- Email: [email protected]
A fuzzy multi-criteria decision making approach to assess building energy performance
Accepted Manuscript Title: A fuzzy multi-criteria decision making approach to assess building energy performance Author: Mehmet Kabak Erkan K¨ose O˘guzhan Kırılmaz Serhat Burmao˘glu PII: DOI: Reference:
S0378-7788(14)00008-5 http://dx.doi.org/doi:10.1016/j.enbuild.2013.12.059 ENB 4751
To appear in:
ENB
Received date: Revised date: Accepted date:
8-10-2013 20-12-2013 27-12-2013
Please cite this article as: M. Kabak, E. K¨ose, O. Kirilmaz, S. Burmao˘glu, A fuzzy multicriteria decision making approach to assess building energy performance, Energy and Buildings (2014), http://dx.doi.org/10.1016/j.enbuild.2013.12.059 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ip t
A fuzzy multi-criteria decision making approach to assess building energy performance
cr
Mehmet Kabak**,* Department of Industrial and System Engineering, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey.
us
Erkan Köse Defense Science Institute, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey.
an
Oğuzhan Kırılmaz Department of Industrial and System Engineering, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey. Serhat Burmaoğlu Leadership R&D and Application Center, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey.
M
ABSTRACT
Ac ce p
te
d
Due to an increasing demand for energy and rising energy prices, efficiency in energy consumption is fastbecoming a topic of significance. The building and construction sector has seen an increase of approximately 30 to 40% of overall energy consumption occurred; this has exceeded other major sectors such as industry and transport. Given the number of buildings and the cost of energy required to support these buildings, the developing of new approaches in the construction sector will be likely. This situation forces the various stakeholders to implement energy rating procedures to assess buildings’ energy performance. The most commonly utilized building environment assessment method currently used in Europe is the Building Research Establishment Environmental Assessment Method (BREEAM). Parallel to Europe, Turkey started its National Building Energy Performance Calculation Methodology (BEP-TR) in 2010. BREEAM and BEP-TR like other methods, require a lot of detailed and particular information in order to be implemented, and the procedure is fairly complicated. In addition, decision support systems can involve assessments, developed as a result of imprecise data in a qualitative manner. “Fuzzy set theory” can play a significant role in this kind of decision-making situation. This paper examines a “fuzzy multi criteria decision making (MCDM)” approach in order to analyze BEP-TR. This approach was applied to categorize alternative buildings according to their overall energy performance. Results are discussed in terms of developing a new and practical building rating system. Keywords: Fuzzy, Analytic Network Process, Energy performance assessment, BEP-TR
* Corresponding author: [email protected] , Tel: +90 312 4175190 ext. 4413.
Page 1 of 26
Page 2 of 26
d
te
Ac ce p us
an
M
cr
ip t
1. Introduction Expanding human population, growing complexities of civilization, economic growth, industry, and other related
ip t
issues are forcing countries to demand more energy sources. The fact that the world has limited resources requires countries to study energy and energy policies in detail. As a result of fossil fuel formation over the course of millions of
cr
years, alternative energy sources such as wind, biofuels, solar thermal and photovoltaic sources, are increasingly being considered as fuel sources. Another consideration in meeting this increased energy demand is to reduce energy
us
consumption by improving energy efficiency.
Energy consumption of the building and construction sector accounts for around 30-40% of total primary energy use
an
worldwide; consequently this sector contributes a great deal to greenhouse emissions and global warming [1]. In addition, this sector accounts for the use of approximately 40% of the world’s natural resources extracted in industrialized countries; 12% of available potable water, and the production of 45-65% of the waste later sent to landfills
M
[2]. Due to the projected increase of the global population from 6.5 billion in 2005 to approximately 9.0 billion in 2035, these figures are expected to continue to grow, [3]. In light of this consideration, organizations have been investing
d
significant resources to create sustainably built environments, stressing sustainable building renovation processes to
te
reduce energy consumption and carbon dioxide emissions [4]. Those who design and operate buildings need methods to evaluate the environmental impacts of their actions.
Ac ce p
Various rating methods, referred to as building rating systems (BRS), have been developed as important tools in measuring and evaluating the environmental performance of a building. The Building Research Establishment Environmental Assessment Method (BREEAM) and Leadership in Energy and Environmental Design (LEED) are the two most widely recognized environmental assessment methodologies used globally in the building and construction industry today.
BREEAM was launched in the UK in 1990 to provide an environmental assessment and labeling mechanism for buildings [5,6]. There are nine different sections/criteria, including management, health & well-being, energy, transport, water, materials, waste, land use & ecology, and pollution. There are also sub-criteria (not mentioned in this paper). Each criterion consists of a number of issues; each issue is scored and each criterion is weighted for the overall evaluation of a
Page 3 of 26
building. The building is then categorized and labeled according to its score. Table 1 shows the corresponding scores for each category [7]. The US Green Building Council (USGBC) formed its major green building code known as LEED in 2000 [8]. LEED
ip t
provides building owners and operators with a framework for identifying and implementing practical and measurable green building design, construction, operations and maintenance solutions. It validates that a building is designed and
cr
constructed to achieve high performance in key areas of human and environmental health. Buildings are scored according to the criteria and weights shown in Table 1 [9]. Extra “innovation” credit can be applied for instances where a
us
design can demonstrate a reduction in a building’s impact on the environment in an innovative way.
The individual credits are added up and then weighted in line with the Table 1 to give a final rating. The building
an
must meets all prerequisites and can achieve the minimum number of points necessary to earn the certified level. The certification levels for BREEAM and LEED are explained in Table 2 [7, 9].
M
d
Although the BREEAM and LEED systems are more commonly applied in industry, there are other national rating
te
systems. The Taiwan Architecture and Building Research Institute developed a similar system named “Green Building Evaluation and Labeling System” (GBELS). GBELS uses nine criteria including biodiversity, greenery, soil water
Ac ce p
content, daily energy savings, carbon dioxide emission reduction, waste reduction, indoor environment, water resource, and sewage and garbage improvement [10]. The Green Building Tool (GBT) is an evolving assessment system sponsored by National Resources Canada. Comprehensive Assessment System for Building Environmental Efficiency (CasBee) in Japan, Minergie in Switzerland, National Australian Built Environment Rating System (NABERS) in Australia, DGNB (Deutsche Gesellschaft für Nachhaltiges Bauen) in Germany, Green Rating for Integrated Habitat Assessment (GRIHA) in India are examples for BRS [11]. These methods used in different countries require expertise to assess the energy performance of buildings and the process is generally performed by external experts, resulting in a cost for each assessment. The originality of this paper lies in the way the owner-occupiers’ viewpoint is included in the assessment process. These stake-holders can easily apply the model proposed herein to assess the energy performance of a building without the help of the external experts.
Page 4 of 26
In addition, as seen in the process of BREEAM, LEEDS, BEP-TR and other evaluation systems, energy performance of buildings consist of multiple criteria.
ip t
Multi-criteria decision making (MCDM) comprises a finite set of alternatives, among which the decisions makers (DMs) must select, evaluate or rank according to the weights of a finite set of criteria. The multi-criteria nature of the
cr
buildings' energy performance assessment problem makes MCDM methods ideal to cope with the complexity of the problem. DMs consider many criteria simultaneously, with various weights, then evaluate the alternatives. Some
us
methods apply to the evaluation of qualitative criteria while others are suitable for quantitative criteria. But the buildings' energy performance assessment problem requires both qualitative and quantitative criteria. These methods also involve
an
subjective assessments, resulting with imprecise data in qualitative manner. Due to the availability and uncertainty of information in the decision process, as well as the ambiguities of human feeling and recognition, it is often difficult to make an exact evaluation and convey the feeling and recognition of objects for DMs. “Fuzzy set theory” can play a
M
significant role in this kind of decision situation [12, 13]. Hence since DMs generally fail to make a good numerical prediction for criteria, evaluation is expressed in linguistic terms. Fuzzy linguistic models permit the translation of verbal
d
expressions into numerical ones. Thereby, when dealing quantitatively with imprecision in the expression of the impor-
te
tance of each criterion, some multi-criteria methods based on fuzzy relations are used [14]. As a result, a MCDM method based on fuzzy logic is proposed to assess practically the energy performance of buildings. For this aim, National
Ac ce p
Building Energy Performance Calculation Methodology of Turkey (BEP-TR) is analyzed as a MCDM problem. There are seven criteria: location and climate data, geometrical shape, building envelope, mechanical systems, lighting system, hot water system and renewable energy & cogeneration in BEP-TR [15]. Since there is a set of criteria which affect each other in the BRS, the Analytic Network Process (ANP), as a wellknown MCDM method, is better able to analyze the impact of each criterion on every other criterion using pair wise comparisons [16]. It should be noted that experts who participate in this evaluation process cannot always explain their judgments about certain attributes such as quality or performance with distinct and discreet scales. The fuzzy set theory enables them to explain their evaluations in linguistic terms. This problem is further analyzed by using ANP based on the fuzzy set theory in this paper. The remainder of this paper is organized as follows: Section 2 describes buildings’ energy performance in Turkey. Main characteristics of the fuzzy set theory and fuzzy ANP (FANP) method are detailed in Sections 3 and 4,
Page 5 of 26
respectively. The steps of the proposed FANP model are explained and the application for BEP-TR is detailed in Section 5. Finally, the obtained results, conclusions and future research directions are discussed in Section 6.
ip t
2. Building energy performance in Turkey Scoring systems which use their own criteria and weights vary according to geography, country policy, people’s
cr
awareness about environment, and so forth, and thus have some differences. Prior to 2010, Turkey did not have its own voluntary rating system. On the other hand, the international BRS, BREEAM and LEED in particular were received with
experience in reducing energy consumption in buildings to Turkey.
us
interest by stakeholders of the Turkish construction industry. These stakeholders intend to bring other countries’
an
Legislation in Turkey on energy performance in buildings, in parallel with EPBD (Energy Performance of Buildings Directive), was announced by the Ministry of Public Works and Settlement on 5 December 2008, after consultation with
M
relevant ministries and universities. The Turkish government intends to use this method as a driving force towards green construction and start a transformation in the building and construction industry. In 2009, various international workshops were organized to exchange ideas and share experiences, discuss common problems and local solutions with
d
academics,, and energy experts. Using European standards, their adaptation to local conditions and their contribution to
te
further developments were discussed. BEP-TR was prepared in accordance with EU Building Energy Performance Directive (2002/91/EC) with respect to Turkey’s conditions [17]. BEP-TR is a set of criteria which is designed to guide
Ac ce p
the building and construction sector towards energy efficiency. In the evaluation process of BEP-TR, a reference building is identified by experts. Energy performance of the building and reference one are calculated based on seven criteria labeled in Table 3 [15]. A detailed depiction of how each section value is calculated by using thresholds, formulas require a complicated analysis which exceeds the borders of this paper. As a result, calculation of Energy Performance ( E p ) value only is explained through Eq.1. Energy class of the building is defined according to the E p score, as shown in Table 4 [15, 18].
