Putting into practice a decision-making framework for a thorough performance and location evaluation of solar photovoltaic plants in India from distinctive climate zones

solar photovoltaic plant (from the composite climatic zone) outperforms other selected photovoltaic plants and is determined to be the most appropriate plant. This is due to the lowest levelized cost of energy and lowest total cost for the Unchahar solar photovoltaic plant, as these criteria are the most significant per the study. This study ’ s findings will be useful to energy decision-makers and may serve as a reference for constructing and developing future solar plants. Further, this study will make recommendations for future solar photovoltaic plant development. The study also includes sensitivity analysis to determine the data ’ s robustness. This study is very robust, as the normalized alternative is constant for each criterion, and the ranking remains the same.


Introduction
Energy is essential to our society to ensure our quality of life and underpin all other economic elements.However, not all energy sources are clean, minimize environmental impact, produce minimum secondary waste, and are sustainable.Electricity generation from natural resources such as solar, wave, and wind (which will result in less air pollution) has the potential to replace the energy from coal-burning power plants and additionally can reduce health hazards and affect climate [1].Government, investors, and private sectors are continuously thinking over the horizon, for sustainable development and low carbon future with a purpose to ensure we leave future generations a world that affords them the same opportunities as we have enjoyed.Hence, to boost clean energy and reduce fossil fuel use, the world is heading toward renewable energy sources.For this, India is aiming to achieve net zero emissions by 2070 and limit the global temperature rise to 1.5 • C, as announced at COP27, Glasgow [2].Given the global carbon footprint, most countries are shifting to carbon-neutral technologies such as solar energy.In terms of global installations, solar energy has grown much faster than any other renewable source [3].
India, with a population of 1.4 billion people and one of the world's fastest-growing economies, has also demonstrated outstanding achievements in the energy sector in recent years.Since 2000, the country has made successful efforts to provide electricity to more than 700 million people [4].To ensure full access to electricity and further development, India's government took additional steps and implemented numerous reforms to create a sustainable, affordable, clean, and secure energy system [5].Among renewables, solar energy significantly impacts the Indian energy scenario.Solar energy accounts for 28% of total renewable energy generation and 43% of total installed renewable energy capacity in India [6].With such massive solar power capacity, which has increased from 2.6 GW in March 2014 to 64.38 GW in February 2023, the country recently surpassed Italy in solar power deployment to take fifth place globally [7,8].The state-wise solar potential of India in GWp is shown in Fig. 1 [7].Also, the country has managed to grab the tag of the lowest-cost solar power producer globally.Hence, solar energy is an opportunity we must seize for low-carbon development, energy security, and poverty alleviation.
Solar energy is harnessed using photovoltaic cells, solar heating, solar thermal energy, molten salt power plants, artificial photosynthesis, solar architecture, and other ever-evolving technologies.Solar photovoltaics is a technology that converts photon energy to electric energy.Fig. 2 depicts an overview of a solar photovoltaic power plant that directly converts sunlight energy into electricity using photovoltaic technology.
When it comes to solar energy, India has a very favorable location (8 • N and 38 • N, Latitude) and is one of the best recipients in the global solar belt (40 • S to 40 • N Latitude) [4].Every year, the country receives approximately 3000 h of sunshine, with nearly 300 days of good sunshine.The majority of the country receives 4-7 kWh of incident energy per square meter per day [4] Also, the cost of energy produced by solar photovoltaics (PV) power plants has decreased significantly in recent years due to a variety of factors such as improved performance efficiency of photovoltaic modules, reduced photovoltaic module prices, higher module reliability, and lower initial investment [9].In 2018, India became the world's lowest-cost producer of electricity using solar power, reaching nearly seven cents (USD 0.068) per kWh from utility-scale solar photovoltaic [10,11].Also, as part of the post-COVID-19 self-reliance strategy, the Indian government announced solar PV manufacturing as one of the strategic sectors [12].Regardless of the numerous benefits of solar energy discussed above, 176 projects with a total capacity of 11,462 MW across five companies have been examined in Moody's Investor Service report (March 2021) [13].As per this report, India's solar power plants are underperforming, with 15-20% of solar projects failing to meet generation targets in fiscal years 2018-19 and 2019-20.
One of the reasons for solar PV power plant underperformance could be the higher-than-expected degradation rate of solar modules.The quality of the materials used in manufacturing, the manufacturing process, the quality of cell assembly and packaging into the module, and the site maintenance levels are all factors in this type of degradation [14].Additional degradations can occur as a result of a variety of factors, such as the impact of the environment on the module's surface (for example, pollution, dirt, etc.), discoloration or haze of the encapsulant or glass, mechanical stress, and humidity on the contacts, cell contact breakdown, and so on.Dirt accumulation can significantly impact system performance [14].As a result, regular module cleaning is critical for improved output.In contrast, low-quality water and improper cleaning can damage the module and other components, lowering system performance.Microfracture, also known as micro-cracks, can occur in modules and cannot be repaired, and replacement of modules is the only solution.Micro-cracks form due to the expansion and contraction of solar cells caused by thermal cycling.Furthermore, minor flaws in silicon cells can lead to larger microcracks.Other causes of microcracks include humidity and freezing, cyclic pressure loads and wind loading, heavy snowfall, hail, incorrect packaging, inefficient transportation methods, improper handling, and so on.Micro-cracks in a solar photovoltaic system can impact both energy output and system lifetime [15].The inverter Load Ratio (LDR), which describes the inverter's conversion of DC to AC and transmission to the grid, could also be an important factor.Module prices have fallen in recent years compared to inverter prices, prompting some developers to load more modules onto each inverter.This means that inverters are frequently operating at or near full capacity, increasing the overall capacity factor of the array (in AC terms).It can, however, cause power 'clipping,' which occurs when an inverter prevents modules from generating at full output in order to avoid overloading [16].

