Journal of Cleaner Production 142 (2017) 3728e3740

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Integrated QFD-MCDM framework for green supplier selection

Morteza Yazdani a, *, Prasenjit Chatterjee b, Edmundas Kazimieras Zavadskas c,
Sarfaraz Hashemkhani Zolfani d
a
Department of Business and Management, Universidad Europea de Madrid, 28670, Madrid, Spain
b
Department of Mechanical Engineering, MCKV Institute of Engineering, West Bengal, India
c _
Research Institute of Smart Building Technologies, Faculty of Civil Engineering, Vilnius Gediminas Technical University, Sauletekio al. 11, LT-1022, Vilnius,
Lithuania
d
Technology Foresight Group, Department of Management, Science and Technology, Amirkabir University of Technology (Tehran Polytechnic), P.O. Box
1585-4413, Tehran, Iran

a r t i c l e i n f o a b s t r a c t

Article history: Supplier evaluation and selection is a significant strategic decision for reducing operating costs and
Received 21 February 2015 improving organizational competitiveness to develop business opportunities. Moreover, with increasing
Received in revised form concern towards environmental protection and sustainable development, it becomes important to pay
15 October 2016
more attention to environmental requirements and evaluating the potential suppliers by incorporating
Accepted 19 October 2016
Available online 26 October 2016
green factors into the selection process. Thus, the aim of this paper is to put forward an integrated
approach for green supplier selection by considering various environmental performance requirements
and criteria. The proposed approach addresses the inter-relationships between the customer re-
Keywords:
Green supplier selection
quirements (CRs) with the aid of decision-making trial and evaluation laboratory (DEMATEL) method
Decision support system while constructing a relationship structure. Quality function deployment (QFD) model is used to
Complex proportional assessment establish a central relationship matrix in order to identify degree of relationship between each pair of
Multi-objective optimization based on ratio supplier selection criteria and CRs. Finally, complex proportional assessment (COPRAS) applied to pri-
analysis oritize and rank the alternative suppliers. A case study is presented to reveal the potentiality and aptness
Quality function deployment of the proposed methodology.
Customer needs © 2016 Elsevier Ltd. All rights reserved.

1. Introduction In a supply chain network (SCN), supplier evaluation and se-

lection is a deliberated task because of high complications with
Supply chain management (SCM) is a process of systematizing several conflicting, commensurable, ordinal and cardinal factors
and amalgamating diverse activities, starting from the customer's involved in the decision-making process (Yazdani et al., 2016). In
order to end product delivery in a well-organized manner. The this network, organizations are forced to harmonize their pur-
success or failure of any SCM largely depends upon a suitable sys- chasing activities in order to gain advantage. In managing pur-
tem and appropriate suppliers. Today's fast changing SCM envi- chasing activities proficiently, supplier evaluation and selection
ronment is characterized by low profit boundary, high quality plays a crucial role and has become a very fundamental component
expectations with less waiting time for orders (Ke et al., 2015) Even for viable benefits of any organization (Rezaei et al., 2014). In
over the last two decades the world economy has been spectacu- addition, selecting suitable suppliers considerably reduces material
larly changed due to a variety of reasons. Modern day's business purchasing cost, increases flexibility and product quality and
environment is frequently distinguished by increasing intricacy, eventually helps to accelerate the process of material purchasing.
ambiguity, unsteadiness and unpredictability. Thus, organizations The main objectives of supplier selection are thus to reduce the risk
must take each and every opportunity to advance their operational of purchasing and development of a long-term relationship be-
performance to stay competitive in the worldwide marketplace by tween buyer and supplier (Omurca, 2013).
appropriately selecting its trading partners. Mostly, there are two aspects in the context of supplier evalu-
ation and selection. Evaluation of suppliers depends on several
criteria. Some criteria such as price, quality, delivery performance,
* Corresponding author. branding or reputation are frequently selected for comparison and
E-mail address: [email protected] (M. Yazdani).

http://dx.doi.org/10.1016/j.jclepro.2016.10.095
0959-6526/© 2016 Elsevier Ltd. All rights reserved.
 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740 3729

assessment (Rajesh and Ravi, 2015). Traditional studies on suppler for supplier selection problems and tools and techniques for
selection have concentrated broadly on relevant supplier selection selecting green suppliers. In Section 3, research significance is
methods and criteria without considering the voice of the cus- addressed. Section 4 presents the proposed integrated framework
tomers or customer requirements (CRs). However, for an efficient for green supplier selection. Section 5 illustrates an example on
SCM, establishing a clear understanding about customers' needs, green supplier selection to validate the proposed integrated model.
specifically their subjective performance requirements play a vital The results and discussions and managerial implications are also
role. Also, large portion of these studies have evaluated suppliers presented in this section. Section 6 concludes the paper. Appendix
under economic criteria such as cost, flexibility and delivery rather includes all the required methods and formulas.
than highlighting buyer satisfaction and characteristics related to
the supplier's criteria. Consequently, there is a lack of connecting 2. Literature review
supplier evaluation criteria to the CRs which are prerequisites to be
fulfilled for the benefit of the stakeholders. This section presents a brief review of literature on traditional
It is observed that the previous studies have put concentrations and green supplier selection criteria and traditional and green
only on economic efficiency of suppliers rather than ecological ef- supplier selection methods. Both green and sustainable supplier
ficiency and rarely noticed and took into account the environ- selection are common research in SCM framework of study. Sus-
mental issues to the supplier selection studies (Kumar et al., 2014). tainability is a bigger picture of strategies for managing the world
Environmental deteriorations have forced public and private sec- from different aspects but in level of companies. Green supplier
tors to think over environmental and sustainable issues. To survive selection can be considered as the vital part of bigger picture,
in the global market in wake of developing competition, and due to sustainability. The environment is the main part of sustainable
the increased awareness about environmental concerns, green supplier selection and its dimensions (Liou et al., 2016). The main
supply chain management (GSCM) is appearing in recent literature act of companies which can be useful in bigger structure is actions
(Govindan et al., 2013; Bai and Sarkis, 2010; Khaksar et al., 2016). about the environment. This research is established in companies’
Since major components of an end product are supplied by the level and their activities. The bigger picture is out of scope for this
external suppliers, hence it becomes important to pay more research. Sustainable Supply Chain Management (SSCM) should be
attention to environmental requirements in the supply chain pro- considered in regional, national and international policies and
cess and evaluating the potential suppliers by incorporating green strategies with a clear vision.
criteria into the selection process (Kannan et al., 2014). Importance It has become increasingly imperative for organizations facing
of environmental sustainability in an SCM highly depends on the competitive, regulatory and community pressures to search for a
purchasing strategies and its suppliers. Since the supplier selection balance between economic and environmental performance.
process is analyzed as a combination of CRs and technical re- Presently, many organizations are putting efforts to go green in
quirements (TRs) or supplier selection criteria. Therefore, supplier their businesses due to their apprehension for ecological sustain-
relation can be observed as a house of quality (HoQ) model that ability (Liao et al., 2015). Activities comprised in GSCM are re-use,
seeks to categorize which of the supplier's attributes have the recycle, remanufacture and reverse logistics etc. Lu et al. (2007)
greatest impact on the attainment of its established goals. addressed materials, energy, solid residue, liquid residue, gaseous
This paper aims at developing an integrated approach for eval- residue, and technology as the fundamental environmental criteria.
uating supplier performance and selecting the best supplier while Tuzkaya et al. (2009) recognized pollution control, green product,
considering both traditional and green supplier selection criteria green process management and environmental and legislative
simultaneously. An attempt is made to evaluate and rank ten management as the most significant GSCM criteria. Awasthi et al.
alternative green suppliers for a reputed Iranian dairy company (2010) proposed 12 environmental criteria including use of envi-
using the integrated approach consisting of decision-making trial ronmental friendly technology, use of environmental friendly ma-
and evaluation laboratory (DEMATEL), quality function deployment terial, partnership with green organization, green market share,
(QFD) model, complex proportional assessment (COPRAS) management commitment, adherence to environmental policies,
(Zavadskas et al., 1994; Hashemkhani Zolfani et al., 2012) and green R&D projects, staff training and etc.
multi-objective optimization on the basis of ratio analysis (MOORA) The need for environmental practices in the food supply chain is
(Brauers and Zavadskas, 2006, 2009) methods. becoming acute. The food industry currently has to contend with
The contribution of our proposed model is some folds; causal multiple competing pressures alongside the new challenges of
relationship and interaction level between the CRs and supplier green production, particularly, where the focus of attention is on
selection criteria are addressed using DEMATEL method, and the environmental and greening supply chains, organizations are more
most important CR having significant influence on the remaining concerned (Marsden et al., 2000).
CRs is also identified. QFD provides the weights of the evaluating Environmental and green development in dairy industry is
criteria, which are derived by the degree of relationship between becoming one of the most prominent topics of our time (Patzelt and
CRs and evaluating criteria. MOORA and COPRAS methods are Shepherd, 2011; Gimenez and Tachizawa, 2012). Glover et al. (2014)
applied to prioritize and compare the green suppliers. We have investigated on the environmental impact of the whole supply
brought up a real case study of a famous dairy company in Iran to chain. Recent food crises have increased consumer awareness of the
implement the proposed model and guide managers to a clearer impact on public health of food production, processing, and dis-
understanding of customer standards and regulations. The adopted tribution in all over the world (Van Kleef et al., 2006). Consumers
multi-criteria approach can assist the decision makers (DMs) to have become more critical and tend to be informed about the
directly assess the attributes according to their prominence degree processes of food procurement, safety levels, production methods,
while providing a ranking preorder to all the alternative suppliers. hygiene, security of transportation, and other environmental issues
The proposed integrated approach would be more realistic and like carbon footprints (Redmond and Griffith, 2003). Nowadays,
practical particularly in supplier selection domain and generally consumers include factors like quality, safety and environmental
with respect to any complex decision-making situations. conformity in their buying decision. Managers in the food industry
Rest of paper interprets the following sections: section 2 de- and dairy companies will have to respond to these changing con-
scribes literature review including traditional and green supplier sumer demands by increasing environmentally protection of pro-
selection requirements and criteria, solo and integrated approaches cesses and products (Wognum et al., 2011).
3730 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740