E p = 100 * ( EPa / EPr )
(1)
E p : Energy performance
EP: Energy consumption per year ( kWh/m 2 ) a: Building for assessment r: Reference building
Page 6 of 26
BEP-TR mainly focuses on energy performance of buildings. BEP-TR is not an exact equivalent of BREEAM or
ip t
LEED, but it will define the path towards a green BRS in Turkey. Therefore its approach to sustainability, tools and units of assessment it employs and the improvements it requires will be crucial for the future of sustainability discussions in Turkey’s building and construction industry. It should be noted that weights for criteria in BREEAM and LEED, and not
cr
BEP-TR, are basically declared. In this paper, the authors calculate criteria weights for BEP-TR by using FANP,
us
evaluate and label three buildings in Ankara, Turkey. 3. The fuzzy set theory
an
Since decision makers (DMs) generally fail to make a good numerical prediction for criteria, evaluation is expressed in linguistic terms. Additionally, human judgment on qualitative attributes tends to be subjective and imprecise.
M
Therefore, the fuzzy set theory is commonly used in decision making problems. Fuzzy numbers expand on the idea of the confidence interval and are defined over a fuzzy subset of real numbers. A triangular fuzzy number (TFN) shown in Fig. 1 is a type of fuzzy number and, should possess the some basic properties [19, 20].
d
A fuzzy number M defined on ℜ is a TFN if its membership function μ M~ ( y) : ℜ → [0,1] is equal to
te
~
Ac ce p
⎧( y − l ) /( m − l ), l ≤ y ≤ m ⎪
μ M~ ( y ) = ⎨(u − y ) /(u − m), m ≤ y ≤ u ⎪0, ⎩
(2)
otherwise
where l, u, m are real numbers and l ≤ m ≤ u . The linguistic variable scale and the corresponding TFNs used in this study are shown in Table 5 [21].
In this paper, the fuzzy set theory is incorporated with ANP (Analytic Network Process) through an evaluation form that uses linguistic variables. The value of the linguistic variables that a DM has assigned to the pair-wise comparison between each two criteria is converted into TFN scores. The FANP methodology that is applied for this paper is detailed in next section.
Page 7 of 26
4. The FANP method The Analytic Network Process (ANP) allows for complex inter-relationships among decision levels and attributes
ip t
[16]. The ANP feedback approach replaces hierarchies (Fig. 2a) with networks (Fig. 2b) in which the relationships between levels cannot be easily represented as higher or lower, dominant or subordinate, direct or indirect [22].
cr
In ANP, the modeling process can be divided to three steps, which are described as follows [23]:
us
Step 1: The pair-wise comparisons and relative weight estimation
Before performing the pair-wise comparisons, all criteria and clusters compared are linked to each other. The pair-
an
wise comparisons are made depending on the scale shown on Table 5. The inconsistency ratio (IR) is useful for identifying possible errors in judgments as well as actual inconsistencies in the judgments themselves. For example, if A is more important than B and B is more important than C, C cannot be important than A. If the value of IR is smaller or
M
equal to 10%, the inconsistency is acceptable. If the IR is greater than 10%, we need to revise the subjective judgment. While doing pairwise comparisons, the inconsistency value is considered in all stages[25, 26]. In the pair-wise
d
comparison matrix, the score of a ij represents the relative importance of the component on row (i) over the component
te
on column (j), i.e., aij = wi / w j . The reciprocal value of the expression ( 1 / a ij ) is used when the component j is more important than the component i . Then, a local priority vector (eigenvector) w is computed as an estimate of the relative
Ac ce p
importance accompanied by the elements being compared by solving the following equation:
Aw = λmax w where
(3)
λmax is the largest eigenvalue of matrix A.
Step 2: Formation of the initial supermatrix
The obtained vectors are further normalized to represent the local weight vector. A supermatix is formed, after which local weight vectors are entered in the appropriate columns of the matrix of influence among the elements, to obtain global priorities. The supermatrix representation of a network with three levels is given as follows (Fig. 2b): G C 0 ⎛0 W = Criteria (C ) ⎜ ⎜W21 W22 Alternatives ( A) ⎜ ⎝ 0 W32 Goal (G )
A 0⎞ ⎟ 0⎟ Ι ⎟⎠
(4)
Page 8 of 26
W 21 is a vector that represents the impact of the goal on the criteria, W22 is a vector that represents impact of the
interdependences among criteria, W 32 is also a vector that represents the impact of criteria on each of alternatives, and I
ip t
is the identity matrix. Any zero value in the super-matrix can be replaced by a matrix if there is an interrelationship of elements within a cluster or between the clusters. Step 3: Formation of the weighted super-matrix
cr
An eigenvector is obtained from the pair-wise comparison matrix of the row clusters with respect to the column
us
cluster, which in turn yields an eigenvector for each column cluster. The first entry of the respective eigenvector for each column cluster is multiplied by all the elements in the first cluster of that column, the second by all the elements in the second cluster of that column, and so on. In this way, the cluster in each column of the super-matrix is weighted, and the
an
result, known as the weighted super-matrix, is stochastic [27]. A detailed discussion regarding the mathematical processes of the ANP is provided in different papers [15,21,28,29].
M
In the proposed methodology, the pairwise comparison matrices are formed, and FANP has been used to determine weights of BRS’s criteria. The FANP can easily accommodate the interrelationships existing among the functional
te
interrelationships.
d
activities. The concept of supermatrices is employed to obtain the composite weights that overcome the existing
Pairwise comparison matrices are structured by using TFNs (l, m, u). The fuzzy matrix can be given as follows:
Ac ce p
l m u l m u ⎛ (a11 , a11 , a11) (a12 , a12 , a12 ) ... (a1ln , a1mn , a1un ) ⎞ ⎟ ⎜ l m u l m u , a21 , a21) (a22 , a22 , a22 ) ... (a2l n , a2mn , a2un ) ⎟ ~ ⎜ (a21 A=⎜ ⎟ M M M M ⎟ ⎜ ⎜ l m u l m u l m u ⎟ ⎝ (am1, am1, am1 ) (am1, am1, am1 ) ... (amn, amn, amn) ⎠
(5)
The a mn represents the of comparison m (row) with component n (column). The pair-wise comparison matrix (Ã) is assumed as reciprocal. ⎛ (1,1,1) ⎜ ⎜ 1 1 ⎜( u , m a12 ~ ⎜ a12 A= ⎜ M ⎜ 1 1 ⎜( ⎜ au , am ⎝ 1n 1n
l m u (a12 , a12 , a12 ) ... ( a1ln , a1mn , a1un ) ⎞ ⎟ ⎟ 1 , l ) (1,1,1) ... ( a 2l n , a 2mn , a 2un ) ⎟ a12 ⎟ ⎟ M M M ⎟ 1 1 1 1 ⎟ , l ) ( u , m , l ) ... (1,1,1) ⎟ a1n a 2 n a 2 n a 2 n ⎠
(6)
Page 9 of 26
~ . The In this study, the logarithmic least squares method is used for getting estimates for fuzzy priorities w i logarithmic least squares method for calculating triangular fuzzy weights can be given as follows [27]:
)
k = 1,2,3,..., n
ip t
(
~ W = Wkl , Wkm , Wku where
(∏ a ) = ∑ (∏ a ) n j =1
n i =1
s 1/ n kj
m 1/ n n j =1 ij
, s ∈ {l , m, u}
cr
Wks
(7)
(8)
us
DMs tend to evaluate alternatives by using linguistic scales. Pairwise comparisons, rating that is preferred on the situation of high numbers of criteria and candidates or other ranking methods i.e. TOPSIS, PROMETHEE could be used
an
to evaluate the performance of the alternatives [30,31]. In this paper, rating is used; overall priorities for alternatives and crisp values are calculated by using matrix operations. Crisp values for fuzzy weights are calculated as follows [32]: (9)
M
Valuecrisp = (l + 2m + u) / 4
5. A real case application
d
Then, the alternative with the largest priority should be selected.
te
Turkey is energy importing country, whereby 70% of the total energy consumption supplied is by imported energy
Ac ce p
[33]. This situation forces the country to develop domestic energy resources and to reduce energy consumption by improving energy efficiency. The Turkish government has developed its own BRS as BEP-TR to be able to use it as a driving force towards green construction and start a transformation in the building and construction industry. BEP-TR focuses on energy use of buildings and categorizes them according to their energy consumption. It requires many complex, technical formulas, thresholds, and standards for every criterion in the calculation process. This paper analyzes criteria weights for BEP-TR by using FANP. As a real case application, three buildings in Ankara, Turkey are categorized according to their energy performance by using the proposed fuzzy MCDM approach. Firstly the subject matter experts (SMEs), those persons who will evaluate criteria weights in BEP-TR and case buildings, are determined. Two of the experts are academic personnel who have published papers on energy and green buildings; three are experts on energy affairs in the Turkish Ministry of Energy and Natural Resources, and the
Page 10 of 26
remainder are the authors of this paper. The case study of this research is taken from a housing estate located in Ankara, Turkey, which is in Climate Zone III according to the zoning in TS 825 [34]. The following steps are applied to determine the weights of the criteria in BEP-TR:
can be measured by the FANP technique. The structure established is shown in Fig. 3.
ip t
Step 1: The problem is converted into a hierarchical structure in order to transform the criteria into a state in which they
cr
In this structure, the aim is determining the criteria weights and ranking them according to the importance levels.
Step 2: After forming the decision hierarchy, the weights of criteria to be used in evaluation process are assigned by
us
using FANP method. In this phase, assuming that there is no dependence among the criteria, the experts are given the task of forming individual pairwise comparison matrix as shown in Table 6. Geometric means of these values are found
an
to obtain the pairwise compassion matrix on which there is a consensus. The pairwise comparison matrix is analyzed by ~ using Microsoft Excel software, and the fuzzy eigenvector (W21 ) is obtained.
d te
Ac ce p
~ W21
⎡C1 ⎤ ⎡(0.20, 0.24, 0.29)⎤ ⎢ ⎥ ⎢ ⎥ ⎢C 2 ⎥ ⎢(0.03, 0.03, 0.04) ⎥ ⎢C 3 ⎥ ⎢(0.07, 0.09, 0.11) ⎥ ⎢ ⎥ ⎢ ⎥ = ⎢C 4 ⎥ = ⎢(0.08, 0.09, 0.11) ⎥ ⎢C ⎥ ⎢(0.11, 0.13, 0.15) ⎥ ⎢ 5⎥ ⎢ ⎥ ⎢C 6 ⎥ ⎢(0.10, 0.11, 0.14) ⎥ ⎢ ⎥ ⎢(0.24, 0.31, 0.32) ⎥ ⎦ ⎣C 7 ⎦ ⎣
M
Step 3: Sections in BRS are generally not independent; therefore, inter-dependence among the criteria is determined by
analyzing the impact of each criterion on every other criterion using fuzzy pairwise comparisons. These dependencies are determined by experts on the basis of a group study. The arrow from “Building envelope” to “Geometrical shape” means that the “Building envelope” affects the “Geometrical shape” in the determination of building energy performance. The arrow between “Building envelope” and “Renewable energy & Cogeneration” means they affect each other in the process. The other relations between criteria are as shown in Fig. 4. Based on these inner dependencies, pairwise comparison matrices are formed for the factors. The following question, ‘‘what is the relative influence of Mechanical Systems when compared with Renewable energy & Cogeneration on controlling Geometrical shape?’’ may arise in pairwise comparisons. The resulting fuzzy eigenvectors are calculated and
Page 11 of 26
~ using the computed relative fuzzy importance weights, the inner dependence matrix of the criteria (W 22 ) is formed. The
value of (0, 0, 0) means, there is no relation between two criteria.