List of abbreviation
Other influential factors, aside from module degradation, could be non-optimal plant design, poor quality of workmanship practices, module handling and other system balance items, and selection of the wrong technology and/or wrong site.Site parameters influence PV system performance and reliability.The entire Indian subcontinent has very different climate zones with different seasonal patterns, each of which impacts the spectrum differently.This would have a variety of effects on module performance depending on the absorber material properties such as band gap, absorption coefficient, thermal expansion coefficient, and device structure [16].Solar radiation, near and far shadow, ambient temperature, airflow and ventilation, basic wind speed, orientation, dust level and pollution, humidity, extreme weather condition, cloud and haze, heat interaction, and other variables are likely to differ from one site to the next, even within the same geographical area [17].Another factor to consider when selecting a site is that transmission loss increases with increasing distance between the power plant and the utilization location [18].
The dependability, lifetime performance, and location of solar PV plants are directly related to a number of variables such as average global solar radiation, average maximum temperature, overall efficiency, various associated costs, connectivity to a nearby city/town, and so on.Also, PV systems are subject to a variety of losses due to environmental and technical factors, system design limitations, manufacturing defects, and so on which degrade PV system performance.These factors must be examined and compared because they significantly impact the system's efficiency, performance, and risk of investing in this sector.However, this kind of decision-making is complex and includes various critical criteria to solve.Multi-criteria decision-making (MCDM) methods can be employed to overcome these obstacles to choose amongst available options based on these critical criteria.MCDM is an analytical method for weighing the advantages and disadvantages of various alternatives based on a variety of criteria [19].MCDM has been used in a variety of applications, including risk assessment in hydroelectric projects [20], fuel selection [21], food supply chain [22], selection of biomass resources [23], and optimal location selection for offshore wind-PV-seawater pumped storage power plant [24].However, conventional MCDM methods fail when decision-making has an uncertain data [25].To cope with uncertainties, fuzzy MCDM methods can be used to produce more realistic results than conventional MCDM methods when there is a high degree of vagueness and ambiguity [25].Imprecision in decision-making is caused by a number of factors, including unquantifiable information, incomplete information, unobtainable information, and partial ignorance, which can be reduced using fuzzy sets.When accurate data accumulation is impossible due to a lack of information or data uncertainties, fuzzification aids in obtaining concrete results.
Hence, this manuscript comprehensively analyses the performance of seven solar PV plants installed by NTPC Limited (formerly known as National Thermal Power Corporation Limited), an Indian public-sector company involved in power generation and related activities, based on their performance and location criteria.Fuzzy Stepwise Weight Assessment Ratio Analysis (F-SWARA) [26], Fuzzy COmplex PRoportional ASsessment (F-COPRAS) [27], and Fuzzy ELimination Et Choix Traduisant la REalité (ELimination Et Choice Translating REality) (F-ELECTRE) [28] methods are employed for this purpose.F-SWARA is used for criterion weightage, while F-COPRAS and F-ELECTRE are used for assessment value and ranking.These methods were chosen because they have numerous advantages over other methods.For example, the SWARA method has an advantage over other criteria weightage methods as it allows decision-makers to choose the significance of the evaluation criteria based on their judgment.Furthermore, it is an efficient, novel, and straightforward method for estimating the accuracy of decision-makers' weighted criteria.Furthermore, the SWARA method uses only 'n-1' comparisons for 'n' criteria, which is significantly less than methods such as AHP [29].The COPRAS assessment method has the advantage of simultaneously considering both the ideal and ideal-worst solutions.It is a simple compensatory method.The method evaluates a multi-attribute variable system for minimizing and maximizing values.It enables easy comparison and verification of the final results [30].In comparison to other methods, the ELECTRE method, despite its relative complexity, is less sensitive to changes in data, making it more stable and reliable.It analyses both qualitative and quantitative criteria [31].
The seven power plants or the alternatives selected are Dadri solar PV, Port Blair solar PV, Ramagundam solar PV, Talcher Kaniha solar PV, Faridabad solar PV, Unchahar solar PV, and Singrauli solar PV.These power plants were chosen because, first and foremost, they are all built by the same company, and secondly, they cover most of India's climatic conditions.Dadri, Faridabad, and Unchahar are in the composite zone, Port Blair is in the tropical zone, Ramagundam and Singrauli are in the hot and dry zone, and Talcher Kaniha is in the warm and humid zone.As the performance of solar PV is heavily dependent on climatic conditions, selecting power plants from different climatic zones will be extremely beneficial in studying the sites.Following that, these power plants will be compared using various criteria.Economic, technical, environmental, and connectivity criteria have been chosen.These criteria were chosen because they cover the pillars of sustainability, with the exception of social criteria.Since solar PV plants generally have a minimum discernible impact on society compared to other renewable alternatives, such as emitting minimum emission to the environment during its operational phase, compared to hydropower, the land degradation or habitat loss for wildlife is less.Hence, social criteria for this study can be disregarded.The economic criteria are required for reliable energy to be affordable to everyone and for energy security so that each country has constant access to sufficient energy.The environmental criteria are required for the energy source to emit the fewest greenhouse gases and have the least impact on the environment and people.
