Integrating Social Impact Assessment and Environmental Conflict Analysis on a Hydrocarbon Exploration Project in Spain

Social impact assessment (SIA) has become a key factor for environmental conflicts prevention, which makes necessary to integrate SIA and environmental conflict analysis (ECA). In this article, we integrate SIA and ECA using a method based on grey systems and Shannon entropy. A case study was conducted on a hydrocarbon exploration project located in the Sea of the Gulf of Valencia, Spain. Four stakeholder groups and four evaluation criteria were identified. The results revealed that for group of specialists the project would have negative social impact; and contrary perceptions were found between the group ofprimary activities populationand the group of retirees. It was also noted that the criteria most likely to generate environmental conflicts were the percentage of unemployment and the GDP per capita. These results could help central and community governments to make the best decision on the project. The method showed interesting results and could be apply to manage other projects or programs from point of view of social factors.

integrated. The combination of the GC method and the EW method could integrate SIA and ECA. First, the GC method assesses social impact by quantifying of information from stakeholder groups. And then, the EW method identifies criteria, for which, there is the most divergence between stakeholder groups within of project under scrutiny.
Subsequently, in order to apply and test the integrated method, we conducted a study of SIA and ECA on a hydrocarbon exploration project in the Sea of the Gulf of Valencia, Spain.This hydrocarbon exploration project consists of the application of ultrasound technology, in order to determine the existence of hydrocarbon deposits in the marine subsoil [25]. The company presented environmental impact assessment (EIA) to Spain government in 2012, but at the present (2016) this project is paused due to the fact that a part of the population of Valencia city manifests opposition to the implementation of the project.
The specific objectives of this article are to: 1. Integrate SIA and ECA using the GC method and the EW method. 2. Apply the integrated method to the concrete context of the hydrocarbon exploration project in the Sea of the Gulf of Valencia, Spain.
Section 2 provides details of the methodology to integrate SIA and ECA. In Section 3 the case study is described, followed by the results and discussion in Section 4. Conclusions are provided in Section 5.
II. METHODOLOGY This section describes SIA using the GC method, ECA using the EW method, and provides the details of the integrated method for SIA and ECA.

A. The GC method for SIA
The GC method was developed to classify objects of observation into definable classes, and can be performed by means grey incidence matrices or whitenization weight functions. Whitenization weight functions are mainly used to test whether the objects of observation belong to predetermined classes. In this study, we use centerpoint triangular whitenization weight functions (CTWF), because typically people tend to be more certain about the center-points of grey classes in comparison with other points of the grey class. So, the conclusions based on this cognitive certainty could be more scientific and reliable [5].
The GC method based on CTWF can be described as follows [5], [7], [26]: first, assume that there are a set of m objects, a set of n criteria, and a set of s grey classes; according to the sample value x ij (i=1, 2 ,…, m; j=1, 2, …, n). Then, the steps of the GC method based on CTWF can be developed as follows: Step 1: The ranges of the criteria are divided into s grey classes, and then center-points λ 1 , λ 2 ,…, λ s of grey classes 1, 2, …, s are determined.
where is the CTWF of the kth grey class of the jth criterion, and η j is the weight of criterion j.
Step 4:Ifmax * , we decide that object i belongs to grey class k*. When there are several objects in grey class k*, these objects can be ordered according to the magnitudes of their comprehensive clustering coefficients.

∑ 3
Step 2: The entropy H j of each criterion C j is calculated by Eq. (4).
Step 3: The degree of divergence div j of each criterion C j is calculated by Eq. (5). 1 5 Step 4: The entropy weight w j of each criterion C j is calculated by Eq.(6).

∑ 6
C. Integrating SIA and ECA The integrated method consists of five steps, of which the three first steps correspond to SIA, which are based on the GC method; and the two final steps correspond to ECA, which are based on the EW method, as shown in Fig. 2.