~ and W 21 as follows:
cr
us
Wcriteria(crisp )
⎡C1 ⎤ ⎡0.27⎤ ⎢ ⎥ ⎢ ⎥ ⎢C 2 ⎥ ⎢0.07⎥ ⎢C 3 ⎥ ⎢0.10⎥ ⎢ ⎥ ⎢ ⎥ = ⎢C 4 ⎥ = ⎢0.16⎥ ⎢C ⎥ ⎢0.12⎥ ⎢ 5⎥ ⎢ ⎥ ⎢C 6 ⎥ ⎢0.08⎥ ⎢ ⎥ ⎢0.20⎥ ⎦ ⎣C 7 ⎦ ⎣
an
~ ~ ~ Wcriteria = W22 * W21
⎡C1 ⎤ ⎡(0.15, 0.20, 0.25) ⎤ ⎢ ⎥ ⎢ ⎥ ⎢C 2 ⎥ ⎢(0.04, 0.05, 0.07) ⎥ ⎢C 3 ⎥ ⎢(0.05, 0.07, 0.09) ⎥ ⎢ ⎥ ⎢ ⎥ = ⎢C 4 ⎥ = ⎢(0.09, 0.12, 0.17)⎥ ⎢C ⎥ ⎢(0.07, 0.10, 0.11) ⎥ ⎢ 5⎥ ⎢ ⎥ ⎢C 6 ⎥ ⎢(0.04, 0.06, 0.08) ⎥ ⎢ ⎥ ⎢(0.12, 0.15, 0.20) ⎥ ⎦ ⎣C 7 ⎦ ⎣
ip t
Step 4: In this step, the interdependent fuzzy priorities of the criteria are calculated by using the data given in Tables 7
M
According to the results, the three most significant criteria are C1, C7, and C4, respectively. Significant differences
d
are observed in the results obtained for the criteria priorities when interdependent priorities of the criteria ( wcriteria ) and
te
dependencies are not taken into account. For example, the result for C1 changes from (0.20, 0.24, 0.29) to (0.15, 0.20, 0.25).
Ac ce p
Step 5: In this step, we calculate the performance value of the housing estates with respect to each fuzzy criteria. To
demonstrate, three buildings in Ankara, Turkey are evaluated using BEP-TR criteria. The first building (B1) is a typical residential building which was constructed in 2009. It serves as both a residential structure as well as an office (commercial) structure. The second, (B2), is a 35-year-old building and its health & well-being is unknonwn, but estimated to be inadequate. The last building (B3) was constructed in 2001 and is estimated to need some modifications. Using Microsoft Excel software, the fuzzy eigenvectors are computed by analyzing these matrices and the verbal evaluation values in Table 8. Finally, the overall performance value of the housing estate reflecting the interrelationships within the criteria are calculated as shown in Table 8. Once these performance values are normalized, categories are determined for the buildings. The FANP analysis results indicate that the B1, B2 and B3 are labelled as A, E and B, respectively.
Page 12 of 26
Two different methods, fuzzy AHP and fuzzy TOPSIS, are used for the same calculations to compare the results obtained with FANP. While the scores for buildings change, the labels remain same within the fuzzy AHP analysis. The difference in scores is an expected result based on AHP that does not take into account interrelationships within the
ip t
criteria. TOPSIS is based on the concept that the most preferred alternative should have the shortest distance from the positive ideal solution and have the longest distance from the negative ideal solution. Scores and categories are
different MCDM techniques.
us
cr
determined by using the fuzzy TOPSIS as shown in Table 8. Finally, similar conclusions are obtained by using the three
an
5.1. The sensitivity analysis
Sensitivity analysis of the proposed model leads to a much better understanding of the determining criteria for
M
building energy performances in Turkey. A sensitivity analysis based on criteria weights is performed in this section. For this aim, seven cases are created and the sensitiveness of criteria weights is investigated. When current weights of the
d
criteria are increased individually, the category of the building "B2" rises from E to D while others stay at the same
te
category. Decreasing criteria's weights results that the category for the building "B1" goes to B and stays same for others. For instance, values for the criterion C4 are increased as shown in Fig. 5. Values which are equal or greater than 0.19
Ac ce p
makes the categories of buildings B1, B2, and B3 as A, D, and B, respectively. On the other hand, values equal and less than 0.15 changes categories as B, E, B. The cases are summarized in Table 9.
6. Conclusion
This research illustrates how to make implement a more simplistic and easily-applied energy code, BEP-TR, for buildings, and shows how to develop the performance evaluation model and calculate the evaluation result by using a fuzzy MCDM approach. From the evaluation model’s result, construction companies can determine which part will be the priority criteria in order to improve the effectiveness and efficiency of their construction. By the algorithm of this study, the competition capability for construction companies could be increased. Following additional development and
Page 13 of 26
testing, the proof of the validity and applicability of such an energy code could support both the construction industry and Turkish government energy policies. BEP-TR focuses on buildings’ energy consumption and is not an exact equivalent of BREEAM or LEED. For this
ip t
reason, criteria weights can be evaluated by using the proposed method in further research and applied for determination of green building construction in Turkey. Additionally, BEP-TR will define the path towards a national green BRS.
cr
Revised BEP-TR. BRS must address sustainability and should go beyond the borders of energy, water and recyclables. Green building design practices must encompasses concepts, such as economic performance, contribution to
us
employment, protection of biodiversity, labor processes, security, safety, public health, education and cultural preservation.
an
In the future, this study could be used to develop a new national BRS. The obtained feedback should prove useful to improve the methodology and to build a software program for the BRS, and the role BRS will play in raising awareness
M
in the construction industry for a sustainable future.
d
References
(3) (2008) 394–398.
te
[1] L. Perez-Lombard, J. Ortiz, C. Pout, A review on buildings energy consumption information, Energy and Buildings 40
Ac ce p
[2] D. Castro-Lacouture, J.A. Sefair. L. Florez, A.L. Medaglia, Optimization model for the selection of materials using a LEED based green building rating system in Colombia, Build Environ 44 (2009) 1162–1170. [3] M.K. Dixit, J.L. Fernandez-Solis, S. Lavy, C.H. Culp, Identification of parameters for embodied energy measurement: a literature review, Energy Build 42 (2010) 1238–1247. [4] Y.K. Juan, P. Gao, J. Wang, A hybrid decision support system for sustainable office building renovation and energy performance improvement, Energy and Buildings 42 (3) (2010) 290–297. [5] R. Baldwin, A. Yates, N. Howard, S. Rao, BREEAM 98 for Offices (Building Research Establishment, Watfordt,1998). [6] V. Olgyay and J. Herdt, The application of ecosystems services criteria for green building assessment, Solar Energy 77 (2004) 389–398. [7] HRS
Services.
BREEAM
code
for
Sustainable
Homes
2008.
Available
from:
http://www.hrsservices.co.uk/pages/breeam/breeam_explained.html: April 2012.
Page 14 of 26
[8] J.H. Scofield, Do LEED-certified buildings save energy? Not really, Energy and Buildings 41 (12) (2009) 1386–1390. [9] USGBC.
LEED
for
existing
buildings
operations
and
maintenance
2009.
Available
from
http://www.usgbc.org/ShowFile.aspx?DocumentID=8876; April 2012.
ip t
[10]V.W.Y. Tam, The effectiveness of the green building evaluation and labeling system, Architectural Science Review 50 (4) (2007) 323-330.
cr
[11] M. Bayraktar, E. Kalaycioglu, A.Z. Yilmaz, Proceedings of Building Simulation 2011, in 12th Conference of
[12] L.A. Zadeh, Fuzzy sets. Information and Control 8 (1965) 338–353.
us
International Building Performance Simulation Association, (Sydney, 14–16 November 2011).
[13] G. Büyüközkan and G.Çifçi, A novel hybrid MCDM approach based on fuzzy DEMATEL, fuzzy ANP and fuzzy
an
TOPSIS to evaluate green suppliers, Expert Systems with Applications 39 (3) (2012) 3000-3011. [14] Z. Güngör, G. Serhadlıoğlu, and S.E. Kesen, A fuzzy AHP approach to personnel selection, Applied Soft Computing 9
M
(2009) 641-646.
[15] Energy Performance for Buildings. Regulation, calculation method, reference building & energy classification, software, process, Turkish Republic the Ministry of Public Works and Settlement 2009. Available from
d
http://www.csb.gov.tr/iller/ dosyalar/dosya/il_webmenu8900.pdf; March 2012.
Pittsburgh, 1996).
te
[16] T.L. Saaty, Decision making with dependence and feedback: The Analytic Network Process (RWS Publications,
Ac ce p
[17] National Building Energy Performance Calculation Methodology of Turkey. (No: YİG/2010-02)), Turkish Official Journal 2010.
[18] A.K. Yener, A method for the determination of building’s lighting energy performance: BEP-TR, in 10th National Plant Engineering Congress, (Izmir, 13-16 April 2011).
[19] G. Facchinetti and R.G. Ricci, A characterization of a general class of ranking functions on triangular fuzzy numbers, Fuzzy Sets and Systems 146 (2004) 297–312.
[20] M. Salehi and R. Tavakkoli-Moghaddam, Project selection by using a fuzzy TOPSIS technique, Engineering and Technology 30 (2008) 85–90.
[21] H.F. Lin, An application of fuzzy AHP for evaluating course website quality, Computers & Education 54 (2010) 877– 888.
Page 15 of 26
[22] T.L. Saaty and L.G. Vargas, Diagnosis with dependent symptoms: Bayes theorem and the analytic network process, Operations Research 46 (4) (1998) 491–502.
telecommunication company, Expert Systems with Applications 36 (2009) 3887-3895.
ip t
[23] S. Onut, S.S. Kara, E. Isık, Long term supplier selection using a combined fuzzy MCDM approach: A case study for a
textile firm, Information Sciences 177 (2007) 3364–3382.E.A.
cr
[24] I. Yüksel, I. and M. Dağdeviren, Using the analytic network process (ANP) in a SWOT analysis-A case study for a
[25] Demirtaş and Ö. Üstün, An integrated multiobjective decision making process for supplier selection and order
us
allocation, The International Journal of Management Science 36 (2008) 76-90.
[26] M.S. Ozdemir, Validity and inconsistency in the analytic hierarchy process, Applied Mathematics and Computation
an
161 (2005) 707-720.
[27] J. Ramik, A decision system using ANP and fuzzy inputs, in 12th international conference on the foundations and
M
applications of utility, Risk and Decision Theory, (Roma, 2006).
