CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 52, 2016 

A publication of 

 

The Italian Association 
of Chemical Engineering 
Online at www.aidic.it/cet 

Guest Editors: Petar Sabev Varbanov, Peng-Yen Liew, Jun-Yow Yong, Jiří Jaromír Klemeš, Hon Loong Lam 
Copyright © 2016, AIDIC Servizi S.r.l., 

ISBN 978-88-95608-42-6; ISSN 2283-9216 
 

Problematique Approach to Analyse Barriers in Implementing 

Industrial Ecology in Philippine Industrial Parks 

Michael A. B. Promentilla*a, Lindley R. Bacudiob, Michael F. D. Benjamind, Anthony 

S.F. Chiub, Krista D. S. Yuc, Raymond R. Tana, Kathleen B. Avisoa  

aChemical Engineering Department,  De La Salle University Manila, 2401 Taft Avenue, Malate, Manila, Philippines, 0922 
bIndustrial Engineering Department,  De La Salle University Manila, 2401 Taft Avenue, Malate, Manila, Philippines, 0922 
cSchool of Economics, De La Salle University Manila, 2401 Taft Avenue, Malate, Manila, Philippines, 0922 
dResearch Center for the Natural and Applied Sciences, University of Santo Tomas, España Blvd., 1015 Manila, Philippines 

michael.promentilla@dlsu.edu.ph  

Industrial ecology is recognized as an important framework toward a circular economy wherein industrial 

systems minimize their environmental burden by mimicking the material cycles and energy cascades found in 

biological ecosystems. Advocates of such framework in planning of eco-industrial parks suggest that both 

economic and environmental gains can be attained by transforming the industrial production from a linear to a 

closed loop system. However, it is imperative first to understand and analyze systematically the barriers in 

implementing the concept of industrial ecology in industrial parks even at the early planning stage. This work 

thus proposes a problematique approach to understand and analyse such barriers toward a successful 

development of eco-industrial parks. A problematique is a term coined by Warfield referring to concepts and 

tools for a structural model of relationships among members of a set of problems. The problematique is shown 

to be effective in analyzing the structure that underlies problematic situations, thus increasing the potential for 

crafting solution through human intervention. An illustrative case study was presented using a methodological 

framework built from Decision Making Trial and Evaluation Laboratory (DEMATEL), Interpretive Structural 

Modeling (ISM) and Analytic Network Process (ANP). Among the identified barriers in an industrial park 

situated in Philippines, the method reveals the strength and direction of interaction, hierarchical network 

structure, prioritization of components, and the causal loop mapping to aid stakeholders in systems thinking 

and problem solving for such complex issues. 

1. Introduction 

With the growing global resource consumption coupled with increasing population growth, eco-industrial parks 

(EIPs) are being promoted as a promising strategy to achieve sustainability in a circular economy. EIPs use 

the framework of industrial ecology particularly industrial symbiosis (IS) wherein industrial systems minimize 

their environmental burden by mimicking the material cycles and energy cascades found in biological 

ecosystems (Frosch and Gallopoulos, 1989). One of the most popular examples of EIP is that of the 

Kalundborg Park in Denmark (Jacobsen, 2006) which has been found to evolve spontaneously as a result of 

limited resources. Since then, attempts at improving the environmental performance of the industrial system 

has been done within eco-industrial parks such as those found in Australia (Roberts, 2004) and Korea (Park et 

al., 2008) to name a few. However, despite the documented benefits of EIPs, challenges towards its 

implementation still exist and thus it is important to conduct a rigorous evaluation of the problem structure and 

the parameters which affect the implementation of IS in industrial parks. The work of Chiu and Yong (2004) for 

example have emphasized that for Asian Developing Countries (ADC) eco-industrial development should be 

viewed as a strategy for economic development rather than a practical or technical instrument. The barriers to 

implementing IS may also vary depending on local or regional factors as well external or global trends 

(Mannino et al., 2015). It is imperative to have a more rigorous evaluation of how such factors which influence 

the implementation of IS interact with each other in order to provide insights on how strategies can be 

                                

 
 

 

 
   

                                                  
DOI: 10.3303/CET1652136 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Please cite this article as: Promentilla M. A. B., Bacudio L. R., Benjamin M. F. D., Chiu A. S. F., Yu K. D. S., Tan R. R., Aviso K. B., 2016, 
Problematique approach to analyse barriers in implementing industrial ecology in philippine industrial parks, Chemical Engineering 
Transactions, 52, 811-816  DOI:10.3303/CET1652136   

811



developed and where they should be focused on. This paper thus extends the work of Bacudio et al (2016), 

and develops a hybridized method based on problematique (Warfield and Perino, 1999) which integrates 

Decision Making Trial and Evaluation Laboratory (DEMATEL), Interpretive Structural Modeling (ISM) and 

Fuzzy Analytic Network Process (ANP) in one framework.  