Environmental impact of food production has put pressure on domain. This selection process originally includes several factors,
food companies to improve the efficiency of business management. feathers, models, and methods and is called a decision-making
This requires cooperation between all actors in a food supply chain process in the presence of multiple criteria and alternatives. From
since ultimately the consumer at the end of the chain decides on the literature survey (Refer to Table 1), it is observed that many past
the premium which is granted for all the efforts (Jones, 2002). researchers have already subjected a pool of methods and wide
Dairy supply chain is an economically important aspect of range of combined approaches to evaluate and choose appropriate
agriculture with international aspects to its supply chain. In the suppliers (Govindan et al., 2015). Integration of MCDM methods
dairy sector, organizations range in size both at each segment of, provides different approaches with particular functionalities and
and across, the supply chain. The industry is highly regulated and characteristics. Therefore, it is the art of the DMs to integrate,
there have been a number of environmental legislations and combine, hybridize and extend MCDM methods for specific ob-
standards introduced over the last few years from both national, jectives and requirement perspectives. It is also found that most of
European and international directives (Dries et al., 2009; Yu and the previous attempts by the past researchers for traditional as well
Huatuco, 2016). The dairy industry has also received high profile as green supplier selection problems have several disadvantages
environmentally related attention in relation to GHG emissions too (Chan and Wang, 2013; Aliev, 2013; Kahraman and Oztayşi, €
from cattle (methane). These kinds of standards and legislations 2014). Moreover, no endeavor has been projected till date to pre-
obligate dairy companies to control, observe, and measure all the sent a sound mathematical model for solving green supplier se-
environmental activities. Zhu and Sarkis (2006) believed supplier lection problems which considers the interaction between the CRs
environmental drivers and responsibility can directly affect the and supplier selection criteria respectively. Thus, in this paper, a
whole supply chain performance. humble effort is taken to balance this space by proposing an inte-
There is far less research that addresses the relationship be- grated analytical approach for selecting green suppliers strategi-
tween a dairy company environmental strategy, its internal inte- cally consisting of DEMATEL, QFD, and COPRAS methods.
gration in the form of the supply chain environmental assessment
and the external integration with stakeholders, customers and 3. Research significance
suppliers.
We explore what stakeholders across the diary supply chain As stated earlier, the supplier selection process is analyzed as a
need in order to increase supply chain efficiency in the context of combination of customer criteria and technical requirements
supplier evaluation. Developing a supplier selection strategy to (supplier criteria). In a food-based production company valuing
satisfy government pressures and international standard requires a specific need and attitude of customers can be a spectacular point
blend of competitive advantage, requiring superior firm resources which can raise its competitive and strategic advantage. Although
and capabilities as well as a fit between the external environment many studies addressed the influence of customer satisfaction
and interval actions. Large firms can use their strong strategies to indices and engineering features on the supplier selection process,
help improve good environmental and social practices across the to the best of our knowledge, none of them dealt with a systematic
supply chain. evaluation of how these variables interact each other. In addition,
This paper focuses attention on environmental and green factors the literature on green performance measurement of supplier is
which are often related to waste reduction, pollution reduction, tremendous (Govindan et al., 2015) a significant gap regarding to a
energy efficiency, emissions reduction, and a decrease in the con- systematic analysis of green supplier is obvious (Cuthbertson and
sumption of hazardous materials. Piotrowicz, 2008).
Suppliers are defined as one of the vital parts of an organization To bridge this gap, it is necessary to link green constructs and
who deliver all the requirements for producing complete product sub-constructs of SCM and all the inter-organizational elements
from raw materials, components and services and a suitable sup- contributing to the performance measurement (Bhattacharya et al.,
plier is the one who meets these requirements at the right time, 2014).
acceptable quality and standards. In an SCN, working with efficient The proposed integrated framework aims at bridging both gaps
suppliers enhances the productivity and profitability of any orga- highlighted above by providing a systematic analysis of the in-
nization (Awasthi and Kannan, 2016). The process of assessment terdependencies existing among costumer variables and technical
and evaluation of suppliers lies in the context of supplier selection supplier criteria. This analysis allows not only to identify a clear

Table 1
Summary of previous researches on green supplier selection methods.

Author(s) Methodology applied Application domain

Lu et al. (2007) AHP electronics industry

Tuzkaya et al. (2009) F-ANP and PROMETHEE manufacturing industry
Lee et al. (2009) Delphi and Fuzzy extended AHP high-tech electronics industry
Hsu and Hu (2009) ANP electronic industry
Awasthi et al. (2010) F-TOPSIS logistics
Kuo et al. (2010) ANN and DEA digital camera manufacturing in Taiwan
Kannan et al. (2013) F-AHP, F- TOPSIS and MOLP automobile manufacturing company
Shen et al. (2013) Fuzzy TOPSIS auto industry
Kannan et al. (2014) Fuzzy TOPSIS Brazilian electronics company
Zhao and Guo (2014) F-Entropy and TOPSIS thermal power equipment
Rostamzadeh et al. (2015) Intuitionistic fuzzy VIKOR manufacturing industry
Azadnia et al. (2015) Rule-based weighted fuzzy method, F-AHP and MOMP packaging films in food industry
Hashemi et al. (2015) ANP and improved GRA automotive industry
Kannan et al. (2015) Fuzzy Axiomatic design plastic manufacturing
Fallahpour et al. (2016) Date envelopment analysis, Gene expression programming textile company
Yazdani et al.(2016) QFD and SWARA automobile industry
Govindan and Sivakumar (2016) Fuzzy TOPSIS and MOLP paper industry
Luthra et al. (2017) AHP and VIKOR automobile industry -India
 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740 3731

hierarchical structure for the all the relevant sustainable factors and to identify the right sustainable factors and sub-factors. Two kinds
sub-factors, but also to weight the decision criteria based on the of factors are collected as a result of these consultations: customer
importance given to customer requirements. requirements (CRs) and technical requirements (TRs).
Attentions to integrate several techniques and convey a mean- The CRs are the costumer variables, that is, the criteria on which
ingful logic for assessing suppliers’ performance are enhancing the costumers will base their choice. The TRs include all the criteria
(Labib, 2011). Several contributions to green supplier selection have to rank the candidate suppliers, that is, all the criteria that must be
been obtained by employing MCDM tools (Bottani and Rizzi, 2008; considered for the CRs to be satisfied. The TRs are the HOWs in QFD.
Chen and Wang, 2009; Awasthi et al., 2010; Tavana et al., 2016).
However, developing supplier selection decision making ap-
proaches through integrated methods can be very synergic since 4.2. Phase II: weighting customer and technical requirements
every specific method involves generic functions that allow for
stable solutions only if suitably integrate together. This phase consists of two sub-phases as:
As it is a typical process for every MCDM problem, selecting the
best suppliers is underlined to some essential objectives including
determining the degree of importance (weight) of the selection 4.2.1. Phase II.1
criteria and evaluating the suppliers respecting to these criteria The CRs are assigned a global weight using DEMATEL method to
(Ordoobadi, 2009). In this study, we develop a case-focused model acquire a perceptive relationship between the supplier selection
where these objectives are attained through an integrated MCDM criteria and CRs causally and visually.
model. Firstly, all the relevant factors/criteria assigned to suppliers
and customers are defined. The main factors are interpreted as
costumer requirements (CRs) and used to identify the technical 4.2.2. Phase II.2
requirements (TRs) necessary to weight the suppliers. Secondly, The normalized prominence values of DEMATEL are considered
DEMATEL and QFD are integrated to determine the weights of all as the weights of the respective CRs which will be further used for
the criteria introduced. To the best of our understanding, our choice QFD-based analysis. These weights are used in QFD to weight the
of handling DEAMETL to evaluate the interaction relationship be- TRs.
tween different variables of a complicated system to establish For the sake of completeness, recall that QFD transformations
direct and indirect causal relationships and influence levels among are usually represented by a matrix, known as house of quality
the customer variables through is also novel to the literature. (HoQ). This matrix expresses the relationship between the CRs
Thereafter, QFD is used to build a central relationship matrix in (WHATs) and the TRs (HOWs) incorporating the following items: A)
order to identify degree of relationship between each pair of sup- WHATs matrix, B) HOWs matrix, C) relationship matrix between
plier selection criteria and CRs. Then, pairwise comparison matrices WHATs and HOWs, D) relative importance or weights of WHATs, E)
are composed to evaluate each supplier with respect to each sup- interrelationship between HOWs, and F) weights of HOWs (Tang
plier selection criterion to construct corresponding supplier rating et al., 2005). More details on how QFD works has been given in
matrix. Finally, MOORA and COPRAS methods are applied and then previous comment (Fig. 1).
compared to select the best alternative green supplier. In the first sub-phase, a questionnaire is conducted by pur-
chasing department experts to collect the data to input in the
4. Proposed integrated framework for green supplier comparison matrices of DEMATEL and, hence, determine the global
selection weights of all the CRs. In the second sub-phase, another ques-
tionnaire is conducted to construct the HoQ matrix where the CRs
The main contribution of our approach from the technical are connected to the TRs. In this questionnaire, judgments are
viewpoint is the way it combines DEMATEL and QFD. One of the requested on how each decision criterion is influenced by the
essential tasks in MCDM modeling is seeking new and logical ways customers’ sub-factors and vice versa. The relationships composing
to weight decision factors (attributes). Usually, fuzzy linguistic this matrix are evaluated using a 1,3,6, and 9 scale.
variables, AHP, ANP, and entropy are employed to determine the
weights of all the factors. However, in many decision problems the
reliability of the decision criteria is strictly dependent on the
stakeholders and customers’ preferences. The DEMATEL-QFD phase
evaluation
Supplier