[28] R.P. Mohanty, R. Agarwal, A.K. Choudhury, M.K. Tiwari, A fuzzy ANP-based approach to R&D project selection: A
d
case study, International Journal of Production Research 43 (2005) 5199–5216.
te
[29] M. Kabak, S. Burmaoğlu, Y. Kazançoğlu, A fuzzy hybrid MCDM approach for professional selection, Expert Systems with Applications 39 (2012) 3516-3525.
Ac ce p
[30] M. Dağdeviren, A hybrid multi-criteria decision-making model for personnel selection in manufacturing systems, J Intell Manuf. 21 (2010) 451–460.
[31] M. Sevkli, A. Oztekin, O. Uysal, G. Torlak, A. Turkyilmaz, D. Delen, Development of a fuzzy ANP based SWOT analysis for the airline industry in Turkey, Expert Systems with Applications 39 (2012) 14–24. [32] R.D. Raut, Evaluation of supplier selection criteria by combination of AHP and fuzzy DEMATEL method, Int.J. Business Innovation and Research 5 (4) (2011) 359–392.
[33] İ. Kotçioğlu, Clean and sustainable energy policies in Turkey, Renewable and Sustainable Energy Reviews 15 (9) (2011) 5111–5119. [34] TSI. Thermal insulation requirements for buildings (Vol. TS 825). Ankara: Turkish Standards Institution, 2008.
Page 16 of 26
~
Fig. 1. A triangular fuzzy number, M
ip t
Fig. 2. Hierarchy and network: (a) hierarchy; (b) network. Fig 3. Schematic diagram of the proposed model for BRS.
cr
Fig. 4. Network structure of the BRS problem.
Ac ce p
te
d
M
an
us
Fig. 5. Weight values for the criterion C4
Page 17 of 26
μ M~
Ml(y)
Mr(y)
0.0 ~
Ac ce p
te
d
M
an
Fig. 1. A triangular fuzzy number, M
u
us
m
cr
M
l
ip t
1.0
Fig. 2. Hierarchy and network: (a) hierarchy; (b) network [23, 24].
Page 18 of 26
Determination of decision makers
ip t
Identify the Building Rating System’s factors
us
Approve the model?
Y
FANP
an
N
cr
Structure the decision model
Determine the weights of
~
M
the factors (W21 )
Determine inner dependence
~
d
matrix of the factors (W22 )
Ac ce p
te
Determine the interdependent weights of the factors
Evaluate and label the buildings
ig 3. Schematic diagram of the proposed model for BRS.
Page 19 of 26
ip t
GOAL Determining BRS criteria weights
C7 Renewable energy & Cogeneration
cr
C1 Location and climate data
us
C6 Hot water system
an
C2 Geometrical shape
C3 Building envelope
C5 Lighting system
M
C4 Mechanical Systems
te
d
Alternatives B1, B2, B3
Ac ce p
Fig. 4. Network structure of the BRS problem.
Fig. 5. Weight values for the criterion C4
Page 20 of 26
[35] Table 1 Major building rating systems and weightings. BREEAM (Building fit-out only where applicable to scheme)
ip t
Weighting (%)
Section
Weighting (%)
Health & Wellbeing
17
Indoor Environmental Quality
Energy
21
Energy & Atmosphere
Transport
9
Sustainable Sites
Water
7
Water efficiency
15
cr
Section
LEED (For existing buildings)
35
us
26
14
14
Material & Resources
10
Management
13
Innovation in Operations
6**
Waste
8
Regional Priority
4**
Pollution
11
Innovation
10*
**
Design innovations can add a maximum of 10% onto the score. In addition to 100 points in LEED’s five credit categories, projects can earn up to 10 bonus points.
M
*
an
Materials
% Score
LEED
% Score
Unclassified
<30
Uncertified
<40
Pass
≥30
Certified
≥40
Good
≥45
Silver
≥50
Ac ce p
te
BREEAM
d
Table 2 Certification levels for BREEAM and LEED.
V. Good
≥55
Gold
≥60
Excellent
≥70
Platinum
≥80
Outstanding
≥85
Page 21 of 26
Table 3 BEP-TR’s criteria. Definition
C1
Location and climate data
Location and climate data must be same for the reference building and the building which is given energy identification card.
C2
Geometrical shape
Building planning, roof types, number of floors and total area must be same for both building.
C3
Building envelope
It includes all the building components that separate the indoors from the outdoors. They are exterior walls, foundations, roof, windows and doors.
C4
Mechanical Systems
Heating and cooling systems, ventilation system, fire extinguishing system, waste water system
C5
Lighting system
Electrical installations and lighting system
C6
Hot water system
Natural gas hot water heater (geyser) is selected for residence, central hot water system for others.
C7
Renewable energy & Cogeneration
Specific rules, building area larger than 20,000 m2, are applied to identify buildings which require renewable energy sources like solar, hydro, geothermal, wind and cogeneration systems., etc.
d
M
an
us
cr
ip t
Criteria
Abbr.
Energy Class
E p Score
A
0-39
B
40-79
C
80-99
D
100-119
E
120-139
F
140-174
G
175
Ac ce p
te
Table 4 Classification and scores in BEP-TR.
Page 22 of 26
Table 5 The fuzzy linguistic scale. Linguistic terms for performance
TFN
TFN (reciprocal)
Equal important (E)
Very poor
1, 1, 1
1,1,1
Weak important (W)
Poor
2, 3, 4
1/4,1/3,1/2
Strong important (S) Demonstrated important (D) Absolute important (A)
Fair
4, 5, 6
1/6,1/5,1/4
Good
6, 7, 8
Very good
8, 9, 9
cr
ip t
Linguistic terms for importance
1/8,1/7,1/6
an
us
1/9,1/9,1/8
Table 6 Pairwise comparison of factors without dependence among them.
(1.59, (0.17, (0.55, (0.87, (1.00, (0.55, (2.52,
C5 1.71, 0.21, 0.69, 1.00, 1.00, 0.58, 2.92,
(4.58, (1.00, (2.08, (2.08, (3.63, (5.24, (7.27,
C2 5.74, 1.00, 2.54, 2.54, 4.72, 6.26, 8.28,
6.60) 1.00) 3.17) 3.17) 5.77) 7.27) 8.65)
1.82) 0.28) 0.91) 1.14) 1.00) 0.63) 3.30)
(1.44, (0.14, (0.48, (0.91, (1.59, (1.00, (2.34,
C6 1.91, 0.16, 0.58, 1.00, 1.71, 1.00, 5.00,
2.52) 0.19) 0.72) 1.10) 1.82) 1.00) 3.14)
(3.63, (0.31, (1.00, (0.40, (1.10, (1.39, (2.00,
C3 4.72, 0.39, 1.00, 0.52, 1.44, 1.71, 2.47,
5.77) 0.48) 1.00) 0.69) 1.82) 2.08) 3.00)
(2.52, (0.31, (1.44, (1.00, (0.87, (0.91, (1.59,
(0.50, (0.12, (0.33, (0.38, (0.30, (0.32, (1.00,
C7 0.62, 0.12, 0.41, 0.48, 0.34, 0.20, 1.00,
0.79) 0.14) 0.50) 0.63) 0.40) 0.43) 1.00)
Geometric mean (1.74, 2.12, 2.51) (0.24, 0.27, 0.32) (0.64, 0.78, 0.95) (0.66, 0.78, 0.95) (0.99, 1.13, 1.27) (0.89, 0.94, 1.18) (2.09, 2.68, 2.82)
M
C1 C2 C3 C4 C5 C6 C7
1.00) 0.22) 0.28) 0.40) 0.63) 0.69) 2.00)
d
C1 1.00, 0.17, 0.21, 0.28, 0.58, 0.52, 1.61,
te
(1.00, (0.15, (0.17, (0.22, (0.55, (0.40, (1.26,
Ac ce p
C1 C2 C3 C4 C5 C6 C7
C4 3.56, 0.39, 1.91, 1.00, 1.00, 1.00, 2.08,
4.58) 0.48) 2.52) 1.00) 1.14) 1.10) 2.62)
Page 23 of 26
Table 7 Degree of relative impact for criteria.
C6 (0, 0, (0, 0, (0, 0, (0.50, 0.57, (0.06, 0.07, (0, 0, (0.31, 0.36,
0) 0) 0) 0.66) 0.08) 0) 0.41)
(0.23, (0, (0.08, (0.13, (0.26, (0.15, (0,
C7 0.27, 0, 0.09, 0.16, 0.29, 0.21, 0,
ip t
0.45) 0) 0.27) 0) 0) 0) 0.40)
(0.32, (0.40, (0, (0, (0, (0, (0.16,
cr
(0.27, (0, (0, (0.30, (0, (0, (0.30,
C4 (0.25, 0.26, 0.28) (0.12, 0.16, 0.20) (0.14, 0.15, 0.17) (0, 0, 0) (0, 0, 0) (0, 0, 0) (0.40, 0.42, 0.44)
0.29) 0) 0.10) 0.19) 0.31) 0.25) 0)
M
C5 C1 (0.38, 0.41, C2 (0, 0, C3 (0.22, 0.24, C4 (0, 0, C5 (0, 0, C6 (0, 0, C7 (0.30, 0.35,
0) 0) 0) 0) 0) 0) 0)
C3 0.38, 0.43) 0.43, 0.47) 0, 0) 0, 0) 0, 0) 0, 0) 0.19, 0.23)
us
(0, (0, (0, (0, (0, (0, (0,
C2 0.33, 0.41) 0, 0) 0, 0) 0.33, 0.38) 0, 0) 0, 0) 0.33, 0.38)
an
C1 C2 C3 C4 C5 C6 C7
C1 0, 0, 0, 0, 0, 0, 0,
Table 8 Evaluation for buildings.
d
BEP-TR
(%)
B1
B2
B3
Location and climate data
27
F
F
F
Geometrical shape
7
G
P
VG
Building envelope
10
VG
F
G
Mechanical Systems
16
G
P
F
Lighting system
12
VG
P
G
Hot water system
8
G
F
F
Renewable energy & Cogeneration
20
VG
P
F
B1
B2
B3
Score
7.3
3.9
5.8
Category
A
E
B
Score
7.5
3.9
5.7
Category
A
E
B
Score
7.9
2.4
4.9
Category
A
F
C
Ac ce p
te
Criteria
Buildings
Method ANP
AHP
TOPSIS
Page 24 of 26
Table 9 Allowable values of criteria weights Wcriteria ≥ Wincreased Wincreased
C1
0.27
C2 C3 C4
0.07 0.09 0.16
Category Wdecreased
B2
B3
B1
0.29
B
E
B
-
A
0.19 0.15 0.27
B A B
E D E
B B B
0.01 0.07 0.10
A B A
B2
B3
E
B
D E D
B B B
us
B1
cr
Current Wcriteria
ip t
Wcriteria ≤ Wdecreased
Category
0.12
0.78
A
F
B
0.10
B
E
B
C6
0.08
0.14
A
D
B
-
A
E
B
C7
0.20
0.41
A
E
C
0.19
B
E
B
Ac ce p
te
d
M
an
C5
Page 25 of 26
ip t
Highlights The paper proposes an approach which is easier to implement. The model based on fuzzy ANP is under linguistic variables. 7 criteria are used to evaluate and label buildings. These criteria are location and climate data, geometrical shape, building envelope, mechanical systems, lighting system, hot water system, renewable energy & cogeneration.