 

Figure 1: Methodological framework for the problem analysis and prioritization model in implementing eco-

industrial park 

2. Methodology 

Figure 1 describes the methodological framework used in this study. Firstly, problem structuring is done 

through focus group discussion (FGD), literature review and key informant interview (KII) to identify for 

example, the barriers in designing and implementing eco-industrial parks. In a multi-participant decision 

making environment, nominal group thinking (NGT) and Delphi technique could be used to facilitate toward a 

group consensus on the definition of the problem and its decomposition. After the group agreed to the list of n 

barriers or sub-problems, the structural inter-relation matrix (SIM) typically used in ISM can facilitate the 

elicitation from stakeholders as regard to any inter-relationship among these barriers. This includes the 

identification of whether there is a direct relation or not; whether the direction of influence is one-way or two 

ways; and if it’s one way only, then in what direction.  

DEMATEL approach is then used to populate the direct relation matrix (D) of order n. It is a square matrix 

which shows the direct relation among barriers. This matrix contains the intensity of influence of a barrier in 

the row i to the barrier in the column j. The stakeholders provide the intensity, for example using a 5-point 

rating scale wherein zero means no influence and 4 is for “very high” influence, to populate the direct relation 

matrix. Then, this matrix is transformed to a normalized direct relation matrix (M), for example, by dividing 

each entry in D matrix with its largest row sum. The total relation matrix (T = M (I-M)-1) is then computed from 

the M matrix where I is an identity matrix. Each entry in the total relation matrix (T) contains the intensity that 

accounts for both direct and indirect relationship. The row sum (Pi) of this T matrix is an indicative of the 

barrier’s influence in the system in the said row. On the other hand, the column sum (Qi) of this matrix is an 

indicative of how the barrier is influenced in the system. The sum (Pi + Qi) indicates how prominent the 

connections or interactions of that barrier in the system whereas the difference (Pi - Qi) indicates how the 

barrier can be classified either as a causal (positive value) or effect factor (negative value). The cause-effect 

diagram is just the plot of (Pi - Qi) vs (Pi + Qi). Those barriers above the horizontal axis are causal factors 

whereas those below are effect factors. The DEMATEL-based approach thus provides the classification of 

barriers according to the net intensity of their influence on the other barriers in the system.      

Results from DEMATEL can then be used to complement the results from ISM in elucidating the hierarchical 

network structure of the system. The reachability matrix (R) in ISM is a binary matrix representation of the 

digraph that describes structurally how the barriers are connected to each other. An entry of “1” in this matrix 

indicates that the barrier in the row i can reach and influence the barrier in the column j; otherwise, it is “0”. 

From the total relation matrix (T), a threshold value (τ) can be set to define the significant relationship and 

812



simplify the depiction of the digraph. Accordingly, an entry in the reachability matrix is “1” if the corresponding 

entry in the total relation matrix is greater than the said threshold; otherwise, the entry is “0”. 

Using the reachability matrix, the driving power-dependence diagram can be plotted. The row sum of the 

reachability matrix (Yi) is indicative of the relative driving power of a barrier to influence other barriers whereas 

the column sum (Zi) is indicative of the dependency of that barrier to other barriers in the system. MICMAC 

(cross-impact matrix multiplication applied to classification) analysis is then used to classify the barriers into 

four quadrants. These are the so-called “autonomous”, “dependent”, “linked”, and “driver” cluster. The 

“autonomous” are barriers with weak driving power and weak dependence. They are weakly connected from 

the system and may have few links that are strong. The “dependent” are the barriers with weak driving power 

but have strong dependence on the other barriers of the system. On the other hand, the “linked” are barriers 

with relatively strong driving power as well as strong dependence. The “driver” are those barriers with very 

strong driving power but have weak dependence on the system, i.e., few or no barrier in the system can 

influence them. 

In addition, level partitioning of barriers can also be done from the reachability matrix by identifying the 

reachability sets and antecedent sets. The reachability set consists of the barrier itself and the other barriers 

that it may influence whereas the antecedent set consists of the barrier itself and the other barriers that may 

influence it. The intersection of these sets is derived for all barriers and the hierarchical levels where these 

barriers belong are determined. These levels thus aid in building the digraph of a multi-level hierarchical 

network structure of the fuzzy Analytic Network Process (FANP) model.  

