 0 to 1.0, with 1.0 being the highest score.
Copyright © 2011 SciRes. EPE
These weights are defined according to the analysis of
the actual database, taking into account social, political
and economic aspects related to the particular region
under analysis, e.g. EES installation in a specific region
of Brazil. In addition, the experience of the selected de-
cision makers is another key aspect in determining the
weights. The qualitative criteria considered in this work
load management (LM) related to load leveling and
load following;
technical maturity (TM);
environmental impacts related to visual and biologi-
cal impacts and greenhouse gas emissions;
power quality (PQ) related to transit and end-use
ride-through, spinning reserve and transmission and
distribution stabilizatio n.
The quantitative criteria are expressed through rated
data. The quantitative criteria evaluated in this study are:
efficiency (EF) in %;
costs represented in US$/kW.
As mentioned previously, the simulations will also
consider fast and conventional spinning reserve (FRSR -
CRSR), transit and end-use ride-through (T&ER), and
transmission and distribution stabilization (T&D) for the
power quality parameter. In addition, they consider load
leveling (LL) and load following (LF) for load manage-
ment parameter. Accordingly, it is i mportant to in troduce
some basic concepts regarding these aspects [11].
The fast response spinning reserve category corre-
sponds to the fast generation capacity response that is in
a state of ‘hot-stand-by’. Utilities hold it back in case of
a failure of generation units. Thus, the required power
output for this application is typically determined by the
power output of the largest unit operating on-grid. The
conventional spinning reserve requires a lesser ‘fast as
possible’ response. Storage systems can provide this ap-
plication in competition with standard generation facili-
Transit and end-use ride-through are applications re-
quiring very short durations combined with very fast
response times. They cover electric transit systems with
remarkable load fluctuations and customer power ser-
vices such as voltage stabilization and frequency regula-
tion to preven t events that can affect sensitiv e processing
equipment and can cause data and production losses.
Transmission and distribution stabilization are appli-
cations that require very high power ratings for short
durations in order to keep all components on a transmis-
sion or distribution line in synchronous operation.
Traditional load leveling is a widespread application
for large energy storages, in which cheap electricity is
used during off-peak hours for charging, while discharg-
ing takes place during peak hours, provid ing cost savings
for the operator.
In the case of ramping and load following, energy
storage is used to assist generation to follow the load
changes. An instantaneous match between generation
and load is necessary to maintain the generator rotational
speed and hence the frequency of the system.
The main characteristics of the ESS under analysis
[11,12] according to Power Quality (PQ) and Load Man-
agement (LM) are presented in Table 1.
The selected criteria analyzed in this section are now
presented in Table 2. The rated data described in the lit-
erature [10-12] are used to determine the values of the
quantitative criteria— efficien cy an d co sts. Th e qualita tiv e
criteria—load management, technical maturity, environ-
mental impacts and power quality—are represented by
weights stipulated by the decision makers in the inter vals
from 0 to 1.0, with 1.0 b ei ng the highe s t sco re.
3. Classification of the Criteria
The scenarios simulated in this stud y are evaluated by the
prior classification of the criteria. This classification is
therefore used in both AHP and fuzzy simulations. This
classification was thus created according to the different
relevance observed among these factors. The purpose of
Table 1. ESS—Power Quality (PQ) and Load Management
(LM) characteristics.
LF --X - - X X X
FRSR --X X - X X X
Back UpX-X X - X X X
T&D --X - X X X X
T&ER --X X X X X X
Table 2. Data used in AHP and fuzzy simulations.
EF 75 80 55 90 95 90 80 85
Costs 450 7501200400 3000 1500 20002500
LM 0.65 0.600.800.40 0.25 0.80 0.800.80
TM 0.85 0.850.500.80 0.70 0.75 0.650.60
Impacts 0.55 0.750.800.90 0.90 0.80 0.750.70
PQ 0.40 0.400.850.80 0.65 0.85 0.850.85
Copyright © 2011 SciRes. EPE
the proposed arrangement was to facilitate the develop-
ment of the simulation steps and facilitate understanding
of the method. Both these aspects are essential to cor-
roborate the final solutions aligned with the main problem
statement. The classification defined for each scenario is
presented below:
1) Power Quality Scenario: 1st power quality, 2nd effi-
ciency and load management, 3rd technical maturity, 4th
environment al impacts and 5th costs.