criteria

of the proposed method provides a simple to implement costumer-

dependent weighting method for decision criteria, which plays a
fundamental role in situations where the satisfaction of external
stakeholders and customers enter the decision process. All the
requirements
requirements

methods are appeared in Appendix for more information. To figure

Weights of
Customer

customer

out the whole proposed framework, see the following phases Relationship matrix
(Proposed four-phase method):
In this section we outline the phases composing the integrated
framework that we propose for solving the green supplier selection
of criteria
Weights

problem considered in the case study.

4.1. Phase I: identifying all relevant sustainable factors

(customer and technical requirements)
Normalized
weights for
criteria

In this phase, the relevant green and environmental factors/

criteria are selected considering the cited literature and the specific
features of the company of the case under analysis. Experts from
the company are consulted to gather information and data in order Fig. 1. QFD process for green supplier selection problem.
3732 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740

4.3. Phase III: ranking the suppliers per each technical recyclable materials, establishing automatic pollution control sys-
requirement tems using green and efficient energies, increasing top manage-
ment commitment etc. Therefore, to assure that all suppliers
In this phase, a questionnaire is conducted by purchasing perform appropriately, it is required to establish a proper supplier
department experts to determine the performance rating of selection methodology. In this paper, to build a comprehensive
candidate suppliers. The data collected using this questionnaire is supplier selection structure, eight CRs and seven evaluation criteria
used for the comparison matrices relative to each TR. The questions are identified from the literature survey. Financial stability (FS),
aim at establishing how much one of the suppliers is preferred to environmental management systems (EMS), waste disposal pro-
another one with respect to each specific decision criterion (TR). gram (WDP), management commitment (MC), quality control
The pairwise comparisons are carried out using again the 1e9 scale. systems (QCS), manufacturing (M), facility (F) and reverse logistics
(RL) are considered as CRs, whereas, quality adaptation (QD), price
4.4. Phase IV: ranking the suppliers with COPRAS and MOORA (P), energy and natural resource consumption (ENRC), delivery
speed (DS), green design (GD), re-use and recycle rate (RRR) and
In this phase, COPRAS and MOORA (See Appendix) are applied production planning (PP) are recognized as supplier selection
to rank the candidate suppliers. To implement these methods, both criteria. Among the seven criteria, P and ENRC are considered as
the weights of all the TRs from phase two and the supplier rankings cost criteria and rest of all is regarded as beneficial criteria.
(one per each of the TRs) obtained in phase three are needed. The company has identified alternative suppliers for the anal-
For the sake of completeness, recall that the solution process of ysis. Among suppliers, they provide packaging raw materials, raw
both COPRAS and MOORA starts with a t
n matrix whose generic milk and, supply milk powders and additional ingredients as sugar,
element is the performance rating of the k-th supplier (k ¼ 1,…,t) salt, stabilizer and essence. Among these suppliers A5, A2, A7 are in
upon the j-th decision criterion (j ¼ 1,…,n). The elements of this business for years and they are very flexible in delivery, cooperation
matrix are then normalized and used to calculate the total perfor- and collaboration in rejection of cartoons, nylons and plastic
mance index of each supplier. The two methods differ in the glasses. They are planning to stablish a waste management system
normalization procedure and in the way the evaluation of the total in their company. A2 is a supplier with modern and high tech fa-
performance of the single suppliers is carried out. COPRAS is cilities and very respectful human resource system. Although the
applied to rank the suppliers. The resulting ranking is used to verify cost of their products seems higher than similar suppliers in their
and validate those obtained in Phase IV. range, its capability convinces purchasing manager to keep contact
In MCDM problems, it is customary to compare the results ob- them. A1 and A6 and A10 produce packaging wraps which are used
tained by using a second method as a “mediator”. In this study we for covering or forming the final products like sleeve and shrink,
use the COPRAS method due to its capability in managing complex paper labels, vacuum films, Aluminum foils, laminate foils, sealed
decision systems. Developed by Zavadskas et al., 1994 and caps and PET caps, three layer and laminated cartons. A3 and A9 are
Zavadskas et al. (2007), this method has the ability of accounting in charge of supplying raw milk. This is the key raw material for
for both positive (beneficial) and negative (non-beneficial) criteria, dairy products, and due to the quality sensitivity and high deteri-
which can be assessed independently within the evaluation pro- oration rate; it needs the specific infrastructure of supply, trans-
cess. The most important feature that makes the COPRAS method portation, and storage. Mostly quality of products like posturized
superior to other methods is that it can be employed to estimate the and homogenized milk is measured by the quality of raw milk, and
utility of alternatives indicating the extent to which one alternative the amount of vitamins and proteins. A8 supplier delivers good
is better or worse than the other alternatives (Hashemkhani Zolfani quality raw milk among others. It should be expressed that Milk
and Bahrami, 2014). Further details on COPRAS and MOORA powder products are widely used in dairy industry. Milk powder is
(Brauers and Zavadskas, 2011) are given in Appendix C and D. generally consumed as a substitute of fresh milk and, in confec-
tionaries and bakeries to provide desired flavor and color, and to
5. Case study: selecting green supplier for an Iranian dairy enhance nutritional values. Whey Powder is used to produce
company products such as snack and chips.
Iran is faced to new paradigm and strategy about the environ-
5.1. Short introduction of case study ment. In climate change conference (2015) in Paris, Iran committed
and dedicated a policy for decreasing CO2 up to 12% till 2030. These
The proposed framework for green supplier selection is imple- days selecting green suppliers as a bigger vision for the future seem
mented within procurement and logistics department of Kalleh more necessary in comparison to the past. This fact is not far from
Dairy Company, established in 1991 in Amol, Iran. It has with more reality because processes of so many actions in Iran aren't profes-
than 23 years of experience in diverse types of dairy, meat, and sionally and compatible with the environment. Technologies are
beverage and food products. Kalleh dairy products contain classi- almost old and new investments should be considered for general
fication of UHT processed milk, yoghurt, cheese, dairy drinks, in- economic structure of Iran. In the near future companies will have
dustrial dry milk, industrial powders production, buttermilk and to be well-matched with new regulations and standards in the near
desserts which compose more than 150 types of products. Pro- future. Kalleh Dairy Company as one the most prestigious com-
curement segment of Kalleh Dairy Company, called POLE, is in panies in Iran would be like placed as one of pioneers in this
charge of providing wide range of raw materials and packaging at framework and structure.
the right time, with the least price and the desired quality.
Furthermore, POLE is responsible for purchasing the whole equip- 5.2. Implementation
ment required for the company, including machinery, spare parts,
technical parts, laboratory equipment and material-handling In view of the fact that selecting the best supplier along with its
equipment. To meet the production requirements, POLE deals performance evaluation is actually an intricate multi-criteria
with a wide range of domestic and foreign suppliers. It also fulfills a problem, so it is not appropriate to presume the elements sup-
complete procurement process, involved sourcing, purchasing, plier evaluation system as independent. As, all of the eight identi-
importing and customs releasing, warehousing, maintaining and fied CRs are deemed to be vital, hence, it becomes indispensable to
delivering of goods. The company is encouraging suppliers to utilize find the most important requirement of the evaluation system and
 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740 3733

measure the relationships among them. In order to achieve this, Table 5

DEMATEL is used for envisaging the insightful relationships be- Computation of vectors D and R.

