Prioritizing sustainable renewable energy systems in Tunisia: An integrated approach using hybrid multi-criteria decision analysis

In recent years, renewable energy technologies (RETs) have become increasingly popular worldwide to achieve energy sufficiency, reduce reliance on conventional fuels, and mitigate their devastating environmental impact. Nonetheless, more appears to have stayed the same in emerging economies, such as Tunisia, as various barriers hampered the implementation of these technologies. Hence, the prime objective of this article is to conduct a thoughtful assessment of four prominent renewable energy options for electricity generation and explore the most potential barriers hindering their development in Tunisia. To do so, a two-stage approach was applied. First, CRiteria Importance Through Inter-Criteria Correlation (CRITIC) and Evaluation based on Distance from Average Solution (EDAS) were used to evaluate the considered alternatives, namely solar PV, concentrated solar power (CSP), onshore wind, and biomass. Then, an integrated step-wise assessment ratio analysis and decision-making trial and evaluation laboratory approach (SWARA-DEMATEL) was developed to identify the most relevant barriers. The related criteria and barriers were determined based on an exhaustive literature review and experts’ feedback. The obtained results indicated that solar PV and wind were the most promising renewable technologies, while biomass was the worst option. To validate the outcome and demonstrate the effect of changing input data on the final results, a sensitivity analysis has been carried out. As for the key obstacles, political instability, high upfront costs, and market access mechanisms were the most influential barriers. Hence, prioritizing different RETs and analyzing the inherent barriers to their utilization would offer insightful guidance to policymakers, enabling them to establish efficient strategies for overcoming these obstacles, and accelerating the deployment of these technologies in the country.


Introduction
Energy has become the mainstay of economic growth, development, and well-being of any nation.In light of this fact, energy demand has increased dramatically over time in many places owing to population and economic growth.The majority of supplied energy worldwide is generated from conventional fuels such as coal, oil, and natural gas (REN 21, 2020).However, given the exhaustible nature of fossil fuels, coupled with their volatile prices and drastic environmental impacts, the penetration of renewable energy sources (RES) has gained significant momentum worldwide (Rehman et al., 2021).Driven by favorable incentives, supportive governmental policies, and mechanisms, solar and wind technologies have taken the lead in terms of the fastest-growing renewable energy technologies (RETs) (REN 21, 2023).In recent years, newly added solar PV and wind capacities were estimated at 191 and 75 GWh/year, respectively, and they are expected to increase to 615 and 335 GWh/year by 2050 (IRENA, 2023;Wen et al., 2020).
Therefore, countries around the world and local governments have been trying to diversify their energy systems in an effort to tackle the challenges of the energy-environment-economy nexus as well as achieve the sustainable development goals (SGDs).Currently, the share of RES reached more than 27% in the global power generation with more than 200 gigawatts added in 2019 (REN 21, 2020).
Nevertheless, developing a feasible RES option has to be considered with the utmost care over a variety of aspects.The sensible selection of the most appropriate renewable technology boosts the economic benefits, local employment, energy security, and mitigates the environmental degradation; in contrast, selecting inadequate technology can lead to various dire consequences most notably the financial burdens (Haddad et al., 2017;Wu et al., 2020;Yazdani et al., 2020).Yet, the assessment of RES is not a straightforward task owing to the multidisciplinary data and multiple conflicting criteria involved in the evaluation process (Al Garni et al., 2016;Lehr et al., 2016;Yazdani et al., 2020).Therefore, determining the optimal energy mix should go beyond the pure evaluation of RES potentials in a given region.Moreover, despite their robust influences on environmental, social, and economic factors, implementing RESs is still faced with a multitude of challenges, especially in developing countries, as in the case of Tunisia (Fashina et al., 2018).
Tunisia is a country that has been actively exploring RES for many years.Owing to its meteorological conditions, the country is well endowed with enormous renewable resource potential, most notably solar and wind (Abdelrazik et al., 2022;Rekik andEl Alimi, 2023a, 2023b).In recent years, Tunisia has made significant strides in developing its renewable energy sector.In an attempt to accelerate the penetration of RETs into the energy mix, the government has approved dozens of wind and solar projects (10-200 MW) (Ministère de l'Energie, des Mines et des Energies Renouvelables de Tunisie, 2020).However, a successful shift toward RES is always associated with high upfront costs, an encouraging investment climate, supportive government policies, social acceptance, and many more.Consistent with this, it is crucial to identify and prioritize the barriers inherent in renewable energy projects, as they often involve new and untested technologies (Qazi et al., 2021).Failure to determine these barriers can result in significant financial losses, delays, or even the cancelation of the project (Hulio et al., 2022;Oryani et al., 2021).In this regard, the multi-criteria decision-making (MCDM) tools have been successfully applied to assess and analyze the various issues associated with the promotion of RES projects.
Given the criticality of assessing and selecting the most appropriate renewable technology, this current article seeks to develop a decision support mechanism using a CRITIC-EDAS approach for prioritizing the renewable energy options for electricity generation in Tunisia, mainly solar photovoltaics, concentrated solar power, onshore wind, and biomass; taking into account technical, economic, environmental, and social dimensions.In addition, a SWARA-DEMATEL model has been applied to identify, and prioritize the most significant barriers to implementing RESs in Tunisia.It is thought that this would provide policymakers with valuable insights, which would allow them to develop proper strategies to overcome these barriers and accelerate the deployment of RETs in the country.