A review of the literature on MCDM and other decision-making tools and techniques used in renewable and solar energy fields was conducted and presented in this study.The significance of present work over existing work is discussed further.Sindhu et al., in their work, aim to select an appropriate site for solar farms; however, the region is limited to a state of India only, i.e., Haryana.But in the present case study, power plants across India were considered, and that too from different climatic zones to get more realistic results for the solar plant deployment [32].In another study, Saraswat et al. developed and applied a GIS-MCDM model to evaluate suitable sites for developing two prominent onshore renewable energy sources, i.e., solar and wind energy.The study considers various sustainability criteria giving the highest significance to the technical criteria.Although, technical, social-environment, and economic criteria were considered their sub-criteria were more likely connectivity criteria mainly evaluated in distances.In the presented work, the criteria selected are more extensive, and connectivity criteria have been considered separately along with economic, technical, and environmental criteria [33].A study conducted by Ishfaq et al., investigated Pakistan's various renewable energy sources for investment purposes, taking economic, technological, and environmental factors into account.However, the presented paper focuses on solar energy only considering the rapid growth of this source and discusses the various underperforming criteria that can be taken as suggestions in building future plants [34].Another study by Solangi et al. also optimized the solar site selection utilizing the AHP fuzzy-VIKOR method.The region of study is Pakistan.For the study, 14 promising cities of Pakistan were optimized based on economic, environmental, social, location, climate, and orography criteria likewise the presented study and further supplemented with 20 sub-criteria [35].Kannan et al. also investigated the various solar sites east of Iran using a real-world case study.They used the Best Worst method (BMW), Grey relational analysis (GRA), and the VIKOR method to evaluate sustainable locations.Aside from MCDM methods, Monte Carlo simulation was used to evaluate the robustness of the methods, and it was discovered that the VIKOR method has higher sensitivity than the GRA method.Compared with the published work, the present case study uses the newest and novel MCDM methods along with fuzzy to generate more crisp and concrete results [36].In another study, Sánchez-Lozano et al. used a GIS tool with MCDM to select sites for upcoming solar power plants, whereas the presented case study checks on underperforming criteria and also suggests sites for upcoming plants using MCDM methods based on performance comparison analysis.Many of the criteria used in the former are also used in this study, such as location criteria (distance to roads and distances to villages) and climatic criteria (solar irradiation potential and average temperature) [37].Another similar study was done by Rekik et al., based on a similar GIS-based MCDM approach for solar and wind farm site selection through spatial analysis [38].As far as the SWARA, COPRAS, and ELECTRE methods are concerned, several studies have been done on these methods, especially in the field of solar, such as Pythagorean F-SWARA and F-VIKOR methods were utilized before for the performance evaluation of solar panel selection by Rani et al., [39].Lozano et al. used the TOPSIS-ELECTRE method for the optimal site of the photovoltaic solar farm in Spain [40].Similarly, Seker et al. used an integrated hybrid method based on the Intuitionistic Fuzzy AHP and COPRAS approach for solar power plant selection in Turkey [41].
The literature review shows that MCDM methods are primarily used in the energy sector to rank optimal renewable energy, energy policy, and energy investment planning.However, very few papers deal with the study incorporating the faults for solar plant underperformance.Based on underperforming criteria, compare the solar plants and suggest a development measure.Also, when using a mathematical tool like MCDM, the most commonly used methods are AHP, TOPSIS, and ANP, despite their many limitations compared to other methods.AHP is based on measurements of probability and potential measures and has a subjective nature in the modelling process that constitutes an AHP limit and therefore does not guarantee a concise decision-making outcome [42].If the hierarchical level rises, the number of pair comparisons also increases, which will take much longer time and more effort to build an AHP model [42].The TOPSIS method has some drawbacks, too.TOPSIS can cause the phenomenon known as rank reversal, which is one of the problems it causes.When an alternative is added to or removed from the decision problem, the order of preference of the alternative changes.In some cases, this may result in total rank reversal, in which the order of preferences is completely inverted, i.e., the alternative considered best becomes the worst as a result of the inclusion or removal of an alternative from the process [43].ANP, on the other hand, heavily relies on expert judgment and experience to produce its results.Furthermore, in ANP, a large number of factors result in an unwieldy model [44].
These studies find importance considering the negative environmental effects caused by non-renewable energy sources and the continuous depletion of fossil fuels.Solar energy development gives a positive change in the world due to its connection with environmental sustainability.In comparison with solar energy, the development of other renewable energy sources, such as hydropower, is affected by land availability, large infrastructure costs, resettlement, rehabilitation, and environmental damage [45].Wind power is clean energy, but offshore wind turbines, electricity transmission, and operation maintenance are huge challenges.By considering the societal barrier of hydropower and some technical limitations of wind power, solar energy gains its importance for the study.
Based on the discussion so far, the present study withdraws the following objectives-• To comprehensively analyze solar PV energy, the most technologically mature and future renewable energy source.Every aspect of the case study is considered for performance analysis, such as the reasons for solar PV underperformance, and critical criteria are chosen based on those considerations.• To select the most optimum power plant site, considering different climatic zones for a more accurate comparison of solar PV systems in India.