Ac ce p
te
d
M
an
us
cr
[36]
Page 26 of 26
S0378-7788(14)00008-5 http://dx.doi.org/doi:10.1016/j.enbuild.2013.12.059 ENB 4751
To appear in:
ENB
Received date: Revised date: Accepted date:
8-10-2013 20-12-2013 27-12-2013
Please cite this article as: M. Kabak, E. K¨ose, O. Kirilmaz, S. Burmao˘glu, A fuzzy multicriteria decision making approach to assess building energy performance, Energy and Buildings (2014), http://dx.doi.org/10.1016/j.enbuild.2013.12.059 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ip t
A fuzzy multi-criteria decision making approach to assess building energy performance
cr
Mehmet Kabak**,* Department of Industrial and System Engineering, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey.
us
Erkan Köse Defense Science Institute, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey.
an
Oğuzhan Kırılmaz Department of Industrial and System Engineering, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey. Serhat Burmaoğlu Leadership R&D and Application Center, Turkish Military Academy. Bakanlıklar, Ankara, 06654, Turkey.
M
ABSTRACT
Ac ce p
te
d
Due to an increasing demand for energy and rising energy prices, efficiency in energy consumption is fastbecoming a topic of significance. The building and construction sector has seen an increase of approximately 30 to 40% of overall energy consumption occurred; this has exceeded other major sectors such as industry and transport. Given the number of buildings and the cost of energy required to support these buildings, the developing of new approaches in the construction sector will be likely. This situation forces the various stakeholders to implement energy rating procedures to assess buildings’ energy performance. The most commonly utilized building environment assessment method currently used in Europe is the Building Research Establishment Environmental Assessment Method (BREEAM). Parallel to Europe, Turkey started its National Building Energy Performance Calculation Methodology (BEP-TR) in 2010. BREEAM and BEP-TR like other methods, require a lot of detailed and particular information in order to be implemented, and the procedure is fairly complicated. In addition, decision support systems can involve assessments, developed as a result of imprecise data in a qualitative manner. “Fuzzy set theory” can play a significant role in this kind of decision-making situation. This paper examines a “fuzzy multi criteria decision making (MCDM)” approach in order to analyze BEP-TR. This approach was applied to categorize alternative buildings according to their overall energy performance. Results are discussed in terms of developing a new and practical building rating system. Keywords: Fuzzy, Analytic Network Process, Energy performance assessment, BEP-TR
* Corresponding author: [email protected] , Tel: +90 312 4175190 ext. 4413.
Page 1 of 26
Page 2 of 26
d
te
Ac ce p us
an
M
cr
ip t
1. Introduction Expanding human population, growing complexities of civilization, economic growth, industry, and other related
ip t
issues are forcing countries to demand more energy sources. The fact that the world has limited resources requires countries to study energy and energy policies in detail. As a result of fossil fuel formation over the course of millions of
cr
years, alternative energy sources such as wind, biofuels, solar thermal and photovoltaic sources, are increasingly being considered as fuel sources. Another consideration in meeting this increased energy demand is to reduce energy
us
consumption by improving energy efficiency.
Energy consumption of the building and construction sector accounts for around 30-40% of total primary energy use
an
worldwide; consequently this sector contributes a great deal to greenhouse emissions and global warming [1]. In addition, this sector accounts for the use of approximately 40% of the world’s natural resources extracted in industrialized countries; 12% of available potable water, and the production of 45-65% of the waste later sent to landfills
M
[2]. Due to the projected increase of the global population from 6.5 billion in 2005 to approximately 9.0 billion in 2035, these figures are expected to continue to grow, [3]. In light of this consideration, organizations have been investing
d
significant resources to create sustainably built environments, stressing sustainable building renovation processes to
te
reduce energy consumption and carbon dioxide emissions [4]. Those who design and operate buildings need methods to evaluate the environmental impacts of their actions.
Ac ce p
Various rating methods, referred to as building rating systems (BRS), have been developed as important tools in measuring and evaluating the environmental performance of a building. The Building Research Establishment Environmental Assessment Method (BREEAM) and Leadership in Energy and Environmental Design (LEED) are the two most widely recognized environmental assessment methodologies used globally in the building and construction industry today.
BREEAM was launched in the UK in 1990 to provide an environmental assessment and labeling mechanism for buildings [5,6]. There are nine different sections/criteria, including management, health & well-being, energy, transport, water, materials, waste, land use & ecology, and pollution. There are also sub-criteria (not mentioned in this paper). Each criterion consists of a number of issues; each issue is scored and each criterion is weighted for the overall evaluation of a
Page 3 of 26
building. The building is then categorized and labeled according to its score. Table 1 shows the corresponding scores for each category [7]. The US Green Building Council (USGBC) formed its major green building code known as LEED in 2000 [8]. LEED
ip t
provides building owners and operators with a framework for identifying and implementing practical and measurable green building design, construction, operations and maintenance solutions. It validates that a building is designed and
cr
constructed to achieve high performance in key areas of human and environmental health. Buildings are scored according to the criteria and weights shown in Table 1 [9]. Extra “innovation” credit can be applied for instances where a
us
design can demonstrate a reduction in a building’s impact on the environment in an innovative way.
The individual credits are added up and then weighted in line with the Table 1 to give a final rating. The building
an
must meets all prerequisites and can achieve the minimum number of points necessary to earn the certified level. The certification levels for BREEAM and LEED are explained in Table 2 [7, 9].
M
d
Although the BREEAM and LEED systems are more commonly applied in industry, there are other national rating
te
systems. The Taiwan Architecture and Building Research Institute developed a similar system named “Green Building Evaluation and Labeling System” (GBELS). GBELS uses nine criteria including biodiversity, greenery, soil water
Ac ce p
content, daily energy savings, carbon dioxide emission reduction, waste reduction, indoor environment, water resource, and sewage and garbage improvement [10]. The Green Building Tool (GBT) is an evolving assessment system sponsored by National Resources Canada. Comprehensive Assessment System for Building Environmental Efficiency (CasBee) in Japan, Minergie in Switzerland, National Australian Built Environment Rating System (NABERS) in Australia, DGNB (Deutsche Gesellschaft für Nachhaltiges Bauen) in Germany, Green Rating for Integrated Habitat Assessment (GRIHA) in India are examples for BRS [11]. These methods used in different countries require expertise to assess the energy performance of buildings and the process is generally performed by external experts, resulting in a cost for each assessment. The originality of this paper lies in the way the owner-occupiers’ viewpoint is included in the assessment process. These stake-holders can easily apply the model proposed herein to assess the energy performance of a building without the help of the external experts.
Page 4 of 26
In addition, as seen in the process of BREEAM, LEEDS, BEP-TR and other evaluation systems, energy performance of buildings consist of multiple criteria.
ip t
Multi-criteria decision making (MCDM) comprises a finite set of alternatives, among which the decisions makers (DMs) must select, evaluate or rank according to the weights of a finite set of criteria. The multi-criteria nature of the
cr
buildings' energy performance assessment problem makes MCDM methods ideal to cope with the complexity of the problem. DMs consider many criteria simultaneously, with various weights, then evaluate the alternatives. Some
us
methods apply to the evaluation of qualitative criteria while others are suitable for quantitative criteria. But the buildings' energy performance assessment problem requires both qualitative and quantitative criteria. These methods also involve
an
subjective assessments, resulting with imprecise data in qualitative manner. Due to the availability and uncertainty of information in the decision process, as well as the ambiguities of human feeling and recognition, it is often difficult to make an exact evaluation and convey the feeling and recognition of objects for DMs. “Fuzzy set theory” can play a
M
significant role in this kind of decision situation [12, 13]. Hence since DMs generally fail to make a good numerical prediction for criteria, evaluation is expressed in linguistic terms. Fuzzy linguistic models permit the translation of verbal
d
expressions into numerical ones. Thereby, when dealing quantitatively with imprecision in the expression of the impor-
te
tance of each criterion, some multi-criteria methods based on fuzzy relations are used [14]. As a result, a MCDM method based on fuzzy logic is proposed to assess practically the energy performance of buildings. For this aim, National
Ac ce p
Building Energy Performance Calculation Methodology of Turkey (BEP-TR) is analyzed as a MCDM problem. There are seven criteria: location and climate data, geometrical shape, building envelope, mechanical systems, lighting system, hot water system and renewable energy & cogeneration in BEP-TR [15]. Since there is a set of criteria which affect each other in the BRS, the Analytic Network Process (ANP), as a wellknown MCDM method, is better able to analyze the impact of each criterion on every other criterion using pair wise comparisons [16]. It should be noted that experts who participate in this evaluation process cannot always explain their judgments about certain attributes such as quality or performance with distinct and discreet scales. The fuzzy set theory enables them to explain their evaluations in linguistic terms. This problem is further analyzed by using ANP based on the fuzzy set theory in this paper. The remainder of this paper is organized as follows: Section 2 describes buildings’ energy performance in Turkey. Main characteristics of the fuzzy set theory and fuzzy ANP (FANP) method are detailed in Sections 3 and 4,
Page 5 of 26
respectively. The steps of the proposed FANP model are explained and the application for BEP-TR is detailed in Section 5. Finally, the obtained results, conclusions and future research directions are discussed in Section 6.
ip t
2. Building energy performance in Turkey Scoring systems which use their own criteria and weights vary according to geography, country policy, people’s
cr
awareness about environment, and so forth, and thus have some differences. Prior to 2010, Turkey did not have its own voluntary rating system. On the other hand, the international BRS, BREEAM and LEED in particular were received with
experience in reducing energy consumption in buildings to Turkey.
us
interest by stakeholders of the Turkish construction industry. These stakeholders intend to bring other countries’
an
Legislation in Turkey on energy performance in buildings, in parallel with EPBD (Energy Performance of Buildings Directive), was announced by the Ministry of Public Works and Settlement on 5 December 2008, after consultation with
M
relevant ministries and universities. The Turkish government intends to use this method as a driving force towards green construction and start a transformation in the building and construction industry. In 2009, various international workshops were organized to exchange ideas and share experiences, discuss common problems and local solutions with
d
academics,, and energy experts. Using European standards, their adaptation to local conditions and their contribution to
te
further developments were discussed. BEP-TR was prepared in accordance with EU Building Energy Performance Directive (2002/91/EC) with respect to Turkey’s conditions [17]. BEP-TR is a set of criteria which is designed to guide
Ac ce p
the building and construction sector towards energy efficiency. In the evaluation process of BEP-TR, a reference building is identified by experts. Energy performance of the building and reference one are calculated based on seven criteria labeled in Table 3 [15]. A detailed depiction of how each section value is calculated by using thresholds, formulas require a complicated analysis which exceeds the borders of this paper. As a result, calculation of Energy Performance ( E p ) value only is explained through Eq.1. Energy class of the building is defined according to the E p score, as shown in Table 4 [15, 18].