Integration of SIA and ECA
Step 1 Step 2 Step 3 Step 5 Step 4 Criteria  .., m) with respect to the criterion C j (j = 1, 2,..., n) Then, the steps of the Integrated method are described as follows [26]: Step 1:Criteria and grey classes A set of n criteria for SIA, determined by C j (j=1, 2,…, n), is established; and a set of s grey classes, determined by V k (k=1, 2,…, s), is defined.
Step 2:CTWF and Comprehensive clustering coefficient The CTWF values of each stakeholder group are obtained using Eq. (1). Then, the comprehensive clustering coefficients for object i, i=1, 2,…, m, with respect to the grey class k, k=1,…, s, are calculated using Eq. (2).

9
where is the entropy weight for each criterion C j and is the result of SIA for each stakeholder group. The results are represented by the matrix defined by Eq. (10).
III. CASE STUDY SIA and ECA were performed for a project located in the Sea of the Gulf of Valencia in Spain, as shown in Fig. 3. The concerned company proposes to conduct the hydrocarbon exploration by means a campaign of 3D seismic acquisition in zones B, G, AM-1 and AM-2, indicated on the map [25]. Ultrasound technology was proposed to be used to determine the existence of hydrocarbon deposits in the marine subsoil. This study was conducted on the city of Valencia, located into the zone of influence of the project.

A. Stakeholder Groups
During the field work, we identified four different stakeholder groups (k=4), the composition of these groups was determined according to similarities found during the overall assessment on the hydrocarbon exploration project [26]. The sample size in each group was determined by means the principle of saturation of discourse, which establish that information gathering should end when respondents do not produce new information relevant to object of study [30]. The stakeholder groups are presented in Table I:   TABLE I. Stakeholder groups in the case study

G1: Primary activities population
It was composed of those members of the population who are directly linked with the impacts of the project, consisting of people undertaking productive activities related to fishing or tourism (see Fig. 4). This group was made up of thirty interviewees.

G2: University students
It was composed of students with no links to productive activities related to fishing or tourism (see Fig. 5). This group was made up of thirty interviewees.

G3: Retirees
It was composed of retirees (see Fig. 6). This group was made up of fifteen interviewees.

G4: Specialists
It was composed of experts from different fields who are familiar with the area of influence and the characteristics of the environmental and social impacts of hydrocarbon exploration projects (see Fig. 7). This group was made up of eight interviewees.

B. Calculations using the integrated method
The calculations for the case study, based on the integrated method, are preceded as follows.
Step 1: Criteria and grey classes a. Evaluation criteria The criteria for the case study were established by taking into account to the economic and social situation of the city of Valencia and the characteristics of the project, and by consulting with experts. The social criteria are directly linked to the economic criteria, due to the fact that social conflicts in Spain are related to the economic crisis facing the country. Four criteria (n=4) were identified as shown in Fig. 8. The established criteria are described in Table II.

C1
It measured the change in the volume of fishing in the ComunitatValenciana, with the baseline figure being taken as the volume of fishing in 2013, which was 31,29 thousand tonnes of fish [31].

C2
It measured the change in the number of foreign tourists visiting the ComunitatValenciana, with the baseline figure being taken as the number of foreign tourists in 2013, which was 5.97 million [31].

C3
It measured the change in quantity of GDP per capita in the ComunitatValenciana, with the baseline figure being the GDP per capita in 2013, which was 19,500 euros per year [32].

C4
It measured the change in the percentage of unemployment in the ComunitatValenciana, with the baseline figure being the unemployment rate in 2013, which was 28.05% [31].

b. Grey classes
Five grey classes (s = 5) for the case study were established according to the historical information from 2009 to 2013 [31], [32], and by the consultation with experts, in order to satisfy the need to reflect the characteristics of the specific region as accurately as possible [5]. All the criteria had the same weight (η j = 0.250), as they are social criteria [30]. The grey classes established for each criterion are shown in Table III. Step 2: CTWF and the comprehensive clustering coefficient The data obtained from the stakeholder groups were processed using CTWF. The grey classes were extended in two directions by adding the grey classes V 0 and V 6 ("extra negative" and "extra positive", respectively), with their center-points λ 0 and λ 6 . Therefore, the new sequence of center-points was λ 0 , λ 1 , λ 2 , λ 3 , and λ 6 , as shown in Table IV and Fig. 9. The information from stakeholder groups was gathered by means of direct interviews using a structured questionnaire based on the evaluation criteria and grey classes established for the case study. The questions used are presented in Table V Table VI shows the overall results of evaluation from four stakeholder groups (m = 4) for each criterion. These data were aggregated using the arithmetic mean [33].  (15). Subsequently, the comprehensive clustering coefficient ( ) was calculated for each stakeholder group using Eq. (2). The values of CTWF and obtained for group G1 (m=1) are shown in Table VII. Step 3: Percentage system The final stage of SIA for the case study involved the employment of a percentage system defined by the values α 1 , α 2 , α 3 , α 4 , and α 5 ; where α 5 =100, α 1 =100/5=20, α 2 =α 1 +α 1 =40, α 3 =α 1 +α 2 =60, and α 4 =α 1 +α 3 =80; according to five grey classes established, as shown in Table VIII. Then, SIA for group G1 was calculated using Eq. (7). The results are presented in Table IX.