S. Singh and S. Powar
• The current work also considers various suggestions made by regional experts, experts from various environmental associations, energy practitioners, stakeholders, and planning authorities, which will increase the research work's acceptability and viability.• Selection of fuzzy MCDM techniques for criteria weightage and ranking of alternatives.• In order to determine the data's robustness, this paper also used sensitivity analysis to see how the results would change if the criteria weights were changed.
As a result of these objectives, the proposed framework will aid decision-makers in the development of future solar power plants in a variety of ways.

Methodology
In this study, we have applied a Fuzzy MCDM approach including F-SWARA, F-COPRAS, and F-ELECTRE, utilizing the case of NTPC's solar PV plants in India to demonstrate how stakeholders can arrive at the best solar PV plant for construction while taking sustainability into account.The detailed flowchart of the study is shown in Fig. 3.
For this, seven NTPC solar PV plants have been selected, the details of which are shown in Table 1.Four major criteria are considered: economic, technical, environmental, and connectivity.Table 2 lists the main criteria and sub-criteria that were considered for the study.The main criteria of the study are selected to integrate sustainability in the problem of selection of solar power sites.The sub-criteria are chosen based on literature review and brainstorming sessions held with the practitioners from the solar industry in Northern India, especially NTPC and academicians working in the field of solar power.All of these respondents (nineteen in number) have more than 5 years of experience in the field.These experts provided their perspectives on the various criteria under consideration through interviews and questionnaires and also decided upon the significance of selected criteria over the plant location and performance.Fig. 4 also includes highlights of the experts' demography in the form of pie charts to help with visualization.As per the discussion with the experts, high investment and operation and maintenance costs, critical concerns in the solar PV energy system, and part of the economic criteria are considered the most significant study criteria.Another important economic sub-criterion is the cost of funding; however, since the government funded these plants, the cost of financing was not considered.The decision-makers also decide other criteria significance.
The methodology for the study consists of three main steps.
• Calculation of criteria and sub-criteria weights by F-SWARA • Ranking of results by F-COPRAS • Ranking and validation of results by F-ELECTRE