tween the CRs causally and visually. Following the DEMATEL steps CR Dk Rk
as described in Appendix, the relationships among different CRs are FS 0.1636 0.1410
achieved using the integer scale as explained earlier. After the in- EMS 0.0890 0.2070
terrelationships are measured, the initial direct-relation matrix (A) WDP 0.0861 0.1731
is produced, as shown in Table 2. The matrix A is an 8
8 matrix, MC 0.2932 0.0228
QCS 0.1508 0.1405
obtained by pair-wise comparisons in terms of influences and di-
M 0.1090 0.1843
rections between the CRs. F 0.1295 0.1079
From matrix A of Table 2, the normalized direct-relation matrix RL 0.0966 0.1414
(X) is estimated as given in Table 3. Then, total-influence matrix (T)
is derived as shown in Table 4. Now, the sums of rows and columns
as represented by vectors D and R respectively are computed and Table 6
are shown in Table 5. The causal diagram, as exhibited by Fig. 4, is Total and net effects for each green CR.
obtained by plotting the dataset of Table 6. The (D þ R) and (D-R) CR DþR D-R Group
values of Table 6 represent the degree of total influence levels
FS 0.3046 0.0226 Cause
(central roles) and the degree of net influence levels respectively, EMS 0.2961 0.1180 Effect
where the positive values indicate that it will influence other re- WDP 0.2592 0.0869 Effect
quirements more than any other requirement influences it. Table 6 MC 0.3160 0.2704 Cause
indicates that MC requirement has the largest net influence level, QCS 0.2913 0.0104 Cause
M 0.2934 0.0753 Effect
followed by FS.
F 0.2374 0.0216 Cause
Now, looking at the causal diagram of Fig. 2, it is apparent that RL 0.2380 0.0448 Effect
CRs are visually segregated into the cause and effect groups. The
cause group compiles of FS, MC, QCS and F, while the effect group
composes EMS, WDP, M and RL. It is convincingly evident that FS,
relation to others for having maximum D þ R value followed by FS.
MC, QCS and F requirements are the main driving factors for EMS,
Furthermore, MC is also the most persuading factor due to its
WDP, M and RL. Among these eight CRs, MC is recognized as the
highest D-R value. Thus, MC plays a major role in the supplier
most significant one because it has the maximum intensity of
evaluation problem, and it has the utmost effect on the others. On
the converse, EMS is very much influenced by the other re-
quirements, having the lowest (D-R) value. The threshold value (a)
Table 2
is derived from the mean of elements in matrix T, which is calcu-
The initial direct-relation matrix (A).
lated using Eq. (7). The obtained a value is 0.0175. The values of tij in
CR FS EMS WDP MC QCS M F RL Table 4 which are greater than a value (0.0175), are shown as t*ij
FS 0 2 2 1 2 3 3 2 representing the interaction between two CRs, e.g. the value of t12
EMS 1 0 1 1 1 2 2 0 (0.0324) > a (0.0175), the arrow in the diagraph, as shown in Fig. 3,
WDP 0 3 0 0 2 1 0 2
is drawn from FS to EMS. This digraph reveals contextual re-
MC 4 4 4 0 4 4 4 3
QCS 3 3 3 0 0 3 0 2 lationships among the components of the system.
M 3 2 1 0 1 0 1 2 The weights of the CRs are calculated by normalizing the values
F 1 2 2 0 2 3 0 2 of prominence vector (D þ R) of Table 6 and are shown in Table 7.
RL 1 3 3 0 1 1 0 0 From these values, it is observed that as MC is the most influencing
factor and it has the highest weight among other CRs. Following the
process of QFD (Appendix B), a central relationship matrix is con-
Table 3 structed which exhibits the effects and relations between each pair
Normalized direct-relation matrix (X) of green CRs.
of CR and corresponding supplier selection criteria. This task is
CR FS EMS WDP MC QCS M F RL handled by decision team based on their knowledge on suppliers
FS 0 0.0194 0.0194 0.0097 0.0194 0.0291 0.0291 0.0194 which is exhibited in Table 8. Actually this step responds to the
EMS 0.0097 0 0.0097 0.0097 0.0097 0.0194 0.0194 0 question of how CR and supplier evaluation criteria interact and
WDP 0 0.0291 0 0 0.0194 0.0097 0 0.0194 influence through the values given by the experts. As seen in
MC 0.0388 0.0388 0.0388 0 0.0388 0.0388 0.0388 0.0291
Table 8, the values are assigned to the matrix which show how
QCS 0.0291 0.0291 0.0291 0 0 0.0291 0 0.0194
M 0.0291 0.0194 0.0097 0 0.0097 0 0.0097 0.0194 supplier selection criteria can satisfy each CR.
F 0.0097 0.0194 0.0194 0 0.0194 0.0291 0 0.0194 Having central relationship matrix and weights of CRs, the
RL 0.0097 0.0291 0.0291 0 0.0097 0.0097 0 0 weights of each supplier selection criteria is computed. The

Table 4
Total-relation matrix (T) of green CRs (tij* > 0.0175).

CR FS EMS WDP MC QCS M F RL

FS 0.0029 0.0324* 0.0226* 0.0103 0.0218* 0.0323* 0.0306* 0.0313*

EMS 0.0113 0.0018 0.0113 0.0100 0.0112 0.0212* 0.0204* 0.0017
WDP 0.0014 0.0306* 0.0016 0.0003 0.0201* 0.0112 0.0008 0.0201*
MC 0.0426* 0.0445* 0.0436* 0.0012 0.0425* 0.0442* 0.0414* 0.0333*
QCS 0.0307* 0.0322* 0.0311* 0.0006 0.0021 0.0313* 0.0018 0.0216*
M 0.0301* 0.0216* 0.0117 0.0006 0.0112 0.0023 0.0111 0.0209*
F 0.0118 0.0029 0.0213* 0.0099 0.0209* 0.0309* 0.0011 0.0118
RL 0.0107 0.0309* 0.0301* 0.0004 0.0110 0.0113 0.0010 0.0014
3734 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740

0.30 Table 7
0.2704 Weights of green CRs.

0.22 CR FS EMS WDP MC QCS M F RL

Weight 0.1362 0.1324 0.1159 0.1413 0.1303 0.1312 0.1062 0.1064

0.14
D-R

instance, Tables 9 and 10 show the pairwise comparison matrices

0.06
0.0216 0.0226
for alternative suppliers regarding ENRC and GD criteria respec-
0.0104
tively. For rest of the criteria, the same procedure is followed. At the
-0.02
-0.0448 end of this step, performance ratings of alternative suppliers and
-0.0753
-0.0869 criteria weights are integrated to compose the initial decision
-0.10 -0.1180
matrix, as shown in Table 11.
0.22 0.24 0.26 0.28 0.30 0.32 For the purpose of performance evaluation of the ten alternative
D+R suppliers, COPRAS and MOORA methods are now applied respec-
tively. At first, this supplier selection problem is solved using CO-
PRAS method and thereafter MOORA method is applied.
Fig. 2. DEMATEL causal diagram of green CRs. For the application of COPRAS method, the initial decision ma-
trix, as shown in Table 11, is the normalized matrix and the corre-
sponding weighted normalized decision matrix, is given in
α = 0.0175 Table 12. The normalization process is done to establish same scale
for all criteria for the purpose of comparability. Now using Equa-
FS
tions (12) and (13), the sums of the weighted normalized values are
EMS calculated for both the beneficial attributes (Pj) and non-beneficial
RL
attributes (Rj), as given in Table 13. Then, applying Eq. (14), the
relative significance or priority value (Qj) for each alternative sup-
plier is determined, as shown in Table 13. This table also exhibits
the value of quantitative utility (Nj) for each alternative on the basis
of which the complete ranking of the alternative suppliers is ob-
WDP
F
tained. The candidate suppliers are then arranged in descending
order of Nj values. The best choice of supplier for the Iranian dairy
company is S7. S2 is the second choice and the last choice is supplier
S5.
Now, the same supplier selection problem is solved using
QCS MOORA method. Table 14 shows the normalized performance
M MC
scores of the alternatives with respect to the considered criteria, as
obtained using Eq. (16). Then applying Eq. (17), the weighted
normalized values are computed, as shown in Table 15. This table
also exhibits the overall rating of benefit (Sþj) and cost criteria (Sj)
Fig. 3. DEMATEL diagraph for green supplier selection problem.
for all the alternative suppliers, determined employing Eqs. (18)
and (19) respectively. Finally, the overall performance index (Sj) is
calculated using Eq. (20) and a comparative ranking of the alter-
10
COPRAS native suppliers is obtained as
9 MOORA S7 > S2 > S3 > S1 > S4 > S8 > S9 > S6 > S10 > S5 when the Sj values are
8
arranged in descending order. It is observed that similar to COPRAS
7 method, suppliers S7 and S2 emerge out as the first and second
6 choices respectively. Also, COPRAS and MOORA methods confirm
Rank

5 that S10 and S5 are the worst suppliers. Moreover, to compare the
4 ranking performances of the two MCDM methods, Spearman cor-
3 relation coefficient (Chatterjee et al., 2014) is computed as 0.88
2 which reflects a strong agreement between the rankings provided
1 by the two methods. MOORA method is a tool which is recognized
0 as the process of simultaneously optimizing two or more conflict-
1 2 3 4 5 6 7 8 9 10 ing criteria subjected to certain constraints, while COPRAS method
ranks alternatives based on utility degree and their significances.
Supplier
Fig. 4 shows the comparative ranking preorder of the alternative
Fig. 4. Comparative ranking of suppliers. suppliers.
From the initial decision matrix of Table 11, it is observed that
supplier S7 and S2 outperforms most of the other alternative sup-
normalized weights of all criteria are obtained which are shown in pliers with respect to higher values of QD and RRR, and lower value
Table 8. From this table, it becomes clear that re-use and re-cycle of ENRC criteria. On the other hand, the main reason behind the
rate (RRR) is the most important supplier evaluation criterion underperformance of S5 supplier is its very low RRR value, although
among others. it has amazingly attractive values for QD and DS criteria. It is
With respect to each evaluation criterion, suppliers are now identified that this supplier has very less capability in terms of
rated using pairwise comparisons as stated before. It means seven energy and natural resource consumption (ENRC), delivery speed
pairwise comparisons are done for ten alternative suppliers. For (DS), green design (GD), re-use and recycle rate (RRR) and
 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740 3735

Table 8
QFD model for green supplier selection problem.