E p = 100 * ( EPa / EPr )
(1)
E p : Energy performance
EP: Energy consumption per year ( kWh/m 2 ) a: Building for assessment r: Reference building
Page 6 of 26
ip t
LEED, but it will define the path towards a green BRS in Turkey. Therefore its approach to sustainability, tools and units of assessment it employs and the improvements it requires will be crucial for the future of sustainability discussions in Turkey’s building and construction industry. It should be noted that weights for criteria in BREEAM and LEED, and not
cr
BEP-TR, are basically declared. In this paper, the authors calculate criteria weights for BEP-TR by using FANP,
us
evaluate and label three buildings in Ankara, Turkey. 3. The fuzzy set theory
an
Since decision makers (DMs) generally fail to make a good numerical prediction for criteria, evaluation is expressed in linguistic terms. Additionally, human judgment on qualitative attributes tends to be subjective and imprecise.
M
Therefore, the fuzzy set theory is commonly used in decision making problems. Fuzzy numbers expand on the idea of the confidence interval and are defined over a fuzzy subset of real numbers. A triangular fuzzy number (TFN) shown in Fig. 1 is a type of fuzzy number and, should possess the some basic properties [19, 20].
d
A fuzzy number M defined on ℜ is a TFN if its membership function μ M~ ( y) : ℜ → [0,1] is equal to
te
~
Ac ce p
⎧( y − l ) /( m − l ), l ≤ y ≤ m ⎪
μ M~ ( y ) = ⎨(u − y ) /(u − m), m ≤ y ≤ u ⎪0, ⎩
(2)
otherwise
where l, u, m are real numbers and l ≤ m ≤ u . The linguistic variable scale and the corresponding TFNs used in this study are shown in Table 5 [21].
In this paper, the fuzzy set theory is incorporated with ANP (Analytic Network Process) through an evaluation form that uses linguistic variables. The value of the linguistic variables that a DM has assigned to the pair-wise comparison between each two criteria is converted into TFN scores. The FANP methodology that is applied for this paper is detailed in next section.
Page 7 of 26
4. The FANP method The Analytic Network Process (ANP) allows for complex inter-relationships among decision levels and attributes
ip t
[16]. The ANP feedback approach replaces hierarchies (Fig. 2a) with networks (Fig. 2b) in which the relationships between levels cannot be easily represented as higher or lower, dominant or subordinate, direct or indirect [22].
cr
In ANP, the modeling process can be divided to three steps, which are described as follows [23]:
us
Step 1: The pair-wise comparisons and relative weight estimation
Before performing the pair-wise comparisons, all criteria and clusters compared are linked to each other. The pair-
an
wise comparisons are made depending on the scale shown on Table 5. The inconsistency ratio (IR) is useful for identifying possible errors in judgments as well as actual inconsistencies in the judgments themselves. For example, if A is more important than B and B is more important than C, C cannot be important than A. If the value of IR is smaller or
M
equal to 10%, the inconsistency is acceptable. If the IR is greater than 10%, we need to revise the subjective judgment. While doing pairwise comparisons, the inconsistency value is considered in all stages[25, 26]. In the pair-wise
d
comparison matrix, the score of a ij represents the relative importance of the component on row (i) over the component
te
on column (j), i.e., aij = wi / w j . The reciprocal value of the expression ( 1 / a ij ) is used when the component j is more important than the component i . Then, a local priority vector (eigenvector) w is computed as an estimate of the relative
Ac ce p
importance accompanied by the elements being compared by solving the following equation:
Aw = λmax w where
(3)
λmax is the largest eigenvalue of matrix A.
Step 2: Formation of the initial supermatrix
The obtained vectors are further normalized to represent the local weight vector. A supermatix is formed, after which local weight vectors are entered in the appropriate columns of the matrix of influence among the elements, to obtain global priorities. The supermatrix representation of a network with three levels is given as follows (Fig. 2b): G C 0 ⎛0 W = Criteria (C ) ⎜ ⎜W21 W22 Alternatives ( A) ⎜ ⎝ 0 W32 Goal (G )
A 0⎞ ⎟ 0⎟ Ι ⎟⎠
(4)
Page 8 of 26
W 21 is a vector that represents the impact of the goal on the criteria, W22 is a vector that represents impact of the
interdependences among criteria, W 32 is also a vector that represents the impact of criteria on each of alternatives, and I
ip t
is the identity matrix. Any zero value in the super-matrix can be replaced by a matrix if there is an interrelationship of elements within a cluster or between the clusters. Step 3: Formation of the weighted super-matrix
cr
An eigenvector is obtained from the pair-wise comparison matrix of the row clusters with respect to the column
us
cluster, which in turn yields an eigenvector for each column cluster. The first entry of the respective eigenvector for each column cluster is multiplied by all the elements in the first cluster of that column, the second by all the elements in the second cluster of that column, and so on. In this way, the cluster in each column of the super-matrix is weighted, and the
an
result, known as the weighted super-matrix, is stochastic [27]. A detailed discussion regarding the mathematical processes of the ANP is provided in different papers [15,21,28,29].
M
In the proposed methodology, the pairwise comparison matrices are formed, and FANP has been used to determine weights of BRS’s criteria. The FANP can easily accommodate the interrelationships existing among the functional
te
interrelationships.
d
activities. The concept of supermatrices is employed to obtain the composite weights that overcome the existing
Pairwise comparison matrices are structured by using TFNs (l, m, u). The fuzzy matrix can be given as follows:
Ac ce p
l m u l m u ⎛ (a11 , a11 , a11) (a12 , a12 , a12 ) ... (a1ln , a1mn , a1un ) ⎞ ⎟ ⎜ l m u l m u , a21 , a21) (a22 , a22 , a22 ) ... (a2l n , a2mn , a2un ) ⎟ ~ ⎜ (a21 A=⎜ ⎟ M M M M ⎟ ⎜ ⎜ l m u l m u l m u ⎟ ⎝ (am1, am1, am1 ) (am1, am1, am1 ) ... (amn, amn, amn) ⎠
(5)
The a mn represents the of comparison m (row) with component n (column). The pair-wise comparison matrix (Ã) is assumed as reciprocal. ⎛ (1,1,1) ⎜ ⎜ 1 1 ⎜( u , m a12 ~ ⎜ a12 A= ⎜ M ⎜ 1 1 ⎜( ⎜ au , am ⎝ 1n 1n
l m u (a12 , a12 , a12 ) ... ( a1ln , a1mn , a1un ) ⎞ ⎟ ⎟ 1 , l ) (1,1,1) ... ( a 2l n , a 2mn , a 2un ) ⎟ a12 ⎟ ⎟ M M M ⎟ 1 1 1 1 ⎟ , l ) ( u , m , l ) ... (1,1,1) ⎟ a1n a 2 n a 2 n a 2 n ⎠
(6)
Page 9 of 26
~ . The In this study, the logarithmic least squares method is used for getting estimates for fuzzy priorities w i logarithmic least squares method for calculating triangular fuzzy weights can be given as follows [27]:
)
k = 1,2,3,..., n
ip t
(
~ W = Wkl , Wkm , Wku where
(∏ a ) = ∑ (∏ a ) n j =1
n i =1
s 1/ n kj
m 1/ n n j =1 ij
, s ∈ {l , m, u}
cr
Wks
(7)
(8)
us
DMs tend to evaluate alternatives by using linguistic scales. Pairwise comparisons, rating that is preferred on the situation of high numbers of criteria and candidates or other ranking methods i.e. TOPSIS, PROMETHEE could be used
an
to evaluate the performance of the alternatives [30,31]. In this paper, rating is used; overall priorities for alternatives and crisp values are calculated by using matrix operations. Crisp values for fuzzy weights are calculated as follows [32]: (9)
M
Valuecrisp = (l + 2m + u) / 4
5. A real case application
d
Then, the alternative with the largest priority should be selected.
te
Turkey is energy importing country, whereby 70% of the total energy consumption supplied is by imported energy
Ac ce p
[33]. This situation forces the country to develop domestic energy resources and to reduce energy consumption by improving energy efficiency. The Turkish government has developed its own BRS as BEP-TR to be able to use it as a driving force towards green construction and start a transformation in the building and construction industry. BEP-TR focuses on energy use of buildings and categorizes them according to their energy consumption. It requires many complex, technical formulas, thresholds, and standards for every criterion in the calculation process. This paper analyzes criteria weights for BEP-TR by using FANP. As a real case application, three buildings in Ankara, Turkey are categorized according to their energy performance by using the proposed fuzzy MCDM approach. Firstly the subject matter experts (SMEs), those persons who will evaluate criteria weights in BEP-TR and case buildings, are determined. Two of the experts are academic personnel who have published papers on energy and green buildings; three are experts on energy affairs in the Turkish Ministry of Energy and Natural Resources, and the
Page 10 of 26
remainder are the authors of this paper. The case study of this research is taken from a housing estate located in Ankara, Turkey, which is in Climate Zone III according to the zoning in TS 825 [34]. The following steps are applied to determine the weights of the criteria in BEP-TR:
can be measured by the FANP technique. The structure established is shown in Fig. 3.
ip t
Step 1: The problem is converted into a hierarchical structure in order to transform the criteria into a state in which they
cr
In this structure, the aim is determining the criteria weights and ranking them according to the importance levels.
Step 2: After forming the decision hierarchy, the weights of criteria to be used in evaluation process are assigned by
us
using FANP method. In this phase, assuming that there is no dependence among the criteria, the experts are given the task of forming individual pairwise comparison matrix as shown in Table 6. Geometric means of these values are found
an
to obtain the pairwise compassion matrix on which there is a consensus. The pairwise comparison matrix is analyzed by ~ using Microsoft Excel software, and the fuzzy eigenvector (W21 ) is obtained.
d te
Ac ce p
~ W21
⎡C1 ⎤ ⎡(0.20, 0.24, 0.29)⎤ ⎢ ⎥ ⎢ ⎥ ⎢C 2 ⎥ ⎢(0.03, 0.03, 0.04) ⎥ ⎢C 3 ⎥ ⎢(0.07, 0.09, 0.11) ⎥ ⎢ ⎥ ⎢ ⎥ = ⎢C 4 ⎥ = ⎢(0.08, 0.09, 0.11) ⎥ ⎢C ⎥ ⎢(0.11, 0.13, 0.15) ⎥ ⎢ 5⎥ ⎢ ⎥ ⎢C 6 ⎥ ⎢(0.10, 0.11, 0.14) ⎥ ⎢ ⎥ ⎢(0.24, 0.31, 0.32) ⎥ ⎦ ⎣C 7 ⎦ ⎣
M
Step 3: Sections in BRS are generally not independent; therefore, inter-dependence among the criteria is determined by
analyzing the impact of each criterion on every other criterion using fuzzy pairwise comparisons. These dependencies are determined by experts on the basis of a group study. The arrow from “Building envelope” to “Geometrical shape” means that the “Building envelope” affects the “Geometrical shape” in the determination of building energy performance. The arrow between “Building envelope” and “Renewable energy & Cogeneration” means they affect each other in the process. The other relations between criteria are as shown in Fig. 4. Based on these inner dependencies, pairwise comparison matrices are formed for the factors. The following question, ‘‘what is the relative influence of Mechanical Systems when compared with Renewable energy & Cogeneration on controlling Geometrical shape?’’ may arise in pairwise comparisons. The resulting fuzzy eigenvectors are calculated and
Page 11 of 26
~ using the computed relative fuzzy importance weights, the inner dependence matrix of the criteria (W 22 ) is formed. The
value of (0, 0, 0) means, there is no relation between two criteria.