Calculation of criteria and sub-criteria weight by F-SWARA
In 2010, Kersuliene et al. [60], proposed the SWARA method for the very first time for selection of rational dispute resolution method.SWARA method allows decision makers to choose the significance of the evaluation criteria based on their own judgement or organizations' strategies or plans.It gives freedom to researches to remove criteria and indicators that are not so effective.Following are the steps involved in F-SWARA.
Step 1. Sort the criteria in decreasing order of importance and determine comparative importance of average value (s j ) using fuzzy scale (Electronic Supplementary Information Table S1) [61] as shown in Electronic Supplementary Information Table S2 for main-criteria and Table S3 for sub-criteria.
Step 2. To determine the coefficient value (k j ) as given in equation ( 1) Step 3. Find the recalculated weight value (q j ) as given in equation ( 2) Step 4. Determine the related weights (w j ) of the evaluation criteria following equation (3).S. Singh and S. Powar where, w j is the relative fuzzy weight of the jth criterion and n indicates the number of criteria.
The coefficient value (k j ), recalculated weight value (q j ) and related weights (w j ) for main criteria are shown in Electronic Supplementary Information Table S2 and for sub criteria it is given in Electronic Supplementary Information Table S3.
Step 5. To determine the final weight (w) for sub-criteria.
Final weight (w) as shown in Table 3 can be calculated by multiplying the weight (w j ) of main criteria (refer Electronic Supplementary Information Table S2) with weight (w j ) of sub-criteria (refer Electronic Supplementary Information Table S3).

Ranking of results by F-COPRAS
Zavadskas invented the COPRAS method in 1994 [62].The steps of F-COPRAS are as follows.
Step 1. Gather the data as per the study and construct the initial data matrix.
The initial data matrix is shown in Table 4.
Step 2. Construct the fuzzy decision matrix (refer Electronic Supplementary Information Table S4) as per the scale shown in Table S1 (refer Electronic Supplementary Information).
Step 3. Normalize the fuzzy decision matrix (f ij ).
It is formed by dividing each entry by the sum of all the entries in each column to remove anomalies with different measurement units to form dimensionless criteria.Table S5 (refer Electronic Supplementary Information) shows the fuzzy normalized decision matrix.
Step 4. Calculate the weighted normalized decision matrix (x ij ) as per equation (4) It is calculated by multiplying the weight of evaluation indicators (w j ) with normalized decision matrices (Electronic Supplementary Information Table S6).
Step 5. P i and R i values calculations.
Sums P i (refer equation ( 5)) of criteria values are the one whose larger values are required (maximise the optimization) for the computation of each alternative (line of the decision-making matrix) and Sums R i of criteria values are whose smaller values are desirable (minimize the optimization) computation for (refer equation ( 5)) each alternative (line of the decision-making matrix) (refer Electronic Supplementary Information Table S7).
Step 6. Decide the minimal value of R i as per equation (6).
Step 7. Estimate the relative weight of each alternative Q i using equation ( 7) (refer Electronic Supplementary Information Table S7).
Step 9. Allocate the priority of the alternatives.
The greater is the relative weight of alternative Q i , the higher is the priority or rank of the alternatives.In the case of Q max , the satisfaction degree is the highest.
Step 10.Determine the utility degree of each alternative using equation (9).
where Q i and Q max are the weight of the projects.Table 5 shows the utility degree value and gives rank according to these values to each alternative.

Ranking and validation of results by F-ELECTRE
ELECTRE method is one of the earliest outranking methods proposed by B. Roy in 1968 [63].It is used to resolve the ambiguity of concepts that are associated with decision maker's judgments.Electre method is one the most popular outranking methods in its family.Following are the steps for F-ELECTRE.
Step 1. Determine the initial fuzzy decision matrix.Fuzzy decision matrix is the same matrix that we used for F-COPRAS and shown in Electronic Supplementary Information Table S4.
Step 2. Computation of normalized decision matrix using equations 10-14 as shown in Electronic Supplementary Information Table S8.
Linear scale normalization is used here to ensure that all values of the decision matrix have homogeneous and comparable units.