Related literature
The selection of the most feasible and sustainable RET involves several conflicting criteria and multidisciplinary data.In this context, the MCDM methods have emerged as significantly flexible tools in assisting the decision makers to map out the problem by handling and bringing together a wide range of variables.In the literature, a variety of MCDM approaches have been successfully used in the energy-related projects with two main focuses, assigning weights to the considered criteria and ranking the alternatives according to the defined criteria Yazdani et al., (2020).Recently, more robust models have been introduced on the topic.Among those techniques, CRITIC and EDAS have been frequently used by numerous scholars either extended or integrated with other models.Shi et al. (2021) developed a CRITIC decision making tool to identify and assess the power quality problems associated with microgrid systems at the occurrence of large capacity load changes.Also in Nigeria, to address the sustainability of solar PV microgrid in rural communities in Nigeria, a STEEP framework based on CRITIC and PROMETHEE under the fuzzy environment was proposed by Akinyele et al. (2019).In a similar study, Babatunde and Ighravwe (2019) presented a viable analysis to deploy a hybrid renewable energy system within low-income household using the CRITIC-TOPSIS technique.The same model was propounded by Gu and Liu (2022) to address the evaluation of the main power grid resilience under the background of energy transformation considering the impacts of extreme disasters.In Bangladesh, Ali et al. (2020) developed a novel CRITIC-CODAS model to investigate the feasibility of deploying hybrid RES within the coastal regions.In another study, Narayanamoorthy et al. (2021) introduced an extended version of the MCDM approach based on NWHF-CRITIC and NWHF-MAUT to select the optimal wind turbine considering four factors, namely, capacity, voltage, power level, and quality.
As a multiple attribute decision-making (MADM) technique, the EDAS combined with the Shannon Entropy was proposed by Yazdani et al. (2020) to evaluate the potentiality of five RES in Saudi Arabia, namely, solar PV, solar thermal, wind power, biomass, and geothermal.Zhang et al. (2019) developed an integrated model based on EDAS, WASPAS, and TOPSIS to select the most feasible micro-generation alternative in Lithuania.Ramezanzade et al. (2021) used hybrid MCDM methods including EDAS, ARAS, MOORA, and VIKOR under fuzzy environment to prioritize renewable energy projects in Northern Khorasan, Iran.To take the optimal investment decision in the renewable energy sector in Turkey, Karatop et al. (2021) integrated EDAS with AHP and FMEA.Asante et al. (2020) combined EDAS with MULTIMOORA to address the different barriers hindering the development the renewable technologies in Ghana.In another work, Babatunde et al. (2022) and Brand and Missaoui (2014) used an integrated decision-making approach using CRITIC and EDAS methods to analyze the ideal off-grid hybrid renewable energy system in a high-rise institutional building in Nigeria.In order to select the most economically feasible system, the authors stressed that a proper trade-off among the various criteria should be reached as the choice of the best alternative may change if the sustainability was the prime concern rather than the total cost.Similarly, Moitra et al. (2021) proposed a decision supportive system using EDAS and CRITIC to select the optimum battery energy storage system.Based on the literature survey CRITIC was used to determine the relative importance of the considered criteria, whereas, EDAS was used for ranking the alternatives with respect to the given criteria.
In order to successfully increase the deployment of RESs, it is critical to identify and analyze the various barriers associated with them.To this end, MCDM models such as SWARA and DEMATEL have been widely applied to find appropriate solutions to overcome the identified restrictions.
Owing to its simplicity and straightforwardness, SWARA has been increasingly used as a weighting technique in various fields, including renewable energy systems.For instance, a SWARA was applied by Zolfani and Saparauskas (2013) and Vafaeipour et al. (2014) to evaluate the sustainability of solar projects in Iran.Badi et al. (2021) integrated SWARA and DEMATEL to select the well-suited sites for solar farms in western Libya.In Turkey, an optimal marine power plant was identified using a combined SWARA-WASPAS model (Yücenur and Ipekçi, 2020).From another perspective, the DEMATEL technique has emerged as a useful approach in visualizing the interdependencies inherent in the various factors related to RESs' implementation.
As an example, Azizi et al. (2014) proposed a GIS-based DEMATEL approach to investigate the interrelationships among the different factors for selecting the optimal wind sites in Iran.In Turkey, Büyüközkan and Güleryüz (2016) developed an integrated DEMATEL-ANP model to prioritize various power generation scenarios.Qiu et al. (2020) used fuzzy DEMATEL in conjunction with TOPSIS and VIKOR to assess the systematic risks pertinent to wind energy projects in seven countries, including China, Brazil, India, Indonesia, Mexico, Russia, and Turkey.The authors stated that exchange rates, political stability, and social conflicts were the most potential barriers.Recently, several studies have addressed the various obstacles hindering the promotion of RES in emerging economies using the DEMATEL approach (Gedam et al., 2021;Payel et al., 2023;Siraj et al., 2022).According to the reviewed literature, it has been demonstrated that the use of SWARA and DEMATEL techniques were very effective in exploring and determining the most potential obstacles as well as understanding the interdependencies among the various factors associated with RETs' installation.
Despite the significant body of literature dedicated to utilizing RESs in Tunisia as showcased by Attig-Bahar et al. (2021); Balghouthi et al. (2016); Rekik and El Alimi (2023a);Trabelsi et al. (2016), there is still a lack of research focusing on prioritizing and analyzing the various obstacles hindering RES implementation.This holds especially true since earlier commitments have not been followed through Ben Ammar (2022); Ben Rouine and Roche (2022).Addressing this gap requires a systematic and data-driven approach to prioritizing different renewable technologies and analyzing barriers pertinent to their utilization.This article uses hybrid MCDM models, namely CRITIC, EDAS, SWARA, and DEMATEL for these purposes.First, CRITIC-EDAS ranks feasible and sustainable RETs in Tunisia.Then, prominent obstacles hindering RES acceleration are highlighted through the SWARA-DEMATEL approach.This would offer policymakers a clearer vision on how to come up with appropriate strategies to foster the implementation of these RESs.