~ and W 21 as follows:
cr
us
Wcriteria(crisp )
⎡C1 ⎤ ⎡0.27⎤ ⎢ ⎥ ⎢ ⎥ ⎢C 2 ⎥ ⎢0.07⎥ ⎢C 3 ⎥ ⎢0.10⎥ ⎢ ⎥ ⎢ ⎥ = ⎢C 4 ⎥ = ⎢0.16⎥ ⎢C ⎥ ⎢0.12⎥ ⎢ 5⎥ ⎢ ⎥ ⎢C 6 ⎥ ⎢0.08⎥ ⎢ ⎥ ⎢0.20⎥ ⎦ ⎣C 7 ⎦ ⎣
an
~ ~ ~ Wcriteria = W22 * W21
⎡C1 ⎤ ⎡(0.15, 0.20, 0.25) ⎤ ⎢ ⎥ ⎢ ⎥ ⎢C 2 ⎥ ⎢(0.04, 0.05, 0.07) ⎥ ⎢C 3 ⎥ ⎢(0.05, 0.07, 0.09) ⎥ ⎢ ⎥ ⎢ ⎥ = ⎢C 4 ⎥ = ⎢(0.09, 0.12, 0.17)⎥ ⎢C ⎥ ⎢(0.07, 0.10, 0.11) ⎥ ⎢ 5⎥ ⎢ ⎥ ⎢C 6 ⎥ ⎢(0.04, 0.06, 0.08) ⎥ ⎢ ⎥ ⎢(0.12, 0.15, 0.20) ⎥ ⎦ ⎣C 7 ⎦ ⎣
ip t
Step 4: In this step, the interdependent fuzzy priorities of the criteria are calculated by using the data given in Tables 7
M
According to the results, the three most significant criteria are C1, C7, and C4, respectively. Significant differences
d
are observed in the results obtained for the criteria priorities when interdependent priorities of the criteria ( wcriteria ) and
te
dependencies are not taken into account. For example, the result for C1 changes from (0.20, 0.24, 0.29) to (0.15, 0.20, 0.25).
Ac ce p
Step 5: In this step, we calculate the performance value of the housing estates with respect to each fuzzy criteria. To
demonstrate, three buildings in Ankara, Turkey are evaluated using BEP-TR criteria. The first building (B1) is a typical residential building which was constructed in 2009. It serves as both a residential structure as well as an office (commercial) structure. The second, (B2), is a 35-year-old building and its health & well-being is unknonwn, but estimated to be inadequate. The last building (B3) was constructed in 2001 and is estimated to need some modifications. Using Microsoft Excel software, the fuzzy eigenvectors are computed by analyzing these matrices and the verbal evaluation values in Table 8. Finally, the overall performance value of the housing estate reflecting the interrelationships within the criteria are calculated as shown in Table 8. Once these performance values are normalized, categories are determined for the buildings. The FANP analysis results indicate that the B1, B2 and B3 are labelled as A, E and B, respectively.
Page 12 of 26
Two different methods, fuzzy AHP and fuzzy TOPSIS, are used for the same calculations to compare the results obtained with FANP. While the scores for buildings change, the labels remain same within the fuzzy AHP analysis. The difference in scores is an expected result based on AHP that does not take into account interrelationships within the
ip t
criteria. TOPSIS is based on the concept that the most preferred alternative should have the shortest distance from the positive ideal solution and have the longest distance from the negative ideal solution. Scores and categories are
different MCDM techniques.
us
cr
determined by using the fuzzy TOPSIS as shown in Table 8. Finally, similar conclusions are obtained by using the three
an
Sensitivity analysis of the proposed model leads to a much better understanding of the determining criteria for
M
building energy performances in Turkey. A sensitivity analysis based on criteria weights is performed in this section. For this aim, seven cases are created and the sensitiveness of criteria weights is investigated. When current weights of the
d
criteria are increased individually, the category of the building "B2" rises from E to D while others stay at the same
te
category. Decreasing criteria's weights results that the category for the building "B1" goes to B and stays same for others. For instance, values for the criterion C4 are increased as shown in Fig. 5. Values which are equal or greater than 0.19
Ac ce p
makes the categories of buildings B1, B2, and B3 as A, D, and B, respectively. On the other hand, values equal and less than 0.15 changes categories as B, E, B. The cases are summarized in Table 9.
6. Conclusion
This research illustrates how to make implement a more simplistic and easily-applied energy code, BEP-TR, for buildings, and shows how to develop the performance evaluation model and calculate the evaluation result by using a fuzzy MCDM approach. From the evaluation model’s result, construction companies can determine which part will be the priority criteria in order to improve the effectiveness and efficiency of their construction. By the algorithm of this study, the competition capability for construction companies could be increased. Following additional development and
Page 13 of 26
testing, the proof of the validity and applicability of such an energy code could support both the construction industry and Turkish government energy policies. BEP-TR focuses on buildings’ energy consumption and is not an exact equivalent of BREEAM or LEED. For this
ip t
reason, criteria weights can be evaluated by using the proposed method in further research and applied for determination of green building construction in Turkey. Additionally, BEP-TR will define the path towards a national green BRS.
cr
Revised BEP-TR. BRS must address sustainability and should go beyond the borders of energy, water and recyclables. Green building design practices must encompasses concepts, such as economic performance, contribution to
us
employment, protection of biodiversity, labor processes, security, safety, public health, education and cultural preservation.
an
In the future, this study could be used to develop a new national BRS. The obtained feedback should prove useful to improve the methodology and to build a software program for the BRS, and the role BRS will play in raising awareness
M
in the construction industry for a sustainable future.
d
References
(3) (2008) 394–398.
te
[1] L. Perez-Lombard, J. Ortiz, C. Pout, A review on buildings energy consumption information, Energy and Buildings 40
Ac ce p
[2] D. Castro-Lacouture, J.A. Sefair. L. Florez, A.L. Medaglia, Optimization model for the selection of materials using a LEED based green building rating system in Colombia, Build Environ 44 (2009) 1162–1170. [3] M.K. Dixit, J.L. Fernandez-Solis, S. Lavy, C.H. Culp, Identification of parameters for embodied energy measurement: a literature review, Energy Build 42 (2010) 1238–1247. [4] Y.K. Juan, P. Gao, J. Wang, A hybrid decision support system for sustainable office building renovation and energy performance improvement, Energy and Buildings 42 (3) (2010) 290–297. [5] R. Baldwin, A. Yates, N. Howard, S. Rao, BREEAM 98 for Offices (Building Research Establishment, Watfordt,1998). [6] V. Olgyay and J. Herdt, The application of ecosystems services criteria for green building assessment, Solar Energy 77 (2004) 389–398. [7] HRS
Services.
BREEAM
code
for
Sustainable
Homes
2008.
Available
from:
http://www.hrsservices.co.uk/pages/breeam/breeam_explained.html: April 2012.
Page 14 of 26
[8] J.H. Scofield, Do LEED-certified buildings save energy? Not really, Energy and Buildings 41 (12) (2009) 1386–1390. [9] USGBC.
LEED
for
existing
buildings
operations
and
maintenance
2009.
Available
from
http://www.usgbc.org/ShowFile.aspx?DocumentID=8876; April 2012.
ip t
[10]V.W.Y. Tam, The effectiveness of the green building evaluation and labeling system, Architectural Science Review 50 (4) (2007) 323-330.
cr
[11] M. Bayraktar, E. Kalaycioglu, A.Z. Yilmaz, Proceedings of Building Simulation 2011, in 12th Conference of
[12] L.A. Zadeh, Fuzzy sets. Information and Control 8 (1965) 338–353.
us
International Building Performance Simulation Association, (Sydney, 14–16 November 2011).
[13] G. Büyüközkan and G.Çifçi, A novel hybrid MCDM approach based on fuzzy DEMATEL, fuzzy ANP and fuzzy
an
TOPSIS to evaluate green suppliers, Expert Systems with Applications 39 (3) (2012) 3000-3011. [14] Z. Güngör, G. Serhadlıoğlu, and S.E. Kesen, A fuzzy AHP approach to personnel selection, Applied Soft Computing 9
M
(2009) 641-646.
[15] Energy Performance for Buildings. Regulation, calculation method, reference building & energy classification, software, process, Turkish Republic the Ministry of Public Works and Settlement 2009. Available from
d
http://www.csb.gov.tr/iller/ dosyalar/dosya/il_webmenu8900.pdf; March 2012.
Pittsburgh, 1996).
te
[16] T.L. Saaty, Decision making with dependence and feedback: The Analytic Network Process (RWS Publications,
Ac ce p
[17] National Building Energy Performance Calculation Methodology of Turkey. (No: YİG/2010-02)), Turkish Official Journal 2010.
[18] A.K. Yener, A method for the determination of building’s lighting energy performance: BEP-TR, in 10th National Plant Engineering Congress, (Izmir, 13-16 April 2011).
[19] G. Facchinetti and R.G. Ricci, A characterization of a general class of ranking functions on triangular fuzzy numbers, Fuzzy Sets and Systems 146 (2004) 297–312.
[20] M. Salehi and R. Tavakkoli-Moghaddam, Project selection by using a fuzzy TOPSIS technique, Engineering and Technology 30 (2008) 85–90.
[21] H.F. Lin, An application of fuzzy AHP for evaluating course website quality, Computers & Education 54 (2010) 877– 888.
Page 15 of 26
[22] T.L. Saaty and L.G. Vargas, Diagnosis with dependent symptoms: Bayes theorem and the analytic network process, Operations Research 46 (4) (1998) 491–502.
telecommunication company, Expert Systems with Applications 36 (2009) 3887-3895.
ip t
[23] S. Onut, S.S. Kara, E. Isık, Long term supplier selection using a combined fuzzy MCDM approach: A case study for a
textile firm, Information Sciences 177 (2007) 3364–3382.E.A.
cr
[24] I. Yüksel, I. and M. Dağdeviren, Using the analytic network process (ANP) in a SWOT analysis-A case study for a
[25] Demirtaş and Ö. Üstün, An integrated multiobjective decision making process for supplier selection and order
us
allocation, The International Journal of Management Science 36 (2008) 76-90.
[26] M.S. Ozdemir, Validity and inconsistency in the analytic hierarchy process, Applied Mathematics and Computation
an
161 (2005) 707-720.
[27] J. Ramik, A decision system using ANP and fuzzy inputs, in 12th international conference on the foundations and
M
applications of utility, Risk and Decision Theory, (Roma, 2006).
[28] R.P. Mohanty, R. Agarwal, A.K. Choudhury, M.K. Tiwari, A fuzzy ANP-based approach to R&D project selection: A
d
case study, International Journal of Production Research 43 (2005) 5199–5216.
te
[29] M. Kabak, S. Burmaoğlu, Y. Kazançoğlu, A fuzzy hybrid MCDM approach for professional selection, Expert Systems with Applications 39 (2012) 3516-3525.