HOWs (CRs) WHATs (Criteria) Weight of CR

QD P ENRC DS GD RRR PP

FS 6 3 0.1362
EMS 3 3 3 0.1324
WDP 3 0.1159
MC 3 3 3 3 1 0.1413
QCS 3 6 1 3 0.1303
M 1 3 3 3 0.1312
F 3 3 3 0.1062
RL 6 0.1064
1.2120 1.2410 1.2710 1.4940 0.8330 2.0490 1.1030 9.2020
Normalized criteria weight 0.1320 0.1350 0.1380 0.1620 0.0900 0.2230 0.1200

Table 9
Pairwise comparison of green suppliers for ENRC criterion.

ENRC S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 Weights

S1 1 6 3 3 4 0.11 0.33 1 5 7 0.15

S2 0.167 1 0.25 7 0.5 9 1 2 3 4 0.108
S3 0.333 4 1 4 6 2 1 4 4 0.13 0.128
S4 0.333 0.143 0.25 1 0.14 6 3 0.17 3 1 0.058
S5 0.25 2 0.167 7 1 0.2 4 8 2 3 0.125
S6 9 0.111 0.5 0.167 5 1 6 0.17 8 0.25 0.15
S7 3 1 1 0.333 0.25 0.167 1 2 1 0.333 0.056
S8 1 0.5 0.25 6 0.125 6 0.5 1 6 2 0.086
S9 0.2 0.333 0.25 0.333 0.5 0.125 1 0.167 1 6 0.043
S10 0.143 0.25 8 1 0.333 4 3 0.5 0.167 1 0.097

Table 10
Pairwise comparison of green suppliers for GD criterion.

GD S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 Weights

S1 1 2 1 7 4 3 0.5 2 1 3 0.156
S2 0.5 1 4 0.2 1 0.143 2 0.333 5 2 0.082
S3 1 0.25 1 1 0.5 0.25 5 3 3 7 0.108
S4 0.143 5 1 1 3 5 6 2 2 1 0.132
S5 0.25 1 2 0.33 1 0.333 5 4 0.25 0.333 0.074
S6 0.333 7 4 0.2 3 1 7 0.5 4 0.25 0.134
S7 2 0.5 0.2 0.167 0.2 0.143 1 5 6 9 0.122
S8 0.5 3 0.333 0.5 0.25 2 0.2 1 3 1 0.062
S9 1 0.2 0.333 0.5 4 0.25 0.167 0.333 1 0.25 0.05
S10 0.333 0.5 0.143 1 3 4 0.111 1 4 1 0.08

Table 11 Table 12
Initial decision matrix for supplier evaluation. Weighted normalized matrix for COPRAS method.

Weight 0.132 0.135 0.138 0.162 0.090 0.223 0.120 Supplier QD P ENRC DS GD RRR PP

Supplier QD P ENRC DS GD RRR PP S1 0.0089 0.0089 0.0207 0.0159 0.0141 0.0253 0.0118
S2 0.0103 0.0103 0.0149 0.0221 0.0075 0.0382 0.0126
S1 0.068 0.066 0.150 0.098 0.156 0.114 0.098 S3 0.0206 0.0154 0.0176 0.0136 0.0097 0.0251 0.0157
S2 0.078 0.076 0.108 0.136 0.082 0.171 0.105 S4 0.0140 0.0188 0.0081 0.0120 0.0119 0.0187 0.0144
S3 0.157 0.114 0.128 0.083 0.108 0.113 0.131 S5 0.0135 0.0252 0.0173 0.0286 0.0067 0.0143 0.0068
S4 0.106 0.139 0.058 0.074 0.132 0.084 0.120 S6 0.0138 0.0112 0.0207 0.0083 0.0121 0.0210 0.0135
S5 0.103 0.187 0.125 0.176 0.074 0.064 0.057 S7 0.0180 0.0171 0.0077 0.0217 0.0110 0.0266 0.0137
S6 0.105 0.083 0.150 0.051 0.134 0.094 0.113 S8 0.0132 0.0110 0.0119 0.0098 0.0056 0.0242 0.0111
S7 0.137 0.127 0.056 0.133 0.122 0.119 0.114 S9 0.0070 0.0070 0.0059 0.0162 0.0045 0.0173 0.0075
S8 0.100 0.082 0.086 0.060 0.062 0.109 0.093 S10 0.0124 0.0100 0.0134 0.0142 0.0072 0.0120 0.0127
S9 0.053 0.052 0.043 0.100 0.050 0.078 0.063
S10 0.094 0.074 0.097 0.087 0.080 0.054 0.106

cumulative Qj or Sj of 20e80%) and this second group can be

production planning (PP). adjudged as an improvement target for suppliers S10 and S5 (cu-
A graphical view of the Pareto analysis results based on Qj and Sj mulative Qj or Sj of 0e20%).
values of the ten alternative suppliers is presented in Fig. 5 (a) and
(b) respectively for COPRAS and MOORA methods. From Fig. 5 (a) 5.3. Managerial tips
and (b), it is observed that suppliers S7 and S2 (with a cumulative Qj
or Sj of 80e100%) can be considered as the benchmarks for sup- The outcomes of this study are headlines for the Kalleh dairy
pliersS9, S3, S1, S4, S8 and S6, (which form the second group with a company to enhance the quality and deliver fitted product based on
3736 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740

Table 13 (a)
Pj, Rj, Qj and Nj values in COPRAS method. 0.12 100
Qj
Supplier Pj Rj Qj Nj Rank 0.11 Cumulative Qj % 80

Cumulative Qj%
S1 0.0761 0.0295 0.0992 84 5 0.1
S2 0.0906 0.0252 0.1177 99 2 60

Qj value
0.09
S3 0.0847 0.033 0.1054 89 4
S4 0.071 0.0269 0.0964 81 6 0.08 40
S5 0.0699 0.0425 0.086 73 10
S6 0.0687 0.0318 0.0901 76 8 0.07
20
S7 0.091 0.0248 0.1184 100 1 0.06
S8 0.064 0.023 0.0937 79 7
S9 0.0525 0.0129 0.1055 89 3 0.05 0

S5

S6
S8
S4
S1
S3
S9
S2
S7
S10
S10 0.0585 0.0234 0.0877 74 9

Supplier

Table 14
Normalized decision matrix for MOORA method. (b) 0.2 100
Sj
Supplier QD P ENRC DS GD RRR PP 0.18 Cumulative Sj % 80

Cumulative Sj %
S1 0.206 0.1943 0.4437 0.291 0.469 0.343 0.3036
0.16
S2 0.2364 0.2249 0.3197 0.4048 0.2472 0.5167 0.3239 60

Sj value
S3 0.4759 0.3357 0.3788 0.248 0.3228 0.3396 0.4048 0.14
S4 0.3225 0.4108 0.1733 0.2204 0.3948 0.2529 0.3698 40
S5 0.3114 0.5496 0.3724 0.5235 0.2225 0.1933 0.1756 0.12
S6 0.3192 0.2441 0.4439 0.1522 0.4009 0.2839 0.3475
S7 0.4146 0.3736 0.1655 0.3963 0.3661 0.3597 0.3526 0.1 20
S8 0.3042 0.2411 0.2562 0.1786 0.1867 0.3281 0.2867
S9 0.1607 0.1531 0.1262 0.2966 0.1503 0.2338 0.194 0.08 0

S5

S6
S9
S8
S4
S1
S3
S2
S7
S10
S10 0.2863 0.2179 0.2877 0.2595 0.2393 0.1625 0.3263

Supplier

green regulations to finally reward consumers. Moreover, the top-

ranked suppliers' evaluation scores can be outlined for the rest of Fig. 5. (a) Pareto analysis of Qj values for ten alternative suppliers. (b) Pareto analysis
suppliers to comprehend their weaknesses and fulfill logical of Sj values for ten alternative suppliers.
pattern for future plans. We have tried to develop a basis for
generous relation of company with its partners to reduce their
weaknesses. The findings of this study are argued and confirmed to Current work points out an integrated approach to address the real
the Kalleh and POLE executives. The ability to model the supplier's life supplier selection problem involving customer opinions and
performance over a set of criteria will offer Kalleh to launch a request to the core activities of suppliers.
capability building anatomy for productive management
competencies. 6. Conclusions
We believe that the proposed model is sufficient robust and
could be easily implemented in practices for multi-criteria deci- With the revolutionizing change in the state of environment,
sion-making problems. Managers can more effectively form their subsequent public pressure and environmental logistics, environ-
decision structure and detect the relative importance of their mental and social issues are becoming more important in man-
supplier attributes. The released model tries to aid managers to aging any business. GSCM is an approach to advance the
prioritize their supplier's development programs focusing on performance of the process and products in accordance with the
customer and external parameters delivering effective performance environmental requirements and regulations. GSCM envelops all
among suppliers. Evaluating the supplier performance effectively phases of product's life cycle from design, production and distri-
improves performance and behavior of suppliers regarding weak bution and whole supply chain to the use of products by the end
attribute and also allocate more credit to the stronger suppliers to users and its disposal at the end of product's life cycle. Role of
appreciate all practices to the next level. Although, many researches suppliers cannot be ignored.
previously investigated many techniques to undertake the supplier In view of this, the supplier selection criteria have to be rede-
selection problem, mostly are challenging to implement in real life. signed according to green perspective. Hence, the main objective of

Table 15
Weighted normalized matrix and ranking of supplier using MOORA method.