Data and methods
There is no doubt that the sensible prioritization of the available RETs spurs economic growth, ensures energy security, and reduces drastic environmental impacts.However, making the wrong choice or failing to identify the prominent barriers inherent in these types of projects can lead to serious consequences.Thus, in this article, a two-stage approach is applied.Firstly, a decision-making model based on CRITIC-EDAS was proposed to assess four well-known RETs, namely solar photovoltaic, solar concentrated power (CSP), onshore wind, and biomass.
To achieve this objective, a broad review of academic literature, industry reports, government documents, and international energy agency publications has been conducted.The purpose of this step is to identify the factors that previous studies have considered significant in evaluating RETs and the barriers associated with them.This could include economic, environmental, social, and technical factors, among others.To complement the information gathered from the literature, an expert panel consisting of individuals with practical experience and knowledge in the field of renewable energy is often convened.Yet, in MCDM research applications, it is essential to involve highly skilled professionals to assess the significance of one factor over another (Sindhu et al., 2017).Consequently, a panel of four specialists well-versed in the specific energy landscape in Tunisia was asked to provide their input on the proposed criteria.As such, using both the literature review and insights from the expert panel, it was possible to compile a comprehensive list of potential criteria for assessing RETs as well as their pertinent barriers, considering the specific context of Tunisia's economic, environmental, and social landscape (See Tables 1 and 2).Therefore, documenting this multi-step selection process in detail adds depth to the research by showing how the findings are grounded in a solid empirical and theoretical foundation.This also demonstrates the study's robustness and credibility, as it shows that the chosen criteria are not arbitrary but are selected and validated through a systematic process that combats bias and takes into account a range of perspectives.
To avoid the subjective and biased nature of decision-makers, we used the CRITIC method to assign weights to the considered criteria.As the process of determining weight is the most critical aspect of MCDM problems, the use of CRITIC significantly improves the reliability of findings.Subsequently, the EDAS approach was utilized to evaluate and rank the four renewable energy alternatives in Tunisia according to the identified criteria.
Secondly, according to the reviewed literature, macroeconomic, technological, and technical potential, as well as sociopolitical context, are all important factors in determining the success of RETs' deployment.Therefore, if the barriers pertaining to these factors are not thoroughly investigated, renewable energy projects are deemed to fail.In this article, to determine the various barriers hindering the implementation of RETs in Tunisia, we carried out an extensive literature survey and then referred to experts, taking into account the Tunisian energy context.Based on the experts' views, 13 potential barriers were identified and categorized into four classes, as shown in Table 2.Then, the SWARA technique was used to rank those barriers in terms of the degree to which they hinder RETs development in Tunisia.Subsequently, the DEMATEL approach was applied to disclose the complicated interdependencies among the various factors and link indirect relationships into the cause-and-effect model.From a methodological perspective, the steps used in prioritizing the well-suited RETs and identifying the most significant barriers pertinent to their utilization In Tunisia are depicted in Figure 1.