Very negative
The values of SIA for groups G2, G3 and G4 were obtained using the same procedure as for group G1. The results for all stakeholder groups are presented in Table X. Step 4: Entropy-weight method ECA for the case study was carried out by applying the EW method. First, the criteria values shown in Table  X were normalized using Eq. (3). The normalized values are shown in Table XI. Then, H j , div j , andw j were calculated using Eqs. (4)- (6). The results are shown in Table XII. Step 5: Objective assessment ECA for the case study was completed by calculating objective assessment of each stakeholder group i, i=1, 2, 3, 4, for each criterion C j (j=1, 2, 3, 4). The results were obtained using Eq. (9), as shown in Table XIII.

IV. RESULTS AND DISCUSSION
The results and discussion, according to objectives in this study, are presented below.
A. The potential of the integrated method to SIA and ECA First, SIA is a topic with high level of uncertainty; therefore, it should be analysed by methods, which consider the uncertainty. Some classical approaches of multi-criteria analysis, such as Delphi [34], [35] or analytic hierarchy process (AHP) [36], [37], do not consider the uncertainty within their analysis, due to the fact that the importance degrees of criteria and performance scores of alternatives are assumed to be known precisely [38]. In addition, some options to model the uncertainly can be fuzzy logic approaches [39], probabilistic approaches [40] or grey systems approaches [5].
Second, Approaches based on fuzzy logic, such as fuzzy analytic hierarchy process (FAHP) [39], [41], emphasize the investigation of problems with cognitive uncertainty, which research objects possess the characteristic of clear intention and unclear extension. The focus of approaches based on grey systems theory is on the uncertainty problems, which the research objects possess the characteristic of unclear intention and clear extension [5]. SIA has clear extension of the criteria on a study determined; for example, in a historic range of five years, we can know the minimum and maximum value of a social variable under analysis. In addition, affected population of a determined project could be clear about when things were good or bad: before or after project implementation [26].
Third, in statistical approaches the concept of large samples represents the degree of tolerance to incompleteness [5], and considering that one of the criteria for evaluating methods can be the cost [4], in this aspect an approach based in grey systems would have a lower cost with respect to a statistical approach, due to the fact that sample size influences on the cost during the field work. In addition, in 1994, JiangpingQiu and Xisheng Hua established a comparison between statistical regression model and grey model on the deformation and leakage data of a certain large scale hydraulic dam. Their work showed that their grey model could provide a better fit than the statistical regression model [5].
Therefore, it could be argued that the GC method based on grey systems theory would benefit SIA, as it considers the uncertainty within its analysis. In addition, the grey clustering method would be more adequate than approaches based on fuzzy logic, as it considers clear extension for evaluation criteria. Furthermore, the GC method could be more effective and would have a lower cost than other statistical approaches during its application.
In turn, ECA is a social topic, which also has high level of uncertainty. ECA could be conducted by classical multi-criteria methods [4], or by statistical approaches [5]. However, classical multi-criteria methods do not consider the uncertainty within their analysis [38]. In addition, statistical approaches would have high cost during the field work [4]. ECA could be carried out by means the EW method based on Shannon entropy, which is a method that also considers the uncertainty within its analysis [17]. Therefore, the EW method and the GC methodare a good option to integrate SIA and ECA,both under the same philosophy.