Ac ce p
[30] M. Dağdeviren, A hybrid multi-criteria decision-making model for personnel selection in manufacturing systems, J Intell Manuf. 21 (2010) 451–460.
[31] M. Sevkli, A. Oztekin, O. Uysal, G. Torlak, A. Turkyilmaz, D. Delen, Development of a fuzzy ANP based SWOT analysis for the airline industry in Turkey, Expert Systems with Applications 39 (2012) 14–24. [32] R.D. Raut, Evaluation of supplier selection criteria by combination of AHP and fuzzy DEMATEL method, Int.J. Business Innovation and Research 5 (4) (2011) 359–392.
[33] İ. Kotçioğlu, Clean and sustainable energy policies in Turkey, Renewable and Sustainable Energy Reviews 15 (9) (2011) 5111–5119. [34] TSI. Thermal insulation requirements for buildings (Vol. TS 825). Ankara: Turkish Standards Institution, 2008.
Page 16 of 26
~
Fig. 1. A triangular fuzzy number, M
ip t
Fig. 2. Hierarchy and network: (a) hierarchy; (b) network. Fig 3. Schematic diagram of the proposed model for BRS.
cr
Fig. 4. Network structure of the BRS problem.
Ac ce p
te
d
M
an
us
Fig. 5. Weight values for the criterion C4
Page 17 of 26
μ M~
Ml(y)
Mr(y)
0.0 ~
Ac ce p
te
d
M
an
Fig. 1. A triangular fuzzy number, M
u
us
m
cr
M
l
ip t
1.0
Fig. 2. Hierarchy and network: (a) hierarchy; (b) network [23, 24].
Page 18 of 26
Determination of decision makers
ip t
Identify the Building Rating System’s factors
us
Approve the model?
Y
FANP
an
N
cr
Structure the decision model
Determine the weights of
~
M
the factors (W21 )
Determine inner dependence
~
d
matrix of the factors (W22 )
Ac ce p
te
Determine the interdependent weights of the factors
Evaluate and label the buildings
ig 3. Schematic diagram of the proposed model for BRS.
Page 19 of 26
ip t
GOAL Determining BRS criteria weights
C7 Renewable energy & Cogeneration
cr
C1 Location and climate data
us
C6 Hot water system
an
C2 Geometrical shape
C3 Building envelope
C5 Lighting system
M
C4 Mechanical Systems
te
d
Alternatives B1, B2, B3
Ac ce p
Fig. 4. Network structure of the BRS problem.
Fig. 5. Weight values for the criterion C4
Page 20 of 26
[35] Table 1 Major building rating systems and weightings. BREEAM (Building fit-out only where applicable to scheme)
ip t
Weighting (%)
Section
Weighting (%)
Health & Wellbeing
17
Indoor Environmental Quality
Energy
21
Energy & Atmosphere
Transport
9
Sustainable Sites
Water
7
Water efficiency
15
cr
Section
LEED (For existing buildings)
35
us
26
14
14
Material & Resources
10
Management
13
Innovation in Operations
6**
Waste
8
Regional Priority
4**
Pollution
11
Innovation
10*
**
Design innovations can add a maximum of 10% onto the score. In addition to 100 points in LEED’s five credit categories, projects can earn up to 10 bonus points.
M
*
an
Materials
% Score
LEED
% Score
Unclassified
<30
Uncertified
<40
Pass
≥30
Certified
≥40
Good
≥45
Silver
≥50
Ac ce p
te
BREEAM
d
Table 2 Certification levels for BREEAM and LEED.
V. Good
≥55
Gold
≥60
Excellent
≥70
Platinum
≥80
Outstanding
≥85
Page 21 of 26
Table 3 BEP-TR’s criteria. Definition
C1
Location and climate data
Location and climate data must be same for the reference building and the building which is given energy identification card.
C2
Geometrical shape
Building planning, roof types, number of floors and total area must be same for both building.
C3
Building envelope
It includes all the building components that separate the indoors from the outdoors. They are exterior walls, foundations, roof, windows and doors.
C4
Mechanical Systems
Heating and cooling systems, ventilation system, fire extinguishing system, waste water system
C5
Lighting system
Electrical installations and lighting system
C6
Hot water system
Natural gas hot water heater (geyser) is selected for residence, central hot water system for others.
C7
Renewable energy & Cogeneration
Specific rules, building area larger than 20,000 m2, are applied to identify buildings which require renewable energy sources like solar, hydro, geothermal, wind and cogeneration systems., etc.
d
M
an
us
cr
ip t
Criteria
Abbr.
Energy Class
E p Score
A
0-39
B
40-79
C
80-99
D
100-119
E
120-139
F
140-174
G
175
Ac ce p
te
Table 4 Classification and scores in BEP-TR.
Page 22 of 26
Table 5 The fuzzy linguistic scale. Linguistic terms for performance
TFN
TFN (reciprocal)
Equal important (E)
Very poor
1, 1, 1
1,1,1
Weak important (W)
Poor
2, 3, 4
1/4,1/3,1/2
Strong important (S) Demonstrated important (D) Absolute important (A)
Fair
4, 5, 6
1/6,1/5,1/4
Good
6, 7, 8
Very good
8, 9, 9
cr
ip t
Linguistic terms for importance
1/8,1/7,1/6
an
us
1/9,1/9,1/8
Table 6 Pairwise comparison of factors without dependence among them.
(1.59, (0.17, (0.55, (0.87, (1.00, (0.55, (2.52,
C5 1.71, 0.21, 0.69, 1.00, 1.00, 0.58, 2.92,
(4.58, (1.00, (2.08, (2.08, (3.63, (5.24, (7.27,
C2 5.74, 1.00, 2.54, 2.54, 4.72, 6.26, 8.28,
6.60) 1.00) 3.17) 3.17) 5.77) 7.27) 8.65)
1.82) 0.28) 0.91) 1.14) 1.00) 0.63) 3.30)
(1.44, (0.14, (0.48, (0.91, (1.59, (1.00, (2.34,
C6 1.91, 0.16, 0.58, 1.00, 1.71, 1.00, 5.00,
2.52) 0.19) 0.72) 1.10) 1.82) 1.00) 3.14)
(3.63, (0.31, (1.00, (0.40, (1.10, (1.39, (2.00,
C3 4.72, 0.39, 1.00, 0.52, 1.44, 1.71, 2.47,
5.77) 0.48) 1.00) 0.69) 1.82) 2.08) 3.00)
(2.52, (0.31, (1.44, (1.00, (0.87, (0.91, (1.59,
(0.50, (0.12, (0.33, (0.38, (0.30, (0.32, (1.00,
C7 0.62, 0.12, 0.41, 0.48, 0.34, 0.20, 1.00,
0.79) 0.14) 0.50) 0.63) 0.40) 0.43) 1.00)
Geometric mean (1.74, 2.12, 2.51) (0.24, 0.27, 0.32) (0.64, 0.78, 0.95) (0.66, 0.78, 0.95) (0.99, 1.13, 1.27) (0.89, 0.94, 1.18) (2.09, 2.68, 2.82)
M
C1 C2 C3 C4 C5 C6 C7
1.00) 0.22) 0.28) 0.40) 0.63) 0.69) 2.00)
d
C1 1.00, 0.17, 0.21, 0.28, 0.58, 0.52, 1.61,
te
(1.00, (0.15, (0.17, (0.22, (0.55, (0.40, (1.26,
Ac ce p
C1 C2 C3 C4 C5 C6 C7
C4 3.56, 0.39, 1.91, 1.00, 1.00, 1.00, 2.08,
4.58) 0.48) 2.52) 1.00) 1.14) 1.10) 2.62)
Page 23 of 26
Table 7 Degree of relative impact for criteria.
C6 (0, 0, (0, 0, (0, 0, (0.50, 0.57, (0.06, 0.07, (0, 0, (0.31, 0.36,
0) 0) 0) 0.66) 0.08) 0) 0.41)
(0.23, (0, (0.08, (0.13, (0.26, (0.15, (0,
C7 0.27, 0, 0.09, 0.16, 0.29, 0.21, 0,
ip t
0.45) 0) 0.27) 0) 0) 0) 0.40)
(0.32, (0.40, (0, (0, (0, (0, (0.16,
cr
(0.27, (0, (0, (0.30, (0, (0, (0.30,
C4 (0.25, 0.26, 0.28) (0.12, 0.16, 0.20) (0.14, 0.15, 0.17) (0, 0, 0) (0, 0, 0) (0, 0, 0) (0.40, 0.42, 0.44)
0.29) 0) 0.10) 0.19) 0.31) 0.25) 0)
M
C5 C1 (0.38, 0.41, C2 (0, 0, C3 (0.22, 0.24, C4 (0, 0, C5 (0, 0, C6 (0, 0, C7 (0.30, 0.35,
0) 0) 0) 0) 0) 0) 0)
C3 0.38, 0.43) 0.43, 0.47) 0, 0) 0, 0) 0, 0) 0, 0) 0.19, 0.23)
us
(0, (0, (0, (0, (0, (0, (0,
C2 0.33, 0.41) 0, 0) 0, 0) 0.33, 0.38) 0, 0) 0, 0) 0.33, 0.38)
an
C1 C2 C3 C4 C5 C6 C7
C1 0, 0, 0, 0, 0, 0, 0,
Table 8 Evaluation for buildings.
d
BEP-TR
(%)
B1
B2
B3
Location and climate data
27
F
F
F
Geometrical shape
7
G
P
VG
Building envelope
10
VG
F
G
Mechanical Systems
16
G
P
F
Lighting system
12
VG
P
G
Hot water system
8
G
F
F
Renewable energy & Cogeneration
20
VG
P
F
B1
B2
B3
Score
7.3
3.9
5.8
Category
A
E
B
Score
7.5
3.9
5.7
Category
A
E
B
Score
7.9
2.4
4.9
Category
A
F
C
Ac ce p
te
Criteria
Buildings
Method ANP
AHP
TOPSIS
Page 24 of 26
Table 9 Allowable values of criteria weights Wcriteria ≥ Wincreased Wincreased
C1
0.27
C2 C3 C4
0.07 0.09 0.16
Category Wdecreased
B2
B3
B1
0.29
B
E
B
-
A
0.19 0.15 0.27
B A B
E D E
B B B
0.01 0.07 0.10
A B A
B2
B3
E
B
D E D
B B B
us
B1
cr
Current Wcriteria
ip t
Wcriteria ≤ Wdecreased
Category
0.12
0.78
A
F
B
0.10
B
E
B
C6
0.08
0.14
A
D
B
-
A
E
B
C7
0.20
0.41
A
E
C
0.19
B
E
B
Ac ce p
te
d
M
an
C5
Page 25 of 26
ip t
Highlights The paper proposes an approach which is easier to implement. The model based on fuzzy ANP is under linguistic variables. 7 criteria are used to evaluate and label buildings. These criteria are location and climate data, geometrical shape, building envelope, mechanical systems, lighting system, hot water system, renewable energy & cogeneration.
Ac ce p
te
d
M
an
us
cr
[36]
Page 26 of 26