CRITIC method
As an MCDM approach, the CRITIC method takes into account the intensity of the opposition and the dispute in the structure of the problem and uses the correlation to compute the differences between the criteria to determine their objective weights according to their relative importance (Diakoulaki et al., 1995).This approach has been widely utilized across diverse fields, such as manufacturing, construction applications, medical pharmacy, electrical grid systems, and sustainable optimization of energy and environment (Lamas et al., 2020;Markovic ́et al., 2020).The CRITIC methodology entails several essential steps, which are presented in detail as follows (Ali et al., 2020): Step 1: The decision matrix [x ij ] ∀ i = 1, 2, 3, . . .m; j = 1, 2, 3, . . .n is formed: Step 2: Decision matrix is normalized according to the following equation, considering the type of criteria (beneficial or non-beneficial).
Then, the initial matrix is converted into a matrix with generic elements [r ij ].
Step 3: Each vector has a standard deviation, which quantifies the extent of variation in values relative to the mean value for a certain criterion.Therefore, the standard deviation (σ j ) of each criterion has to be computed.
Step 4: Create a correlation matrix for the evaluation process.Then, compute the linear correlation coefficient between the criteria measure of the conflict created by criterion Step 5: Determine the measure of conflict caused by criterion j regarding the decision situation defined by the rest of the criteria.
Step 6: Calculate the quantity of information concerning each criterion.
Step 7: Finally, calculate the objective weights of each criterion.