2) Costs Scenario: 1st costs, 2nd efficiency and power
quality, 3rd load management, 4th technical maturity and
5th environm ental impacts.
3) Environment Scenario: 1st environmental impacts,
2nd efficiency and load management, 3rd technical matur-
ity, 4th power quality and 5th costs.
4. AHP Analysis
The Analytic Hierarchy Process (AHP) was proposed by
Saaty [13]. As described previously, the AHP method is
based on pairwise comparisons. It uses a subjective as-
sessment of relative importance converted into a set of
overall scores (weights), classifying in this way the
struc ture of th e prob lem in a hierarchical way. According
to [14], the use of AHP as the single decision support
tool may be very problematic, largely because this
method can, in some cases, overlook the relationship
between values and judg ments. For this reason, the main
considerations observed in [14] and the consistency ratio
(C.R.) developed by Saaty [13] are verified in this study.
To find the C.R., AHP makes use of a consistency index
(C.I.) that prevents priorities from being accepted if the
inconsistency level is high. In order to measure the de-
viation of a pairwise matrix from “consistency”, the C.I.
is defined by
C.I. 1
n (1)
where λmax n is the dev iation of the judgments from the
consistency approximation .
A random index R.I. (of order n elements) is calcu-
lated as the average of the C.I. of many reciprocal matri-
ces randomly generated from the scale 1 to 9, with re-
ciprocals forced. The values of R.I. for matrices can be
found in [13]. The ratio of C.I. to R.I. for the same order
matrix is called the consistency ratio (C.R.). According
to Saaty, “a consistency ratio of 0.10 or less is consid-
ered acceptable”. That is, an inconsistency is stated to be
a matter of concern if C.R. exceeds 0.1, in which case the
pairwise comparisons should be reexamined.
It is important to emphasize that the fuzzy sets will be
applied in the same practical scenarios previously ana-
lyzed by AHP. Thus, it will provide a good comparison
of the consistency between the two methods.
Before one applies the AHP method, it is necessary to
select the criteria and the alternatives. In addition, it is
essential to consider an actual database for the specific
case under analysis. Subsequently, criteria and alterna-
tives can be placed into an AHP hierarchy, which is then
used to construct th e pairwise comparison matrix (PCM).
For this, the weights of the criteria need to be estimated.
This is done via measurement of AHP, which is based on
the theory defined by Saaty, as presented in Table 3.
Therefore, as shown in Table 4, one weight is as-
sumed for each pairwise comparison defined according
to the analysis in Table 2. In addition, five more tables
are constructed, considering in this way every criteria
and each final possible alternative (ESS). Later, as
shown in Table 5, one weight is assumed for the pair-
wise comparisons, taking into account only the criteria,
and with regard to the priority classification defined ear-
lier for the power quality scenario. This step results in
Table 3. Comparisons defined by saaty.
Weight Comparisons Explanation
1 Equal
Two activities contribute
equally to the objective
3 Moderate (weak)
Experience and judgment
slightly favor one activity
over another
5 Essential (strong)
Experience and judgment
strongly favor one activity
over another
7 Very strong
An activity is favored very
strongly over another; its
dominance demonstrated in
9 Absolute (extreme)
The evidence favoring one
activity over another is of the
highest possible order of
Table 4. PCM: Alternative × alternative—power quality
Criterion 6: Power Quality—C.R. = 0.0399
CAES1. 0.14 0.14 0.140.02
PHS1. 0.14 0.14 0.140.02
H27. 1.00 1.00 1.000.17
FLY5. 1.00 1.00 1.000.17
SUP5.005.000.330.331.00 0.33 0.33 0.330.07
LITH 1.00 0.33 0.330.14
NaS7. 3.00 1.00 1.000.20
VRB7. 3.00 1.00 1.000.20
Copyright © 2011 SciRes. EPE
the construction of two more tables, one for each sce-
Tables 4 and 5 present the consistency ratio (C.R.)
and the relative weights (RW) computed by the simula-
tion of AHP method considering each PCM. According
to Saaty, the results of C.R. must be less than 0.1 for the
analysis of more than five elements (criteria). To find the
RW, the data of each column are summed. Each single
data of the PCM is then divided by the sum of the col-
umn in which this data is placed. Later, the data resulting
from this division is summed to find the RW, consider-
ing each line of the matrix separately.