Supplier QD P ENRC DS GD RRR PP Sþ j S j Sj Rank

S1 0.027 0.026 0.061 0.047 0.042 0.076 0.036 0.230 0.087 0.142 4
S2 0.031 0.030 0.044 0.066 0.022 0.115 0.039 0.273 0.074 0.199 2
S3 0.063 0.045 0.052 0.040 0.029 0.076 0.049 0.256 0.098 0.159 3
S4 0.042 0.055 0.024 0.036 0.036 0.056 0.044 0.215 0.079 0.135 5
S5 0.041 0.074 0.051 0.085 0.020 0.043 0.021 0.210 0.126 0.085 10
S6 0.042 0.033 0.061 0.025 0.036 0.063 0.042 0.208 0.094 0.114 8
S7 0.055 0.050 0.023 0.064 0.033 0.080 0.042 0.274 0.073 0.201 1
S8 0.040 0.033 0.035 0.029 0.017 0.073 0.034 0.193 0.068 0.125 6
S9 0.021 0.021 0.017 0.048 0.014 0.052 0.023 0.158 0.038 0.120 7
S10 0.038 0.029 0.040 0.042 0.022 0.036 0.039 0.177 0.069 0.108 9
 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740 3737

this paper is set to resolve the issue of evaluating and ranking green 2 3
suppliers utilizing an integrated formulation. This paper evaluates 0 a12 ::: a1j ::: a1n
6 a21 0 ::: a2j ::: a2n 7
and elucidates the interaction relationships and impact levels be- 6 7
tween the CRs and supplier selection criteria and also to determine A ¼ 6
6 ::: ::: ::: ::: ::: ::: 7
7 (1)
4 ::: ::: ::: ::: ::: ::: 5
the weights of the CRs and supplier selection criteria respectively.
DEMATEL facilitates to build up the underlying relationship dia-
an1 an2 ::: anj ::: 0
gram, dividing the considered CRs into cause and effect groups. Step 2: Formation of the normalized direct-relation matrix (X):
Based on the results, it is recommended that the organization After the generation of the direct-relation matrix (A), the
should focus on maintaining product and process quality, energy normalized matrix (X) is achieved using Eq. (2). Each element in
and natural resource consumption, green design with increased re- matrix X ranges from 0 to 1.
use and recycle rate according to different environmental regula-
tions. It is figured out strong management commitment is the key X ¼ k$A (2)
driving force for sustainable developments in infrastructure, facility
and quality. It is also the responsibility of the management to where
strongly focus on improved manufacturing and reverse logistics
processes and production planning activities for maintaining an 1
k¼ !; i; j ¼ 1; 2; :::; n (3)
efficient GSCM system. Pn
The findings of this paper put forward some important insights max j¼1 aij
1in
on different attributes which considerably contribute to supplier
performance and efficiencies so that inefficient suppliers can focus Step 3: Computation of the total-relation matrix (T):
on those attributes to improve their performances. From the Pareto The total-relation matrix (T) is obtained by Eq. (4), in which I
analysis, it is observed that the inefficient suppliers have to increase denotes the identity matrix. Each element (tij) of this matrix sym-
their re-use and recycle rate, condense energy and natural resource bolizes the indirect influences that ith criterion imparts on jth
consumption and increase delivery performance and should adopt criterion, and the matrix T reveals the total relationship between
the benchmark policies and techniques of the Pareto efficient each pair of decision variables.
suppliers with respect green design aspects, price, quality adapta-  
tion principles, suitable waste disposal program and appropriate T ¼ tij n
n ; i; j ¼ 1; 2; :::; n
production planning. It is expected that the proposed integrated
framework will serve as a vital tool in devising environmentally
T ¼ X þ X 2 þ X 3 þ :::: þ X k
conscious SCM system which will enable organizations to become  h i
more competitive, while achieving sustainable development. ¼ X I þ X þ X 2 þ ::: þ X k1 ðI  XÞðI  XÞ1
The proposed supplier selection approach enables the DMs to  
better understand the complex relationships of the relevant attri- ¼ X I  X k ðI  XÞ1 Then;
butes in the decision-making process which may subsequently
improve the reliability of the decision and also contributes in pro-
moting sustainable development to some extent. The methodology T ¼ XðI  XÞ1 T; when k/∞; X k ¼ ½0n
n
can easily be adjusted to solve other decision-making problems
involving any number of alternatives and any number of criteria. T ¼ XðI  XÞ1 (4)
Step 4: Determination of the sums of rows and columns of
Acknowledgement matrix T:
In the total-relation matrix T, the sum of rows and sum of col-
Authors are thankful of anonymous reviewers. Also special umns are represented by vectors D and R, as derived using Eqs. (5)
thanks for supports from Kalleh dairy complex especially to Mr. and (6) respectively.
Nikzad. Ask, Mr. Moghadassi, Mrs. Shirvani, Mr. Zare and all the
2 3
others who have contributed in this work. X
n
Di ¼ 4 tij 5 ¼ ½ti n
1 ; i ¼ 1; 2; :::; n (5)
j¼1 n
1
APPENDIX
" #
A) DEMATEL method X
n  
Rj ¼ tij ¼ tj n
1 ; j ¼ 1; 2; :::; n (6)
i¼1 1
n
The application of DEMATEL method consists of the following
seven steps (Tamura and Akazawa, 2005; Tzeng et al., 2007; Ranjan Step 5: Setting a threshold value (a):
et al., 2014). It presumes a system restraining a set of components Since matrix T provides information on how one factor affects
C ¼ {C1,C2, …,Cn}, with pair-wise relations that can be assessed. another, it thus becomes essential for the DM to set a threshold
Step 1: Generation of the direct-relation matrix (A) by scores: value (a) for elucidating the structural relation among criteria while
At first, the DM indicates the relationship between the sets of simultaneously keeping the intricacy of the entire system to a
paired criteria that signifies the direct effect that each ith criterion convenient level. This threshold value is generally determined by
exerts on each jth criterion, as specified by an integer score ranging experts in order to set up the minimum value of influence level. An
from 0 to 4, representing no influence (0), low influence (1), me- influence relationship between two elements is excluded from the
dium influence (2), high influence (3) and very high influence (4). map if their correlation value in matrix T is smaller than a and only
As a result of these assessments, a direct-relation matrix (A) is the effects greater than the set a value are chosen and shown in the
obtained in the form of an n
n matrix, in which the individual digraph. In this paper, the value of a is computed from the average
element (aij) denotes the degree to which ith criterion affects jth of the elements in matrix T, as computed using Eq. (7), where N is
criterion and n denotes the total number of criteria. the total number of elements in matrix T.
3738 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740

Pn Pn   the priority value to the CRs, a 5 point scale representing not

i¼1 j¼1 tij important (1), important (2), much more important (3), very
a¼ (7)
N important (4) and most important (5) is set.
Step 6: Development of a causal diagram: Step 5: Development of the relationship matrix (C) or the HoQ
The causal diagram is a classification of the degree of each cri- by judging the degree of impact between HOWs and WHATs
terion. It shows the criterion which be easily classed as either a expressing how much each HOW affects each WHAT using an
passive one or active one. The horizontal axis vector (Dk þ Rk) appropriate scale (here, a four point scale as 1, 3, 6, and 9 addressing
named ‘prominence’ is computed by adding D to R while weak, moderate, strong and very strong relationships has been
k ¼ i ¼ j ¼ 1 which reveals how much importance the criterion has. used).
Similarly, the vertical axis (Dk-Rk) named ‘relation’ is calculated by Step 6: Once the HoQ matrix is developed, compute the overall
subtracting D from R, which divides the criteria into a cause group priorities of TRs signifying the synthesized importance of the TRs or
and an effect group. The “prominence axis” of the causal diagram HOWs.
indicates the criterion which affects others and be affected by In QFD, the output of each phase (HOWs) is transformed into the
others. The “relation axis” can divide the criteria into the causal and inputs of the next phase (new WHATs). Advantages of applying QFD
effect groups. Generally, when the value of “relation” is positive, the can be counted as; higher customer satisfaction, shorter lead time,
criterion belongs to the causal group and if the value is negative, the better flexibility, quality promotion, reduction of time to market,
criterion belongs to the effect group. Hence, causal diagrams can and knowledge preservation (Khademi-Zare et al., 2010; Ignatius
visualize the complicated causal relationships and interaction in- et al., 2016).
fluence levels between the criteria into a visible structural model,
providing valuable insights for problem solving. Further, with the
help of the causal diagram, the DM can find the driving variables of
C) COPRAS method
the core problem in a complicated system, and plan for suitable
decisions to solve the problem in accordance with attribute type
The computational steps as involved in COPRAS method-based
and influence level.
analysis are now presented below (Zavadskas et al., 1994):
Step 7: Calculation of criteria weights:
Step 1: Let D is a decision matrix, containing the performance
The criteria weights (Ci) are calculated by normalizing the
rating of m number of alternatives with respect to n number of
prominence vector (Dk þ Rk) in which the sum of normalized
criteria, as shown below.
weights equals to 1.
2 3
x11 x12 :::: x1n
B) QFD model 6 x21 x22 ::::: x2n 7
6
D¼4 7 (8)
:::: ::: ::: ::: 5
In a general QFD model, the following items are incorporated in xm1 xm2 :::: xmn
the HoQ, as shown in Fig. 6.
wherexij is the rating of ith decision criteria on jth alternative,
 A: WHATs matrix whereas, m is the number of alternatives and n is the number of
 B: HOWs matrix criteria.
 C: relationship matrix between WHATs and HOWs Step 2: Normalize the decision matrix using Eq. (9).
 D: relative importance or weights of WHATs
 E: interrelationship between HOWs xij
 F: weights of HOWs rij ¼ Pm ; j ¼ 1; 2; :::::; m i ¼ 1; 2; :::; n (9)
j¼1 xij

The general steps of QFD model implementation are as follows: Step 3: Calculate the weighted normalized decision matrix as
Step 1: Identify the WHATs. follows: where wi includes the weights of criteria and given by
Pn
Step 2: Identify HOWs. TRs are specified as the HOWs in the HoQ i¼1 wi ¼ 1;
and positioned on the area marked as ‘B’ of Fig. 6.
Step 3: Development of HOWs matrix (E) indicating inner vij ¼ wi
rij ; j ¼ 1; 2; :::::; m i ¼ 1; 2; :::; n (10)
dependence among the HOWs.
Step 4: Priority weights are assigned to the CRs. For assigning The sum of dimensionless weighted normalized values of each
criterion is always equal to the weight for that criterion.