EDAS method
The use of the EDAS technique as an MADM tool has steadily gained prominence in academic literature due to its simplified and efficient ranking process with fewer computations compared to other MCDM methods (Torkayesh et al., 2023).This technique relies on variables such as positive distance from average (PDA) and negative distance from average (NDA) to calculate alternative distances based on each criterion's average solution (Babatunde et al., 2022).For a consideration of n alternatives and m criteria, the key steps of the EDAS model could be outlined as follows: Step 1: Construct the decision matrix. [ where x ij stands for the performance value of alternative i under criterion j.
Step 2. Compute the average solution of each criterion using equations ( 7) and ( 8) Step 3: Construct the PDA and NDA matrices for the assessment process (equations ( 9)-( 14)) In case of jth criterion is beneficial In case of jth criterion is non-beneficial where PDA ij and NDA ij represent the positive and negative distance of the ith alternative from the average solution in terms of jth criterion.
Step 4. We calculate the weighted sum of PDA and NDA for all alternatives using equations ( 15) and ( 16).
where w j is the criterion j weight.
Step 5: Normalize the alternatives' SP and SN values using the following equations.
Step 6: Finally, appraisal score (AS) for each alternative is calculated based on: SWARA approach.First introduced by Keršuliene et al. ( 2010), the SWARA method has been successfully applied as a weighting technique in various MADM problems.Unlike the other methods, such as AHP and ANP, SWARA is a straightforward approach as it allows the experts to participate more spontaneously without considering lengthy pairwise comparisons or consistency issues (Badi et al., 2021).In this approach, the considered criteria are evaluated and ranked from the most significant to the least significant according to the experts' understanding.
Then, the final ranks are calculated by taking the average of the individual rankings.The key steps of SWARA are explained as follows: 1.Each expert ranks the n criteria based on their significance 2. Compute the average attribute value (A ̄c) obtained from T experts using the following formula: 3. Calculate the comparative importance S j of the average value (A ̄c) starting from the second-ranked criterion, and determine its significance compared to the remaining criteria C j+1 .4. Determine the coefficient K j according to: 5. Determine the recalculated weight q j q j = 6. Compute the final weight according to: Originating at the Battelle Geneva Institute in 1971, the DEMATEL methodology is frequently employed to address intricate causal issues within complex systems (Braga et al., 2021;Kobryn ́, 2017;Yazdi et al., 2020).Its main aim is to reveal connections between different factors and to identify direct and indirect interdependencies among them (Si et al., 2018).
Visualizing influential network relation maps built on graph theory is a way to achieve this.Mapping out these relationships provides a clearer understanding of the significant factors and their causal implications within complex problem structures (Braga et al., 2021;Chauhan et al., 2018;Si et al., 2018;Yazdi et al., 2020).To apply the DEMATEL approach, the following steps should be used as outlined: Therefore, the visual planning makes it easier to comprehend the significant and causal factors of the complex structure of the problem.The DEMATEL's main steps are illustrated below: 1. Construct the direct comparison relation matrix for k experts and n criteria, which displays the direct link between variable i and variable j, according to a 5-point scale (Table 3).
2. Calculate the aggregated matrix of k experts by means of arithmetic mean.

CRITIC results
In line with what was stated before, weights were assigned to the considered criteria by implementing all the specified steps in accordance with the CRITIC methodology, as illustrated in Tables 4-7.
Findings revealed that resource availability (C9) held the highest ranking with a weight of 14.03% as opposed to other criteria (Table 8).Efficiency (C8) followed closely behind with a score of 11.48%.Investment cost (C1) and job creation (C10), both obtained nearly equal relative weights of 10.64% and 10.38%, respectively.Surprisingly, energy cost (C3) was identified as having relatively low significance with a weight of 7.77%.Despite the fact that financial constraints are often regarded as key hindrance to energy project execution globally, the CRITIC technique has failed to select this factor as the most influential parameter.