The proposed integration of DEMATEL-ISM could aid stakeholders not only to describe quantitatively the 

intensity of influence among barriers but also to further illuminate the causal interrelationship in the said 

structural model. However, this structural model may only account for the strength of relationship among these 

barriers but it does not clearly measure the priority of the barrier itself. This hybrid method, which is built from 

fuzzy ANP (Promentilla et al, 2014), provides a systems approach to capture both the inherent strength or 

importance of the barrier itself, and the intensity of influence among these barriers. The supermatrix (S) is 

analogous to a Markov matrix, i.e., a partitioned matrix wherein each submatrix (Wij) expresses a relationship 

between two a priori defined clusters in a system. These submatrices contain local priority vectors (wk). For 

example, when an element in a row has no direct dependence from the element in the column (i.e., no arrow 

that connects the node to the other node in the digraph), the priority of an element is assigned zero. 

Otherwise, the priorities are the normalized ratio-scale weights associated with the dominance of one element 

over the other element within the cluster, or another element in another cluster of the system. Such 

dominance or strength can be interpreted in terms of importance, preference, likelihood of one element over 

the other element in the cluster or subsystem. As for the submatrix in the supermatrix indicating the 

interdependence among barriers, the column-normalized T is used to measure the influence weights among 

barriers. The rest of the submatrices which expresses other inner or outer dependence between clusters or 

levels in the fuzzy ANP model are populated with priorities derived from pairwise comparison matrix (A). Such 

ratio of weights are indicative of the intensity of dominance of element i over element j which is typically 

elicited from stakeholders in order to compute these priorities (wk). In this study, these local priorities are 

derived using the method proposed in Promentilla et al. (2016) wherein the intensity of dominance is 

represented by a calibrated fuzzy scale, and solution ratios (aij=wi/wj) are approximated using a nonlinear-type 

fuzzy preference programming (Promentilla et al., 2015). After the supermatrix of a strongly connected 

hierarchical network is populated with local priorities, the eigenvector of that supermatrix is computed which 

gives the global priority values of an element in the system (Promentilla et al., 2008). This eigenvector also 

provides the limiting priority weights of the barriers as such eigenvector can also be normalized just within the 

cluster. Note that this is analogous to the synthesizing concept of the limit supermatrix (Saaty, 2001) resulting 

from raising the supermatrix into a large power; in doing so, the transmission of influence along all possible 

paths defined in the hierarchical network structure is captured in the process.   

The proposed prioritization model has an advantage over the typical ANP in terms of a facilitated problem 

structuring through the combined DEMATEL-ISM and a lesser number of pairwise comparison questions that 

need to be elicited from stakeholders to derive the priority weights. To demonstrate the method and further 

understand the step-by-step procedure, an illustrative case study is presented in the next section.  

3. An Illustrative Case Study  

This case study considers an initiative of a certain industrial park situated in Luzon, Philippines to showcase 

an industrial symbiosis network among their locators. Table 1 summarizes the ten potential barriers that were 

identified to implement an eco-industrial park as described in detail in Bacudio et al., (2016). Figure 2 

describes the sample output of DEMATEL-ISM based from the input of one of the stakeholders during the 

focus group discussion. The SIM (see Figure 2(a)) provides the pairwise elicitation of the relationship between 

813



the barriers. The stakeholder then provided the rating, i.e., intensity of influence of the pertinent barrier to the 

other barriers in each row of the direct relation matrix (see Figure 2(b)). Succeeding calculations were done 

(see Figures 2(c) - (d)) to obtain the total relation matrix. From this T matrix, the causal-effect diagram can be 

plotted as shown in Figure 4(a). Results indicate causal barriers such as that of B3 (lack of top management 

support), B4 (lack of training for implementing industrial symbiosis, and B5 (lack of policy to incentivize 

initiative of industrial symbiosis). In addition, the effect barriers are identified such as that of B1 (lack of trust 

among locators) and B8 (lack of institutional support for integration, coordination and communication). 

Addressing the problem of these causal barriers such as the lack of top management support (B3) could affect 

for example in resolving the effect barriers such as the problem of lack of institutional support (B8). 