To find the classification of the ESS, the relative
weights (RW) presented in Table 4 are therefore multi-
plied by the relevant RW estimated in Table 5. The Final
Relative Weights (FRW) are described by the sum of
these multiplications. All the calculated FRWs repre-
senting the ESS classification are shown in Table 6, ac-
cording to AHP analysis.
5. Fuzzy Logic
Fuzzy logic was proposed by Zadeh. Fuzzy logic is con-
Tab le 5. PCM: Criterion × criterion—power quality scenario.
Scenario 1: Power QualityC.R. = 0.0551
EF 1.00 1.00 3.00 7.00 5.00 0.330.20
LM 1.00 1.00 3.00 7.00 5.00 0.330.20
TM 0.33 0.33 1.00 3.00 3.00 0.140.08
COST 0.14 0.14 0.33 1.00 0.33 0.110.03
LC 0.20 0.20 0.33 3.00 1.00 0.140.05
PQ 3.00 3.00 7.00 9.00 7.00 1.000.44
Table 6. ESS final classification in AHP—power quality
CAES 0.01 0.02 0.02 0.000.00 0.01 0.057th
PHS 0.01 0.02 0.02 0.000.00 0.01 0.055th
H2 0.00 0.05 0.00 0.000.00 0.09 0.148nd
FLY 0.04 0.01 0.02 0.010.01 0.07 0.167st
SUP 0.04 0.00 0.01 0.000.01 0.04 0.1033rd
LITH 0.04 0.04 0.01 0.000.01 0.06 0.164st
NaS 0.02 0.04 0.00 0.00 0.00 0.09 0.151nd
VRB 0.02 0.04 0.00 0.00 0.00 0.09 0.156
sidered one of the most powerful control methods en-
compassing many fields of application [15]. Fuzzy logic
is tested for the same case study by using MATLAB®
software, under multi-rules-based decisions and multi-set
The multi-rules-base used in this work consists of a
collection of if-then propositions. Using MATLAB®
software, the Mamdani method was applied in the fuzzy
inference process and the center of gravity method in the
defuzzification process [16].
A basic Mamdani fuzzy system accepts numbers as
input, then translates the input numbers into linguistic
terms such as low, medium, high (fuzzification). Rules
map the input linguistic terms into similar linguistic
terms describing the output [17]. Finally, the output lin-
guistic terms are translated into an output number (de-
fuzzification). The main idea of the Mamdani is to de-
scribe process states by means of linguistic variables and
to use these variables as inputs to control rules. The lin-
guistic terms are represented in fuzzy sets with a certain
shape. It is popular to use trapezoidal or triangular fuzzy
sets due to their computational efficiency [18].
The number of linguistic terms in each fuzzy set de-
termines the number of rules. In most applications, cer-
tain states can be neglected either because they are im-
possible or because a control action would not be helpful.
It is therefore sufficient to write rules that cover only
parts of the state space. Definition of linguistic variables
and rules constitute the main design steps when imple-
menting a Mamdani controller. In addition, an appropri-
ate classification of the parameters is essential to cor-
roborate the outcome of the fuzzy method.
The choice of Mamdani controller relates to the fol-
lowing aspects [19]:
it is suitable for engineering systems because its in-
puts and outputs are real-valued varia bles;
it provides a natural framework for incorporating
fuzzy rules from human experts;
there is much freedom in the choices of fuzzifier,
fuzzy inference engine, and defuzzifier;
it provides an effective framework in which to inte-
grate numerical and linguistic in formation.
Regarding the defuzzification process, there are sev-
eral choices to be made and many different methods have
been proposed [20]. Th is study used the so-called Center
of Area (COA) or Center of Gravity (COG) method. This
method chooses the control action that corresponds to the
center of the area with membership greater than zero.
The area is weighted with the value of the membership
function. The solution is a compromise, due to the
fuzziness of the consequences. The choice for COG is
justified because the use of this method is advisable not
only for quantitative but also for qualitative analysis.
Copyright © 2011 SciRes. EPE
Copyright © 2011 SciRes. EPE
To simulate the fuzzy analysis, it is essential to share
the six criteria described in two sub-groups, quantitative
and qualitative criteria, as already explained in section 2.