EDAS ranking
At this stage, we applied the EDAS approach to thoroughly evaluate power source alternatives with respect to the used factors.The process involved computing all initial decision matrix criteria using equations ( 6)-( 8) (see Table 9).Then, meticulous calculations using equations ( 11)-( 14) were carried out to determine both positive and negative distances from average, as illustrated in Tables 10  and 11.Subsequently, the appraisal score of each alternative was computed using equation (19).By employing this comprehensive decision-making method, it became evident that solar  PV technology emerged as the optimal choice among other alternatives as indicated in Table 12.This technology obtained the highest appraisal score of 0.486 followed by onshore wind and CSP which secured second and third positions respectively; whereas biomass recorded the lowest score indicating it is a less favorable option.This outcome corroborates our previous findings which highlighted the enormous potential of solar PV and onshore wind in Tunisia.Despite the substantial growth and development of the MCDM field, these methods are known for exhibiting rank reversal phenomenon (RRP), particularly in subjective ones such as AHP, TOPSIS, PROMETHEE, ELECTRE, and WASPAS (Aires and Ferreira, 2018; Baykasoğlu and Ercan, 2021;García-Cascales and Lamata, 2012;Wang and Luo, 2009).This phenomenon involves a change in alternative rankings when adding or removing an alternative, contradicting the principle of irrelevant alternatives' independence (Aires and Ferreira, 2018;García-Cascales and Lamata, 2012).The literature offers diverse interpretations of RRP; some view it as a significant limitation that can lead to misconceptions about alternative differences (Al Salem and Awasthi, 2018;Anbaroglu et al., 2014).Others argue that reversals are infrequent and do not necessarily indicate faulty decision-making if they occur under specific circumstances (Millet and Saaty, 2000;Saaty, 1994).However, in this study, the final ranking of alternatives was not subjected to RRP as we employed the CRITIC approach, a recognized objective tool used to assign weights regardless experts' opinion, in the initial step.

Sensitivity analysis
While the obtained results suggest a preference for solar PV, it is important to consider that this preference for solar PV may not be definitive, as these findings are based on specific input data.Therefore, to ensure the accuracy and robustness of the decision-making process, it would be valuable to conduct a sensitivity analysis.This approach can help identify other possible scenarios and provide a more comprehensive understanding of the outcomes, which ensures that the decisionmaking process is resilient to variations in subjective judgments and potentially volatile input data (Haddad et al., 2017;Sindhu et al., 2017).This analysis involves altering the weights assigned to different criteria to evaluate the impact on the final ranking of alternatives.Accordingly, the sensitivity analysis was conducted based on the following steps: • Modify the weights assigned to the criteria, one at a time or in combination.This change was based on five scenarios: technical factors (technical maturity, efficiency, and resource abundance) were prioritized, economic (capital and O&M costs) were favored, environmental (water usage, land requirement, and emissions) with higher weights, social criteria were privileged, and finally equal weights.• Recalculate the rankings of the RETs using the modified weights to see if there are any changes in the order of preference or importance.• Compare the new rankings with the original rankings to identify which RETs are most sensitive to changes in the weighting.This indicates which criteria are most influential in the decisionmaking process.• Assess the stability and robustness of the original results by determining if the rankings are consistently similar despite changes in weights.A high level of robustness means that the original rankings are not highly sensitive to changes in the weighting criteria, suggesting that the initial results are reliable.
Through this analysis, it became evident that across all the considered scenarios, solar PV technology emerged as the most dominant alternative closely followed by onshore wind (Figure 2).Notably though, when investment and O&M costs were given priority, biomass and CSP showed nearly identical rankings.