Table 2
Considered main-criteria and its sub-criteria for the assessment and ranking of selected solar photovoltaic plants.
, for cost criteria.
Step 4. Calculate the distances between any two alternatives by using equation ( 16) (refer Electronic Supplementary Information Table S10 (al)).The weighted normalized matrix is used to prepare the concordance and discordance matrices.
Step 5. Prepare the concordance sets of criteria by referring equation ( 17) (refer Electronic Supplementary Information Table S11), and the concordance matrix (refer Electronic Supplementary Information Table S12).
The concordance set is the group of attributes for which ṽxj ≥ ṽyj .The concordance matrix is formed by summing the fuzzy weights of the attributes that are present in the concordance set. where Now calculate the concordance level as per equation ( 18): Step 6. Form a Boolean matrix E (Electronic Supplementary Information Table S13) from the concordance matrix by using equation ( 19).
Step 7. Prepare the discordance sets of criteria (Electronic Supplementary Information Table S14), and the discordance matrix by using equation 20 and 21 (Electronic Supplementary Information Table S15).
The discordance set is the set of the attributes for which ṽxj ≤ ṽyj .The discordance matrix is formed by summing the fuzzy weights of the attributes that are present in the discordance set. Where, Now find the discordance level as per equation ( 22): Step 8. Form a Boolean matrix F from discordance matrix using equation ( 23) (Electronic Supplementary Information Table S16).Step 9. Formation of the global matrix (Electronic Supplementary Information Table S17) The global matrix G is formed by multiplication of the elements of the matrices E and F as given in equation (24).

Sensitivity analysis
Sensitivity analysis is required when the data is uncertain, and the significance of the criteria weightage is decided by the decision makers.In short, data contains some degree of uncertainty and vagueness [64].Sensitivity analysis can be performed by varying the weightage of the evaluated criteria and observing how the ranking of the alternatives change [65].In this study, sensitivity analysis is performed by Super Decision software.

Table 3
Final weights of the sub-criteria identified by fuzzy stepwise weight assessment ratio analysis (F-SWARA) method.

Results and discussion
The study highlights the various reasons for the underperformance of the solar PV power plant and targets these underperforming criteria for analyzing and comparing solar PV plants to assist decision makers for future sustainable energy development.To compare the seven solar PV power plants of NTPC, four main criteria, i.e., economical, technical, environmental, and connectivity, were considered.This study gave Significance to economic criteria because of the large initial investment cost associated with the PV plants and the high operation and maintenance costs.The significance order was decided by the decision-makers who are experts in the field.The selected power plants were from different climatic zones of India, so a clearer picture can be provided to the decision-makers as solar PV plants are very much affected by climatic conditions.To compare these solar plants, hybrid fuzzy MCDM methods have been used, such as F-SWARA, F-COPRAS, and F-ELECTRE.
The initial data matrix having the values for performance and location comparison of different power plants is shown in Tables 4 and it will later be converted to a fuzzy data matrix as in Electronic Supplementary Information Table S4 with the help of scale for the importance rating (refer Electronic Supplementary Information Table S1).The F-SWARA algorithm was used in this study to determine the weightage of the criteria, as shown in Table 3.The F-COPRAS and F-ELECTRE methods were then applied to the initial data matrix to rank the solar PV plants.
Table 7 displays the F-COPRAS and F-ELECTRE ranking results.The rankings produced by these two MCDM methods are similar, but not identical.According to F-COPRAS, the first-ranked solar PV is Unchahar solar PV, and according to F-ELECTRE, it could be Unchahar or Singrauli plant because both these alternatives outrank all other alternatives in its evaluation.After validating the results of F-COPRAS with F-ELECTRE, it is possible to conclude that Unchahar solar PV power plant is the best alternative in terms of performance, location, development, and management, while Faridabad solar PV and Port Blair solar PV plant are the least preferred plants for the same.After validating the results with MCDM methods, sensitivity analysis has been performed to find the robustness of the data.