The data and the weights used to evaluate these criteria
were presented in Table 2. Qualitative criteria are ex-
pressed through weights (defin ed by the selected special-
ists in intervals from 0 to 1.0) to be applied in the fuzzy
The multi-sets that characterize each criterion are dis-
played in Figure 1. The number of membership func-
tions used in each criterion of the fuzzy set and the mul-
ti-rules is determined according to the relevance criteria
(section 3). In addition, bo th shape and position of fuzzy
set are chosen taking into account the need of each
analysis criterion for this study case.
The final classification of the ESS is presented in Ta-
ble 7. It is calculated using fuzzy logic and is associated
with the power quality scenario.
Figure 1. Fuzzy sets for each criterion—ESS analysis.
Copyright © 2011 SciRes. EPE
6. Results and Comparison: AHP and Fuzzy
Table 7 presents the final relative weigh ts and the classi-
fication according to the results achieved by AHP and
fuzzy logic, regarding the power quality scenario.
By observing the data presented in Table 7, it can be
seen that the ESS classification is the same, whether
compared with the AHP method or fuzzy logic outcome.
These results corroborate the use of both methods for the
analysis of the main characteristics of ESS. Accordingly,
the flywheel and the lithium-ion battery are the most
appropriate choices for the power quality scenario. Fur-
thermore, the environmental and costs scenarios were
also evaluated by the AHP method and fuzzy logic. As
predicted, the most appropriate choices for these scenar-
ios computed by the two methods were the same. In
these analyses, the Flywheel was selected again for both
environmental and costs scenarios. The final results are
presented in Table 8, according to the three situations
It is important to emphasize that this study may con-
sider several criteria and scenarios. This can be done
simply by evaluating and changing the AHP and the fuzzy
method (sets and rules) for each case under analysis.
Table 7. Final classification in AHP and Fuzzy—power
quality scenario.
AHP Fuzzy
CAES 0.057 th
4 0.381 th
PHS 0.055 th
4 0.383 th
H2 0.148 nd
2 0.751 nd
FLY 0.167 st
1 0.757 st
SUPERC 0.103 3rd 0.528 3rd
LITH 0.164 st
1 0.756 st
NaS 0.151 nd
2 0.751 nd
VRB 0.156 nd
2 0.751 nd
Table 8. Most appropriate ESS in AHP and Fuzzy—all
scenarios in analysis.
AHP Fuzzy
Power Quality 0.167 FLY 0.757 FLY
Power Quality 0.164 LIT H 0.756 LITH
Costs 0. 201 FLY 0.863 FLY
Environment 0.145 FLY 0.741 FLY
7. Conclusions
This paper presented a study for finding an appropriate
ESS, by evaluating its key operational characteristics in
the contex t of th e power qu ality scenario. To achieve this
objective, the AHP and fuzzy logic related to quantitative
and qualitative criteria were used. During the AHP
simulation, the considerations observed in [14] were
verified. It may be concluded that the AHP method re-
spects the relationship between values and judgments in
the majority of analyses, considering the previous set of
decision makers’ weights and the final results obtained
by the simulations.
A prior classification of criteria was defined in relation
to each scenario. This arrangement facilitated the devel-
opment of the simulation steps and led to better under-
standing of the method. The final results offered the
same ESS classification for both AHP method and fuzzy
logic. These outcomes corroborate the effectiveness of
AHP and fuzzy logic, and also validate the use of the
prior classification of the criteria for both methods. In
addition, they confirm that the relationship between val-
ues and judgments is respected by the AHP analysis in
this case study.
The outcome achieved by both methods for the power
quality scenario—flywh eel and the lithium-ion battery as
the most appropriate choices—is corroborated because
these technologies support the Quality and Load Man-
agement characteristics as a whole, such as load leveling,
fast response spinning reserve, conventional spinning
reserve, transmission and distribution stabilization,
among others.
Regarding ESS selection for costs and environment
scenarios, the flywheel is selected again as the most ap-
propriate technology. It is easily justified because the
flywheel offers high density energy, high efficiency, high
life cycle, low costs and it does not entail any kind of
negative environmental impact.
To summarize, this paper presents essential aspects of
the potential of storage energy systems for the improve-
ment of system management, taking into account not
only power quality, but also costs and environmental
8. Acknowledgements
The authors would like to thank CAPES for its financial
9. References
[1] National Regulatory Agency of Electrical Energy, ANEEL
Incentive Program for Alternative Sources of Energy,
Law 10.438/2002.