SWARA-DEMATEL approach
In the first stage, experts with rich experience in the domain of the energy sector in Tunisia were invited to assign weights to the identified indicators based entirely on their own understanding using the SWARA technique (See Table 13).Then, they were requested to apply the DEMATEL approach to capture the interdependencies among those indicators (see Tables A1-A3).
The SWARA analysis showed that political instability (I 6 ), high upfront costs (I 2 ), and limited access to finances (I 1 ) were the most potential indicators, while the change of policy & regulations (I 7 ) and system requirements (I 12 ) were the least significant ones, as illustrated in Table 14.From the DEMATEL approach, it was observed that the main factors of policy and institutional and macro-economic were in the cause group (R i + C i > 0), while technical and social were in the effect group (R i − C i < 0) (Figure 3).
Likewise, after computing the total relation matrix using the DEMATEL approach (Table 15), it was observed that six indicators (I 6 , I 2 , I 5 , I 1 , I 11 , and I 8 ) were in the cause group, while the remaining ones were in the effect group as shown in Table 16.
An influential relation map (IRM) is a visual representation used to display the relationships between various factors or indicators and how they influence each other.Indicator weights are a measure of the relative importance of each factor within the set of all  factors.In an IRM, higher weights indicate that the indicator has more influence or is more critical within the context of the decision-making process.When applied in an IRM, the high-weighted indicators become focal points, alerting policymakers to the most significant areas that could either promote or hinder the deployment of RETs.Therefore, identifying which barriers have the most impact, guides where to focus efforts and resources for overcoming these obstacles.
In terms of ranking, DEMATEL results were somehow different from the SWARA ones.Even though, I 6 and I 2 had the same ranking as SWARA, I 9 (social unrest) and I 4 (high inflation) were the third and fourth most influential indicators, respectively (Table 9).The overall ranking of indicators was computed using expression 33, as illustrated in Figure 4.  To visualize the influential interdependencies among the identified indicators, the influential network relation map (INRM) was constructed using R i and C i vectors along with the threshold value of the total relationship matrix (α), to filter out the most significant relation lines, as depicted in Figure 5.
As can be seen from the diagram, political instability (I 6 ), high upfront costs (I 2 ), and market access mechanism (I 5 ) were the main causative indicators.Without a favorable and supportive political climate, it would be very difficult to attract investments for deploying RETs, which would lead to excessive delays or even the cancelation of such projects (Pathak et al., 2022;Solangi   Furthermore, high upfront costs (I 2 ) were yet another major obstacle to adopting RETs.Emerging economies, as in the case of Tunisia, find it really hard to allocate the necessary funds to finance these ambitious plans.Additionally, the lack of a clear market access mechanism (I 5 ) such as Feed-in-Tariff (FiT) makes investors reluctant to enter the market.Moreover, limited access to finances (I 1 ) was found to be a significant barrier as the country struggles to reach out to international financial institutions to help provide the much needed funds.In the same cause group, technical skills (I 11 ) and lack of institutional coordination (I 8 ) were also perceived as hindering indicators.Without a skilled workforce and effective coordination among the involved institutions, it would be difficult to implement, operate, and maintain large-scale RETs projects.
In terms of the effect group, even though they do not have a direct impact on the structure, they still have significant importance.Social unrest (I 9 ) was determined to be the third major indicator.A clear example of this is the periodic protests against the deployment of RETs in the southern regions, as local communities were restricted from performing their agricultural activities or even dispossessed of their lands without fair compensation (Ben Ammar, 2022).Therefore, these effect indicators have to be addressed simultaneously along with causative ones.
The perusal of Figure 5 typically provides a clear visual aid to understand complex relationships between different factors, with heavy-weighted indicators being more prominent, which is useful for communicating findings to a wider audience, including non-experts, by visualizing the key takeaways of a study.It simplifies how policymakers and stakeholders interpret the data and the relationships among the indicators, helping them to grasp the complex interdependencies without being statisticians or modelers themselves.This should help in strategic planning by showing each indicator's direct and indirect impacts, informing where changes or interventions could be most effective in improving renewable energy implementation.Thus, with this specific visualization of relationships and weights for Tunisia's renewable energy scenario, its significance would lie in pinpointing exactly where efforts and strategic actions should be directed for the optimal deployment of RETs, according to the study's results.