Ranking results by MCDM
The findings favour Unchahar solar PV because both the LCOE and total cost criteria that fall under the economic criteria were chosen as the most significant criteria for the study.The Unchahar solar PV power plant has the lowest levelized cost of energy (0.00011 USD/kWh) and total cost (579263 USD/MW), clearly outperforming other plants.Also, based on Tables 4 and it is possible to conclude that Unchahar solar PV plants have the second highest overall efficiency (14.15%) which comes under the second most important criterion, technical criteria.In comparison to other solar PV plants, the amount of CO 2 reduction (environmental criteria) in the case of the Unchahar solar PV plant (1389.5 metric tonnes of CO 2 equivalent per annum per MW) is the third best.Regarding location, Unchahar solar PV also has very good connectivity values, with the nearest rail head connectivity (5 km) and nearest city/ town connectivity (3.8 km).
Faridabad solar PV and Port Blair solar PV rank last because the former has the second highest associated LCOE of 0.00024 USD/kWh, total cost of 1346185 USD/MW, and operation and maintenance cost of 7215 USD/MW per year, while Port Blair solar PV has the highest LCOE of 0.00027 USD/kWh, total cost of 1402455 USD/MW, and operation and maintenance cost of 10096 USD/MW per year, but Port Blair solar PV at the same time is also having the highest overall efficiency of 16.43% and lowest average maximum temperature of 33 • C which is desirable and place it above Faridabad solar PV in rank.
Considering the solar PV power plants from the same climatic zone, such as Ramagundam solar PV and Singrauli solar PV from hot and dry climatic zone.Singrauli solar PV is at second place while Ramagundam solar PV is at third place.The reason is that along with the location factors (average temperature and average global solar radiation) in other critical criteria such as LCOE, total cost and operation, and maintenance cost, the difference in values are negligible.However, while comparing Unchahar solar PV, Dadri solar PV, and Faridabad solar PV from the composite climatic zone, the difference is ranking is noticeable as Unchahar solar PV is ranked one while Faridabad solar PV and Dadri solar PV is at last position.This is because economic criteria being the most significant criteria played it role and due to the noticeable difference between the value of LCOE, total cost and operation and maintenance cost criteria of Unchahar solar PV (0.00011 USD/kWh, 579263 USD/MW and 5600 USD/MW per year, respectively) in

Comparison and validation of findings
During the thorough literature review, it has been found that no similar studies have been conducted considering PV plants from different climatic zones of India, focusing mainly on underperforming indicators for solar PV power plants.However there were few studies comparing the location of various energy sources and considering other similar objectives.
One such study of finding the suitable location for solar and wind plants conducted by Saraswat et al. found that Rajasthan has the highest and Uttar Pradesh has the second highest 'highly suitable' land for solar farms.However, the study primarily focused on location only and considered different criteria mainly inclined towards connectivity.
Comparing this with the present case study, Unchahar solar PV plant located in Uttar Pradesh is found to be the best location.Also, the present case study used comparatively more extensive criteria that touched every aspect of sustainability and covered a wide range of factors, such as different climatic zones for fair comparison among power plants for the location [33].Another study conducted by Kanwar et al. investigated that the high initial cost for renewable energy plant installations in India is the greatest challenge.As a developing country, India relies on investments by both the government and private firms for its economic growth.The government and the private sector are discouraged by the high initial cost of renewable energy and the lack of surety of getting a return on the investment.Hence, this paper gives the most significance to the economic criteria [67].

Sensitivity analysis results
Fig. 6 depicts the sensitivity analysis results for all twelve criteria (al).Fig. 6(a-l) shows that the sensitivity analysis results are robust, as changing the weights has a negligible effect on the ranking results, and the Unchahar solar PV plant remains in the first place.For every criterion, the results of sensitivity analysis are almost the same.However, from Fig. 6 (c) and Fig. 6 (k) one can observe that operation and maintenance cost criteria and nearest railway head criteria can be significant criteria for ranking, and the results might change with the change in the criterion value.The worst-performing plant after sensitivity analysis is the Port Blair solar PV plant.Comparing Port Blair solar PV and Faridabad solar PV (last rank according to ranking results by MCDM), later showing more robustness towards sensitivity analysis and showing better results after Unchahar and Singrauli solar PV.

Research limitations
The research has been carried out with the data available from the existing literature, the NTPC website, other internet resources, and through NTPC field visits but the data is not as much updated as commercially available data.The results can be further improved or more criteria can be accommodated for the analysis with well-defined data.However, the uncertainties bound to happen by these is taken care of in this study using improved fuzzy MCDM methods.