X
m
vij ¼ wi (11)
j¼1

Thus, it can be said that the weight, wi of ith criterion is pro-

portionally distributed among all the alternatives according to their
weighted normalized value vij.
Step 4: Calculate the sums of weighted normalized values for
both the beneficial (Pj) and non-beneficial attributes(Rj) using the
following equations:

X
k
Pj ¼ vij (12)
i¼1

Fig. 6. General QFD model. wherek is the number of criteria to be maximized.

 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740 3739

form of an ordinal ranking of the alternatives.

X
nk
Step 5: The ranking for each alternative is obtained by arranging
Rj ¼ vij (13)
the Sj values in descending order. It means higher values of Sj
i¼1
exhibit better priority order and would be preferred.
Where ðn  kÞ is the number of criteria to be minimized.
Step 5: Determine the relative significances or priorities of the References
alternatives as follows:
Aliev, R.A., 2013. Fundamentals of the Fuzzy Logic-based Generalized Theory of
Pm Decisions. Springer-Verlag, Berlin. http://dx.doi.org/10.1007/978-3-642-34895-
j¼1 Rj 2.
Qj ¼ Pj þ Pm 1 ; (14) Awasthi, A., Kannan, G., 2016. Green supplier development program selection using
Rj j¼1 Rj NGT and VIKOR under fuzzy environment. Comput. Industrial Eng. 91, 100e108.
Awasthi, A., Chauhan, S.S., Goyal, S.K., 2010. A fuzzy multicriteria approach for
Step 6: Calculate the quantitative utility (Nj) for jth alternative. evaluating environmental performance of suppliers. Int. J. Prod. Econ. 126 (2),
The degree of an alternative's utility which leads to a complete 370e378. http://dx.doi.org/10.1016/j.ijpe.2010.04.029.
Azadnia, A.H., Saman, M.Z.M., Wong, K.Y., 2015. Sustainable supplier selection and
ranking of the candidate alternatives is determined by comparing
order lot-sizing: an integrated multi-objective decision-making process. Int. J.
the priorities of all the alternatives with the most efficient one and Prod. Res. 53 (2), 383e408. http://dx.doi.org/10.1080/00207543.2014.935827.
can be denoted as below: Bai, C., Sarkis, J., 2010. Green supplier development: analytical evaluation using
rough set theory. J. Clean. Prod. 18 (12), 1200e1210.
Bhattacharya, A., Mohapatra, P., Kumar, V., Dey, P.K., Brady, M., Tiwari, M.K.,
Qj
Nj ¼
100% (15) Nudurupati, S.S., 2014. Green supply chain performance measurement using
Qmax fuzzy ANP-based balanced scorecard: a collaborative decision-making
approach. Prod. Plan. Control 25 (8), 698e714.
whereQmax is the maximum relative significance value. These utility Bottani, E., Rizzi, A., 2008. An adapted multi-criteria approach to suppliers and
products selectiondAn application oriented to lead-time reduction. Int. J. Prod.
values of the alternatives range from 0% to 100%. Thus, this Econ. 111 (2), 763e781.
approach allows for evaluating the direct and proportional Brauers, W.K.M., Zavadskas, E.K., 2006. The MOORA method and its application to
dependence of significance and utility degree of the considered privatization in a transition economy. Control Cybern. 35 (2), 445.
Brauers, W.K.M., Zavadskas, E.K., 2009. Robustness of the multi-objective MOORA
alternatives in a decision-making problem having multiple criteria, method with a test for the facilities sector. Technol. Econ. Dev. Econ. 2,
their weights and performance values of the alternatives with 352e375.
respect to all the criteria. Brauers, W.K.M., Zavadskas, E.K., 2011. MULTIMOORA optimization used to decide
on a bank loan to buy property. Technol. Econ. Dev. Econ. 1, 174e188.
Chan, H.K., Wang, X., 2013. Fuzzy Hierarchical Model for Risk Assessment: Princi-
ples, Concepts, and Practical Applications. Springer-Verlag, London. http://
D) MOORA method dx.doi.org/10.1007/978-1-44715043-5.
Chatterjee, P., Mondal, S., Chakraborty, S., 2014. A comprehensive solution to
The step by step application procedure of MOORA (Brauers and automated inspection device selection problems using ELECTRE methods. Int. J.
Technol. 2, 193e208. http://dx.doi.org/10.14716/ijtech.v5i2.302.
Zavadskas, 2006) method is explained below: Chen, L.Y., Wang, T.C., 2009. Optimizing partners' choice in IS/IT outsourcing pro-
Step 1: To have a dimensionless and comparable element in the jects: the strategic decision of fuzzy VIKOR. Int. J. Prod. Econ. 120 (1), 233e242.
evaluation process, the ratio system of MOORA method first com- Cuthbertson, R., Piotrowicz, W., 2008. Supply chain best practices-identification and
categorization of measures and benefits. Int. J. Prod. Perform. Manag. 57 (5),
putes the normalized decision matrix, as shown below: 389e404.
Dries, L., Germenji, E., Noev, N., Swinnen, J.F., 2009. Farmers, vertical coordination,
xij and the restructuring of dairy supply chains in Central and Eastern Europe.
rij ¼ Pm 2
(16) World Dev. 37 (11), 1742e1758.
j¼1 xij Fallahpour, A., Amindoust, A., Antuchevi _ J., Yazdani, M., 2016. Nonlinear
ciene,
genetic-based model for supplier selection: a comparative study. Technol. Econ.
Step 2: Determine the weighted normalized matrix as: Dev. Econ. 1e18.
Gimenez, C., Tachizawa, E.M., 2012. Extending sustainability to suppliers: a sys-
vij ¼ wi
rij (17) tematic literature review. Supply Chain Manag. An Int. J. 17 (5), 531e543.
Glover, J.L., Champion, D., Daniels, K.J., Dainty, A.J.D., 2014. An Institutional Theory
Step 3: Compute the overall rating of benefit and cost criteria for perspective on sustainable practices across the dairy supply chain. Int. J. Prod.
Econ. 152, 102e111.
all alternatives implementing the following equations: Govindan, K., Sivakumar, R., 2016. Green supplier selection and order allocation in a
low-carbon paper industry: integrated multi-criteria heterogeneous decision-
X
n making and multi-objective linear programming approaches. Ann. Operations
Sþ
j ¼ vij ; i2J Max (18) Res. 238 (1e2), 243e276.
i¼1 Govindan, K., Khodaverdi, R., Jafarian, A., 2013. A fuzzy multi criteria approach for
measuring sustainability performance of a supplier based on triple bottom line
approach. J. Clean. Prod. 47, 345e354.
where JMax is related to the beneficial criteria where higher values
Govindan, K., Rajendran, S., Sarkis, J., Murugesan, P., 2015. Multi criteria decision
are desirable. making approaches for green supplier evaluation and selection: a literature
For non-beneficial criteria, Eq. (19) is changed to: review. J. Clean. Prod. 98, 66e83.
Hashemi, S.H., Karimi, A., Tavana, M., 2015. An integrated green supplier selection
X
n approach with analytic network process and improved Grey relational analysis.
S
j ¼ vij ; i2J Min (19) Int. J. Prod. Econ. 159, 178e191.
Hashemkhani Zolfani, S., Bahrami, M., 2014. Investment prioritizing in high tech
i¼1 industries based on SWARA-COPRAS approach. Technol. Econ. Dev. Econ. 20 (3),
534e553.
where JMin is related to the non-beneficial criteria for which lower Hashemkhani Zolfani, S., Chen, I.S., Rezaeiniya, N., Tamosaitiene, _ J., 2012. A hybrid
values are preferable. MCDM model encompassing AHP and COPRAS-G methods for selecting com-
pany supplier in Iran. Technol. Econ. Dev. Econ. 18 (3), 529e543.
Step 4: Obtain the overall performance index by mutually sub- Hsu, C.W., Hu, A.H., 2009. Applying hazardous substance management to supplier
tracting the overall ratings for beneficial and cost criteria using the selection using analytic network process. J. Clean. Prod. 17 (2), 255e264. http://
following formula: dx.doi.org/10.1016/j.jclepro.2008.05.004.