Discussion
Tunisian energy policies have been increasingly focused on the promotion of RETs as part of the country's efforts to transition away from fossil fuels and reduce its dependence on foreign energy sources.However, renewable technologies face various challenges, including technical obstacles, financing issues, and opposition from entrenched interests in the traditional energy sector (Fashina et al., 2018;Rocher and Verdeil, 2019;Schmidt et al., 2017).Thus, with these contrasting visions of the future of Tunisia's energy landscape, efforts to promote RETs are contested.
While data-driven meta-modeling and AI-based techniques can provide powerful insights, particularly in handling large volumes of complex data, they may require additional efforts to incorporate subjective criteria and expert judgments, which are inherent to MCDM approaches (Abisoye et al., 2023;Sankarananth et al., 2023).Comparing rankings directly might reveal disparities due to different underlying foundations, where MCDM focuses on a holistic evaluation incorporating expert judgment and AI/meta-modeling emphasizes data patterns and predictive accuracy (Gupta and Singh, 2021;Ohalete et al., 2023).Each framework offers unique advantages, and the best approach depends on the specific objectives, the availability of data, and the context of the decisionmaking environment.
This article delves into the nuances of renewable energy in Tunisia by examining RETs and identifying the various barriers hindering their implementation.Interestingly, the outcomes were perfectly in alignment with previous works.Firstly, the preference for solar PV and onshore is not quite surprising, as both sources are more accessible than any other forms of renewable energy in the country (Attig-Bahar et al., 2021;Balghouthi et al., 2016;Rekik and El Alimi, 2023a;Trabelsi et al., 2016).
Considering other emerging economies with similar socio-economic profiles, geographical characteristics, and energy needs, it appears that adopting solar PV and onshore is the most prevalent trend (Table 17).Moreover, political instability, financial constraints, societal resistance, and technological limitations are the major barriers, which indicates that Tunisia's challenges are typical of those similar economies (Table 17).
Secondly, drawing on the works presented by Rocher andVerdeil (2013a, 2013b), Rocher and Verdeil, (2019) and Verdeil, (2014), this analysis underscores the role of political and spatial dynamics in the energy transition in Tunisia.These dynamics, as reflected in the identified barriers, point toward a larger need for policy reform and coherent governance in the sector.Furthermore, the study echoes the assertions made by Rocher andVerdeil (2013a, 2013b), Siraj et al. (2022), andPayel et al. (2023).Conflicts over landuse, regulatory barriers, financial constraints, and social acceptance were particularly the main obstacles hampering the acceleration of solar power projects.
Moreover, improving institutional coordination and technical skills is crucial for advancing these RETs (Rocher andVerdeil, 2013a, 2013b).Therefore, a particular focus on these aspects might pave the way for a smoother and more successful implementation of renewable energy projects into existing energy systems.
Overall, to take advantage of the substantial potential of these resources in any given area, wellcrafted policy instruments are essential for speeding up the transition to renewable energy systems.Incentives like feed-in tariffs, tax exemptions, and renewable portfolio standards offer financial support for investing in clean energy projects (Solangi et al., 2021;Xu and Solangi, 2023).Furthermore, simplified permitting procedures and policies for grid integration make it easier to adopt these systems (Tang and Solangi, 2023).Partnerships between government, industry, and research organizations also encourage sharing knowledge, innovation, and building capacity.These policy measures establish a supportive atmosphere for the growth of renewable energy and contribute to driving the shift toward a more sustainable future (Shah and Solangi, 2019).

Conclusion
This article prioritizes renewable energy options and identifies barriers to their utilization in Tunisia using integrated CRITIC-EDAS and SWARA-DEMATEL approaches.Solar PV and onshore wind are found to be the most viable options, while CSP and biomass are less favorable due to poor performance in most criteria, particularly in terms of capital costs.The sensitivity analysis confirms the dominance of solar PV and onshore wind, suggesting the need for concrete measures to accelerate their integration.The study also identifies macroeconomic and socio-political barriers to implementing RETs in Tunisia, emphasizing the importance of addressing these barriers to establish a sustainable and renewable energy future.
It is worth noting that while the study presents significant insights, its scope is limited to select renewable technologies, omitting others like geothermal and offshore wind, which may also offer substantial benefits.Building on the momentum of the present study, future investigations could delve into other decision-making frameworks like fuzzy multi-criteria decision-making, which could offer a more nuanced handling of uncertainties in the assessment process.Engagement with stakeholders should be a priority, incorporating comprehensive feedback from diverse influences, including local communities, industry stakeholders, and policymakers, to ensure contextual relevance and practical applicability of the findings.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Figure 1.The main steps for prioritizing RETs and identifying their pertinent barriers in Tunisia.

Figure 2 .
Figure 2. Sensitivity analysis with respect to influential criteria.

Figure 3 .
Figure 3. Influential relation map of the main factors.

Figure 5 .
Figure 5. Influential relation map of the indicators

Table 4 .
CRITIC initial input data matrix.

Table 8 .
Obtained weights using CRITIC method.

Table 9 .
EDAS initial input data matrix.

Table 10 .
Positive distance from average.

Table 13 .
Experts' ranking and scoring of indicators.

Table 17 .
Comparison with similar emerging economies.

Table A2 .
Matrices of expert comparisons in pairs of indicators.