Conclusions and prospects
Energy demand is increasing in India as the population rises, and power supply is not catching up.As environmental concerns have risen, the role of environmental factors in energy generation has grown.Considering these other factors, renewable energy is an option to increase production capacity.Among renewable energy, solar energy is one of the cleanest, emission-free, and has the least effect on society.However, solar energy plants are typically underperforming due to various factors, such as the selection of the wrong technology or the wrong location, the environment, pollution, extreme weather, and dirt.
This paper evaluates seven NTPC solar PV plants in India as MCDM problems, with the goal of determining the best-performing solar PV plant in terms of cost, environment, technicality, and location.The power plant that was selected is from a different climatic zone of India to analyze the effect of location on the performance as well.The goal of this research is to make a strategic contribution to decision-making.A novel fuzzy MCDM model approach has been proposed to prioritize solar PV plants based on four main criteria and twelve sub-criteria.The criteria chosen are sustainable criteria in decreasing order of importance: economic, technical, environmental, and connectivity criteria.The F-SWARA method calculates the criteria weights, and the F-COPRAS and F-ELECTRE methods are used to calculate the ranking.The alternative 'Unchahar solar PV' is the best option available, according to the F-COPRAS ranking and F-ELECTRE validation results.Furthermore, a sensitivity analysis was carried out to demonstrate the impact of changing the weightage of criteria on alternative ranking results.It was also discovered that the ranking results did not change significantly after performing sensitivity analysis, indicating that the data is robust.This methodology will assist decision-makers in locating suitable locations for solar power plants (taking into account various climatic zones).Other regions can be considered, and this methodology can be used to identify favorable climatic zones for power plant deployment.This method can also be applied to other renewable energy sources for locating suitable land for energy generation by examining various sustainability indicators.The current framework can also aid in defining the energy policy's goal of ensuring energy security.A sound energy policy will assist states and localities in reducing harmful air pollutants, improving public health, lowering energy costs and the costs of complying with national air quality standards, creating jobs, and improving the nation's energy system's reliability and security.
While the method used in this study assists decision-makers in identifying suitable locations for solar PV plants, additional qualitative factors can be incorporated to ensure a thorough analysis.This study excludes social factors because the societal effects of establishing solar power facilities are generally minor compared to larger infrastructure projects such as hydroelectric plants.Nonetheless, researchers can address this gap by including social criteria in their research, in addition to economic and environmental benchmarks.The readers can enhance the study based on the extended exergy accounting performance method for hybrid or single renewable energy.Exergoeconomic and exergoenvironmental approaches can be considered reliable indicators for locating and sizing photovoltaic solar farms in future studies, along with the criteria given in the current study for comprehensive research.As a future study recommendation, the presented technique can be applied to various decision-making issues.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 6 .
Fig. 6.Sensitivity analysis plot for (a) LCOE (b) Total cost (c) Operation and maintenance cost (d) Overall efficiency (e) Plant load factor (f) Energy generation (g) Average global solar radiation (h) Average maximum temperature (i) Amount of reductions (j) Nearest NH connectivity (k) Nearest Rail head and (l) Nearest city/town.

Table 1
General specification of solar photovoltaic plants selected for comparison study.

Table 4
Decision matrix for seven alternatives (A1-A7) of solar photovoltaic plants (E1-E3 represent sub-criteria of economic criteria, T1-T5 represent sub-criteria of technical criteria, EN1 represent sub-criteria of environmental criteria and C1-C3 represent sub-criteria of connectivity criteria).

Table 5
Ranking results of alternatives by fuzzy complex proportional assessment (F-COPRAS) method.

Table 6
Ranking result of alternatives by fuzzy ELimination Et Choice Translating REality (F-ELECTRE) method.

Table 7
[66]ingh and S. Powarcomparison with Faridabad (0.00024 USD/kWh, 1346185 USD/MW and 7215 USD/MW per year respectively) and Dadri solar PV (0.00024 USD/kWh, 1308222 USD/MW and 4092 USD/MW per year respectively), the latter two are at last positions.Only the operation and maintenance cost for Dadri solar PV (4092.57USD/MWperyear) is less in comparison with Unchahar (5600.35USD/MWperyear) and Faridabad solar PV (7215.84USD/MWperyear) due to installation of 64 robots for dry cleaning on daily basis[66], which place Dadri solar PV at better rank than Faridabad solar PV but not above Unchahar solar PV because of very high LCOE and total cost of Dadri solar PV in comparison to Unchahar solar PV.