Table 1: Identified barriers in the planning, design and implementation of eco-industrial park  

Code Definition 

B1 Lack of trust among locators (i.e., industrial plant) 

B2 Lack of information sharing among locators 

B3 Lack of top management support 

B4 Lack of training for implementing industrial symbiosis 

B5 Lack of policy to incentivize initiative of industrial symbiosis 

B6 Lack of funding to promote industrial symbiosis 

B7 Lack of technology and infrastructure readiness 

B8 Lack of an institutional support for integration, coordination and communication 

B9 Lack of willingness to collaborate 

B10 Lack of awareness of industrial symbiosis concepts 

 

 

Figure 2: Sample matrix output from the integrated DEMATEL-ISM for problem analysis 

 

Figure 3: Visual output from the integrated DEMATEL-ISM for problem analysis  

814



To further understand the problem structure, a digraph was formed using ISM to visualize the significant 

relationship among these barriers and structured the causal relations as a multi-level hierarchical network. A 

threshold based on the median of the computed influence intensity in the total relation matrix was used to 

define the binary relation in the reachability matrix. Note that the influence of the barrier to itself is also 

considered in the R matrix and the level partitions for each barrier were derived as shown in Figures 2(d) and 

3(b). In addition, Figure 3(c) describes how the barriers were classified based on the MICMAC analysis. 

 

Figure 3: An example of Fuzzy ANP model including its supermatrix representation and eigenvector  

Table 2: Summary of results obtained from the problem analysis and prioritization model  

Barriers DEMATEL ISM-MICMAC Fuzzy ANPa 

Priority (Rank) 

Fuzzy ANPb 

Priority (Rank) 

B1 Effect Barrier autonomous 0.048 (10) 0.042 (10) 

B2 Effect Barrier linked-dependent 0.098   (6) 0.097   (7) 

B3 Causal Barrier linked-driver 0.135   (1) 0.134   (1) 

B4 Causal Barrier driver-linked 0.107   (5) 0.101   (6) 

B5 Causal Barrier linked-driver 0.124   (2) 0.118   (3) 

B6 Causal Barrier autonomous 0.081   (8) 0.075   (9) 

B7 Causal Barrier autonomous 0.076   (9) 0.086   (8) 

B8 Effect Barrier dependent 0.090   (7) 0.106   (5) 

B9 Effect Barrier linked 0.121   (3) 0.116   (4) 

B10 Effect Barrier linked 0.119   (4) 0.126   (2) 
aModel 1 assumes equally important barriers regardless of whether it is an internal or external issue  
bModel 2 considers priority weights of barriers with respect to internal or external issue with feedback dependence    

 

Indication suggests that the following barriers in the boundary of driver-linked quadrant namely B3, B4 and B5 

are the key driving barriers in the system. These barriers are perceived to be a relatively strong driving barrier 

with high degree of connectedness with the other barriers (see Figures 3(b) and 3(c)). On the other hand, B1 

and B7 are classified as autonomous which indicate their weak relationship with the other barriers. Although 

the lack of technology and infrastructure readiness (B7) is considered as a causal barrier, it is also classified 

as autonomous and thus may not be a prominent barrier to consider. However, prioritization of these barriers 

based on their interrelationship and intensity of influence to each other may provide an incomplete picture of 

the whole problem if the inherent importance or strength of these barriers is not considered. In such case, the 

proposed Fuzzy ANP model addresses this issue in a more systematic way (see Figure 3).  

This study models the decision problem of prioritizing the barriers in terms of a hierarchical network structure. 

For example, the priority weights of these barriers are influenced by how important they are with respect to 

internal and external issue of implementing an eco-industrial park. Here the differentiation of internal and 

external is defined by the systems boundary of an eco-industrial park which includes the locators and people 

working in the industrial park. As shown in Figure 3(a), the goal, which is to prioritize the barriers that need to 

be addressed, would influence (W21 in the supermatrix) on which internal or external issues should be given 

emphasis as depicted by a downward arrow from the 1st level (L1) to the second level (L2). The downward 

arrow from L2 to the third level (L3) represents the outer dependence (W 32 in the supermatrix) associated with 

the relative importance of each barrier respect to either an internal or external issue. For example, B8 is 

perceived to be the most important internal barrier whereas B7 is the most important external barrier. On the 

815



other hand, the upward arrow from L3 to L2 indicates feedback dependence (W23 in the supermatrix) between 

these two levels or clusters. The arc (loop in the digraph) represents the inner dependence among barriers 

within each level. For example, an identity matrix such as that of W 22 indicates that the internal and external 

issues are mutually independent to each other. On the other hand, the interdependence among the barriers is 

expressed in the submatrix W33 which contain inputs derived from the total relation matrix. Note that the 

supermatrix (Figure 3(d)) is populated with priority weights normalized by the maximum priority in the column 

within the cluster. Summary of the results from such fuzzy ANP model is shown in Table 2 in comparison with 

the output from DEMATEL and ISM. The results from this illustrative case study demonstrate how the 

separate techniques complement with each other to understand the factors which are relevant to the problem 

or issue. 