Ignatius, J., Rahman, A., Yazdani, M., Saparauskas, J., Haron, S.H., 2016. An integrated
fuzzy ANPeQFD approach for green building assessment. J. Civ. Eng. Manag. 22
Sj ¼ Sþ 
j  Sj (20) (4), 551e563.
Jones, A., 2002. An environmental assessment of food supply chains: a case study on
The Sj values indicate cardinal scales which can be compared in dessert apples. Environ. Manag. 30 (4), 560e576.
3740 M. Yazdani et al. / Journal of Cleaner Production 142 (2017) 3728e3740

€
Kahraman, C., Oztayşi, B., 2014. In: Supply Chain Management under Fuzziness: relational analysis approach. J. Clean. Prod. 86, 343e359.
Recent Developments and Techniques, vol. 2014. Springer-Verlag, Berlin. http:// Ranjan, R., Chatterjee, P., Chakraborty, S., 2014. Evaluating Performance of Engi-
dx.doi.org/10.1007/978-3-642-53939-8. neering Departments in an Indian University Using DEMATEL and Compromise
Kannan, D., Khodaverdi, R., Olfat, L., Jafarian, A., Diabat, A., 2013. Integrated fuzzy Ranking Methods, Opsearch. http://dx.doi.org/10.1007/s12597-014-0186-1.
multi criteria decision making method and multi-objective programming Redmond, E.C., Griffith, C.J., 2003. Consumer food handling in the home: a review of
approach for supplier selection and order allocation in a green supply chain. food safety studies. J. Food Protection® 66 (1), 130e161.
J. Clean. Prod. 47, 355e367. http://dx.doi.org/10.1016/j.jclepro.2013.02.010. Rezaei, J., Fahim, P.B., Tavasszy, L., 2014. Supplier selection in the airline retail in-
Kannan, D., Jabbour, A.B.L.D.S., Jabbour, C.J.C., 2014. Selecting green suppliers based dustry using a funnel methodology: conjunctive screening method and fuzzy
on GSCM practices: using fuzzy TOPSIS applied to a Brazilian electronics AHP. Expert Syst. Appl. 41 (18), 8165e8179. http://dx.doi.org/10.1016/
company. Eur. J. Operational Res. 233 (2), 432e447. http://dx.doi.org/10.1016/ j.eswa.2014.07.005.
j.ejor.2013.07.023. Rostamzadeh, R., Govindan, K., Esmaeili, A., Sabaghi, M., 2015. Application of fuzzy
Kannan, D., Govindan, K., Rajendran, S., 2015. Fuzzy Axiomatic Design approach VIKOR for evaluation of green supply chain management practices. Ecol. Indic.
based green supplier selection: a case study from Singapore. J. Clean. Prod. 96, 49, 188e203. http://dx.doi.org/10.1016/j.ecolind.2014.09.045.
194e208. Shen, L., Olfat, L., Govindan, K., Khodaverdi, R., Diabat, A., 2013. A fuzzy multi criteria
Ke, H., Cui, Z., Govindan, K., Zavadskas, E.K., 2015. The impact of contractual approach for evaluating green supplier's performance in green supply chain
governance and trust on EPC projects in construction supply chain perfor- with linguistic preferences, Resources. Conservation Recycl. 74, 170e179. http://
mance. Eng. Econ. 26 (4), 349e363. dx.doi.org/10.1016/j.resconrec.2012.09.006.
Khademi-Zare, H., Zarei, M., Sadeghieh, A., Owlia, M.S., 2010. Ranking the strategic Tamura, H., Akazawa, K., 2005. Stochastic DEMATEL for structural modeling of a
actions of Iran mobile cellular telecommunication using two models of fuzzy complex problematique for realizing safe, secure and reliable society.
QFD. Telecommun. Policy 11 (34), 747e759. http://dx.doi.org/10.1016/ J. Telecommun. Inf. Technol. 4, 139e146.
j.telpol.2010.10.001. Tang, J., Zhang, Y.E., Tu, Y., Chen, Y., Dong, Y., 2005. Synthesis, evaluation, and se-
Khaksar, E., Abbasnejad, T., Esmaeili, A., Tamosaitiene, _ J., 2016. The effect of green lection of parts design scheme in supplier involved product development.
supply chain management practices on environmental performance and Concurr. Eng. 13 (4), 277e289. http://dx.doi.org/10.1177/1063293X05059806.
competitive advantage: a case study of the cement industry. Technol. Econ. Dev. Tavana, M., Yazdani, M., Di Caprio, D., 2016. An application of an integrated
Econ. 22 (2), 293e308. ANPeQFD framework for sustainable supplier selection. Int. J. Logist. Res. Appl.
Kumar, A., Jain, V., Kumar, S., 2014. A comprehensive environment friendly 1e22.
approach for supplier selection. Omega 42 (1), 109e123. http://dx.doi.org/ Tuzkaya, G., Ozgen, A., Ozgen, D., Tuzkaya, U., 2009. Environmental performance
10.1016/j.omega.2013.04.003. evaluation of suppliers: a hybrid fuzzy multi-criteria decision approach. Int. J.
Kuo, R.J., Wang, Y.C., Tien, F.C., 2010. Integration of artificial neural network and Environ. Sci. Technol. 6 (3), 477e490. http://dx.doi.org/10.1007/BF03326087.
MADA methods for green supplier selection. J. Clean. Prod. 18 (12), 1161e1170. Tzeng, G.H., Chiang, C.H., Li, C.W., 2007. Evaluating intertwined effects in e-learning
http://dx.doi.org/10.1016/j.jclepro.2010.03.020. programs: a novel hybrid MCDM model based on factor analysis and DEMATEL.
Labib, A.W., 2011. A supplier selection model: a comparison of fuzzy logic and the Expert Syst. Appl. 32 (4), 1028e1044. http://dx.doi.org/10.1016/
analytic hierarchy process. Int. J. Prod. Res. 49 (21), 6287e6299. j.eswa.2006.02.004.
Lee, H.I., Kang, H.Y., Hsu, C.F., Hung, H.C., 2009. A green supplier selection model for Van Kleef, E., Frewer, L.J., Chryssochoidis, G.M., Houghton, J.R., Korzen-Bohr, S.,
high-tech industry. Expert Syst. Appl. 36 (4), 791e798. http://dx.doi.org/ Krystallis, T., Rowe, G., 2006. Perceptions of food risk management among key
10.1016/j.eswa.2008.11.052. stakeholders: results from a cross-European study. Appetite 47 (1), 46e63.
Liao, C.N., Fu, Y.K., Wu, L.C., 2015. Integrated FAHP, ARAS-F and MSGP methods for Wognum, P.N., Bremmers, H., Trienekens, J.H., van der Vorst, J.G., Bloemhof, J.M.,
green supplier evaluation and selection. Technol. Econ. Dev. Econ. 1e19. 2011. Systems for sustainability and transparency of food supply
Liou, J.J., Tamosaitiene,
_ J., Zavadskas, E.K., Tzeng, G.H., 2016. New hybrid COPRAS-G chainseCurrent status and challenges. Adv. Eng. Inf. 25 (1), 65e76.
MADM Model for improving and selecting suppliers in green supply chain Yazdani, M., Hashemkhani Zolfani, S., Zavadskas, E.K., 2016. New integration of
management. Int. J. Prod. Res. 54 (1), 114e134. MCDM methods and QFD in the selection of green suppliers. J. Bus. Econ.
Lu, L.Y.Y., Wu, C.H., Kuo, T.C., 2007. Environmental principles applicable to green Manag. 1e17.
supplier evaluation by using multi-objective decision analysis. Int. J. Prod. Res. Yu, C., Huatuco, L.H., 2016. Supply Chain Risk Management Identification and
45 (18e19), 4317e4331. http://dx.doi.org/10.1080/00207540701472694. Mitigation: a Case Study in a Chinese Dairy Company. InSustainable Design and
Luthra, S., Govindan, K., Kannan, D., Mangla, S.K., Garg, C.P., 2017. An integrated Manufacturing 2016 (Pp. 475e486. Springer International Publishing.
framework for sustainable supplier selection and evaluation in supply chains. Zavadskas, E.K., Kaklauskas, A., Sarka, V., 1994. The new method of multi criteria
J. Clean. Prod. 140 (P3), 1686e1698. http://dx.doi.org/10.1016/j.jclepro.2016.09.078. complex proportional assessment of projects. Technol. Econ. Dev. Econ. 1 (3),
Marsden, T., Banks, J., Bristow, G., 2000. Food supply chain approaches: exploring 131e139.
their role in rural development. Sociol. Rural. 40 (4), 424e438. Zavadskas, E.K., Kaklauskas, A., Peldschus, F., Turskis, Z., 2007. Multi-attribute
Omurca, S.I., 2013. An intelligent supplier evaluation, selection and development assessment of road design solutions by using the COPRAS method. Baltic J. Road
system. Appl. Soft Comput. 13 (1), 690e697. http://dx.doi.org/10.1016/ Bridge Eng. 2 (4), 195e203.
j.asoc.2012.08.008. Zhao, H., Guo, S., 2014. Selecting green supplier of thermal power equipment by
Ordoobadi, S.M., 2009. Development of a supplier selection model using fuzzy logic. using a Hybrid MCDM Method for sustainability. Sustainability 6 (1), 217e235.
Supply Chain Manag. An Int. J. 14 (4), 314e327. http://dx.doi.org/10.3390/su6010217.
Patzelt, H., Shepherd, D.A., 2011. Recognizing opportunities for sustainable devel- Zhu, Q., Sarkis, J., 2006. An inter-sectoral comparison of green supply chain man-
opment. Entrepreneursh. Theory Pract. 35 (4), 631e652. agement in China: drivers and practices. J. Clean. Prod. 14 (5), 472e486.
Rajesh, R., Ravi, V., 2015. Supplier selection in resilient supply chains: a grey