[2] A. Farret and M. G. Simoes, “Integration of Alternative
Sources of Energy,” John Wiley & Sons, Hoboken, 2006.
[3] E. Løken, “Use of Multicriteria Decision Analysis Meth-
ods for Energy Planning Problems,” Renewable and Sus-
tainable Energy Reviews, Vol. 11, No. 7, 2007, pp. 1584-
1595. doi:10.1016/j.rser.2005.11.005
[4] C. Zopounidis and M. Doumpos, “Multicriteria Classifi-
cation and Sorting Methods: A Literature Review,” Euro-
pean Journal of Operational Research, Vol. 138, No. 2,
2002, pp. 229-246. doi:10.1016/S0377-2217(01)00243-0
[5] M. Beccali, M. Cellura and M. Mistretta, “Decision-
Making in Energy Planning-Application of the ELECTRE
Method at Regional Level for the Diffusion of Renewable
Energy Technology,” Renewable Energy, Vol. 28, No. 13,
2003, pp. 2063-2087.
[6] M. Behzadian, R. B. Kazemzadeh, A. Albadvi and M.
Aghdasi, “PROMETHEE: A Comprehensive Literature
Review on Methodologies and Applications,” European
Journal of Operational Research, Vol. 200, No. 1, 2010,
pp. 198-215. doi:10.1016/j.ejor.2009.01.021
[7] F. Montignac, I. Noirot and S. Chaudourne, “Multi-Cri-
teria Evaluation of on-Board Hydrogen Storage Tech-
nologies Using the MACBETH Approach,” International
Journal of Hydrogen Energy, Vol. 34, No. 10, 2009, pp.
4561-4568. doi:10.1016/j.ijhydene.2008.09.098
[8] B. A. Akash, R. Mamlook and M. S. Mohsen, “Multi-
Criteria Selection of Electric Power Plants Using Ana-
lytical Hierarchy Process,” Electrical Power System Re-
search, Vol. 52, No. 1, 1999, pp. 29-35.
[9] I. J. Ramírez-Rosado and J. A. Domínguez-Navarro,
“Possibilistic Model Based on Fuzzy Sets for the Mul-
tiobjective Optimal Planning of Electric Power Distribu-
tion Networks,” IEEE Transactions on Power System,
Vol. 19, No. 4, 2004, pp. 1801-1810.
[10] W. Clark and W. Isherwood, “Distributed Generation;
Remote Power Systems with Advanced Storage Tech-
nologies,” Energy Policy, Vol. 32, No. 14, 2004, pp. 1573-
1589. doi:10.1016/S0301-4215(03)00017-X
[11] The Electricity Storage Association, Technology Com-
parisons, Electricity Storage Association, 2003.
[12] A. Gonzalez, B. Gallachóir, E. McKeogh and K. Lynch,
“Study of Electricity Storage Technologies and Their Po-
tential to Address Wind Energy Inter. in Ireland,” 2004.
[13] T. L. Saaty, “The Analytic Hierarchy Process,” McGraw-
Hill, New York, 1980.
[14] A. Bana e Costa and J. Vansnick, “A Critical Analysis of
the Eigenvalue Method Used to Derive Priorities in
AHP,” European Journal of Operational Research, Vol.
187, No. 1, 2008, pp. 1422-1428.
[15] L. A. Zadeh, “Fuzzy Logic,” University of California,
Berkeley, 1988.
[16] W. Pedrycz, F. Gomide, “An Introduction to Fuzzy Sets:
Analysis and Design,” The MIT Press, Cambridge, 1998.
[17] W. Siler and J. Buckley, “Fuzzy Expert Systems and
Fuzzy Reasoning,” John Wiley & Sons, Hoboken, 2005.
[18] H. Zimmermann, “Fuzzy Set Theory and Its Applica-
tions,” Springer, Berlin, 2001.
[19] L.-X. Wang, “Adaptive Fuzzy Systems and Control: De-
sign and Stability Analysis,” Prentice Hall, Upper Saddle
River, 1993.
[20] L. C. Barros and R. C. Bassanezi, “Tópicos de Lógica
Fuzzy e Biomatemática,” Unicamp-Imecc, Campinas,
Copyright © 2011 SciRes. EPE