4. Conclusions 

This study proposes a methodological framework that combines DEMATEL, ISM and Fuzzy ANP to analyse 

the barriers of designing and implementing an eco-industrial park in the Philippines. This novel approach not 

only reveals the strength and direction of interaction in a multi-level hierarchical network structure, but also 

provides a systematic way of prioritizing the barriers in the system. Visual output through the causal loop 

mapping and the driving power-dependence diagram could aid stakeholders to understand their mental model 

as regard to the complex inter-relationship of these components underlying the problem. In this illustrative 

case study, indication suggests that the lack of top management (B3), lack of awareness of industrial 

symbiosis concepts (B10), and lack of policy to incentivize initiative of industrial symbiosis (B5) are the key 

barriers that need to be prioritized and addressed to resolve the problem that can be potentially encountered 

in implementing eco-industrial parks. In principle, the proposed problematique approach can also be used to 

other problem domains wherein complex issues require systems thinking and problem analysis. Future studies 

will also incorporate techniques to address uncertainties involved in problem analysis and in the prioritization 

model.  

References  

Bacudio L.R., Benjamin M.F.D., Eusebio R.C.P., Holaysan S.A.K., Promentilla M.A.B., Yu K.D.S., Aviso K.B., 

2016, Analyzing barriers to implementing industrial symbiosis networks using DEMATEL, Sustainable 

Production and Consumption, 7, 57-65. 

Chiu A.S., Yong G., 2004, On the industrial ecology potential in Asian developing countries, Journal of 

Cleaner Production, 12(8), 1037-1045. 

Frosch R.A., Gallopoulos N.E., 1989, Strategies for manufacturing. Scientific American, 261, 144-152. 

Jacobsen N.B., 2006, Industrial symbiosis in Kalundborg, Denmark: a quantitative assessment of economic 

and environmental aspects, Journal of industrial ecology, 10, 239-255. 

Mannino I., Ninka E., Turvani M., Chertow M., 2015, The decline of eco-industrial development in Porto 

Marghera, Italy, Journal of Cleaner Production, 100, 286-296. 

Park H.S., Rene E.R., Choi S.M., Chiu A.S., 2008, Strategies for sustainable development of industrial park in 

Ulsan, South Korea—from spontaneous evolution to systematic expansion of industrial symbiosis, Journal 

of Environmental Management, 87, 1-13. 

Roberts B.H., 2004, The application of industrial ecology principles and planning guidelines for the 

development of eco-industrial parks: an Australian case study, Journal of Cleaner Production, 12, 997-

1010. 

Saaty T.L., 2001, Decision making with Dependence and Feedback: The Analytic Network Process, 2nd ed., 

RWS Publications, Pittsburgh. 

Promentilla M.A.B., Antonio M.R.S.E., Chuaunsu R.M.M., De Serra A.J., 2016, A Calibrated Fuzzy AHP 

Approach to Derive Priorities in a Decision Model for Low Carbon Technologies, In: Proceedings of De La 

Salle University Research Congress <www.dlsu.edu.ph/conferences/dlsu-research-congressproceedings/ 

2016/> accessed 04.04.2016 

Promentilla M.A.B., Aviso K.B., Tan R.R., 2015, A Fuzzy Analytic Hierarchy Process (FAHP) Approach for 

Optimal Selection of Low-carbon Energy Technologies, Chemical Engineering Transactions, 45, 1141-

1146. 

Promentilla M.A.B., Aviso K.B., Tan R.R., 2014, A group fuzzy analytic network process to prioritize low 

carbon energy systems in the Philippines, Energy Procedia 61, 808-811. 

Promentilla M.A.B., Furuichi T., Ishii K., Tanikawa N., 2008, A Fuzzy Analytic Network Process approach for 

evaluation of remedial countermeasures, Journal of Environmental Management, 88, 479-95. 

Warfield J.N., Perino G.H., 1999, The problematique: Evolution of an idea, Systems Research and Behavioral 

Science, 16(3), 221-226. 

816