B. The case study 1. Analysis of findings from calculations
The calculations for the case study produced three important findings, which are discussed below. First, the major tensions among stakeholder groups were identified. Fig. 10 (based on Table X) shows a strong antagonism between groups G1 (primary activities population) and G3 (retirees), despite the fact that the specialists (G4) expressed the opinion that the project would have a negative social impact. The results indicate that G1 and G3, presented contradictory views on the project, these differences suggest potential conflicts between G1 and G3 groups. In order to analyse and more fully understand the mechanisms and forces at play, we need to look at the specific criteria of conflict between G1 and G3, which points to our second important finding.  Table X shows the behaviour of the criteria for G1 and G3 groups: for group G1, all the criteria are in the "very negative" range; for group G3, C1 and C2 are placed in the range of "normal", C3 is found in the range of "positive", and C4 is in the range of "very positive". These results suggest a specific comparison of all these criteria, in order to identify the most controversial criteria among them. Fig. 11. Values of SIA of each criterion for groups G1 and G3 Third, the most divergent criteria between the stakeholder groups, which could imply potential causes of conflicts, were identified. Fig. 12, which is based on Table XIII, shows that the stakeholder groups converge for criteria C1 (volume of fishing) and C2 (quantity of tourists) and diverge for criteria C3 (GDP per capita) and C4 (percentage of unemployment). The convergent criteria can be considered as strengths and the divergent criteria as threats in a possible environmental conflict. The criterion with the greatest divergence is related to unemployment, followed by GDP per capita. Therefore, these issues should be taken into account when implementing measures to prevent environmental conflicts on the hydrocarbon exploration project.

Analysis of conflictive criteria a. Percentage of unemployment
The group G3 (retirees) believe that the project will generate direct and indirect employment, as the hydrocarbon industry demands supplies that would increase the employment in all economic sectors. However, the group G1 (Primary activities population), in concordance with the groups G2 (university students) and G4 (specialists), strongly believe that the project will destroy the employment in sensitive sectors, such as tourism and fishing. Therefore, this fact generates discomfort on a part of the population in Valencia (see Fig. 13), as unemployment is a social problem in Spain, which increased since year 2009, due to the fact that the economic crisis in Europe and particularly in Spain impacts on the unemployment; for example, in Valencia in 2009 was 20.76%, and in 2013 was 28.05% [31].

b. GDP per capita
The group G3 believe that the project will increase the GDP per capita, as there will be investment from the company that will impulse other sectors of the economy. However, for groups G1, G2 and G4, the project will affect to the more important economic sectors of Valencia, which are tourism and fishing. For example, a part of group G1, the fishing cooperative of Valencia strongly believes that the project will affect their economic income, considering the context of lack of employment (see Fig. 14). This fact could be understudied, asin the ComunitatVelenciana, the GDP per capita has been decreased according to increasing of economic crisis since 2009; for example, in 2009 was 20170 euros per year, and in 2013 was 19500 euros per year [31]. This is due to the fact that the employment and the salary have decreased notably.

V. CONCLUSIONS
The methodology applied in this article made possible to integrate SIA and ECA. SIA was conducted by means the GC method, which quantified the qualitative information collected from stakeholder groups, and ECA was performed by means the EW method, which identified the controversial criteria. The results obtained on the hydrocarbon exploration project in the Sea of the Gulf of Valencia, Spain, could help to central government or authorities of the community to make the best decision to manage the use of the Gulf of Valencia.
The main advantages of the integrated method could be summarized as follows: the integrated method would be more effective than other classical multi-criteria methods, as it considers uncertainty within its analysis; would be more appropriate than other approaches based on fuzzy logic, as it considers clear extension of criteria within its analysis;and would have a lower cost than other statistical approaches during its application.
The main limitations of the integrated method could be summarized as follows: the approaches based on grey systems or Shannon entropy are not widely diffused compared to approaches based on multi-criteria analysis, fuzzy logic or statistics models; the Integrated method presents still subjective aspects, during information gathering and the establishment of limits of grey classes; and the calculations are still tedious during the application of the integrated method, this fact could be improved by implementing a computer system.
Finally, the integrated method could be applied, in future studies on social impact assessment and environmental conflict analyses, to other types of programs or projects. The number of stakeholder groups and criteria could be determinate according to each type of project or program and the concrete social situation of the zone of influence.