Sustainable Component-Level Prioritization of PV Panels, Batteries, and Converters for Solar Technologies in Hybrid Renewable Energy Systems Using Objective-Weighted MCDM Models

Abstract

Data-driven prioritization of photovoltaic (PV), battery, and converter technologies is crucial for achieving sustainability, efficiency, and cost-effectiveness in the increasingly complex domain of hybrid renewable energy systems (HRES). Conducting an in-depth and systematic ranking of these components for solar-based HRESs necessitates a comprehensive multi-criteria decision-making (MCDM) framework. This study develops as the most recent and integrated approach available in the literature. To ensure balanced and objective weighting, five quantitative weighting techniques, Entropy, Standard Deviation, CRITIC, MEREC, and CILOS, were aggregated through the Bonferroni operator, thereby minimizing subjective bias while preserving robustness. The final ranking was executed using the measurement of alternatives and ranking according to compromise solution method (MARCOS). Subsequently, comparative validation was conducted across eight additional MCDM methods, supplemented by correlation and sensitivity analysis to evaluate the consistency and reliability of the obtained results. The results revealed that thin-film PV modules (0.7108), hybrid supercapacitor batteries (0.6990), and modular converters (1.1812) emerged as the top-performing technologies, reflecting optimal trade-offs among technical, economic, and environmental performance criteria. Correlation analysis (ρ > 0.9 across nine MCDM methods) confirmed the stability of the rankings. The results establish a reproducible decision-support framework for designing sustainable hybrid systems. These technologies demonstrated superior thermal stability, cycling endurance, and system scalability, respectively, thus laying a foundation for more sustainable and resilient hybrid energy system deployments. The proposed framework provides a reproducible, transparent, and resilient decision-support tool designed to assist engineers, researchers, and policy-makers in developing reliable low-carbon components for the realization of future carbon-neutral energy infrastructures.

1. Introduction

Sustainable development agendas today position global decarbonization and energy security as core priorities, with renewable energy systems projected to supply over two-thirds of global electricity demand by 2050. While substantial progress is being achieved across multiple fronts, the clean energy transition remains constrained by factors such as resource intermittency, geographical variability, and the absence of cost-effective, high-performance components capable of meeting system-level engineering demands. Consequently, hybrid renewable energy systems (HRES) have emerged as an integrated solution, combining synergistic resources including solar, biomass, wind, hydrogen, and batteries to maximize operational stability and minimize life-cycle emissions [1]. As nations worldwide strive to reduce carbon emissions amid rising energy demand, the deployment of solar photovoltaic (PV) technologies has accelerated rapidly. The modularity, scalability, and continual cost reductions of PV systems have established them as the cornerstone of HRES development. However, the realization of their full potential depends critically on the strategic selection of technology components, specifically PV modules, battery storage systems, and power converters, each of which presents distinct techno-economic and environmental trade-offs. Thus, informed decision-making in component selection is essential for ensuring efficient, resilient, and sustainable hybrid energy system design [2].

The rapid global deployment of PV systems has shifted the primary challenge from capacity expansion to the context-sensitive evaluation of module selection and assurance of long-term performance. Badran and Dhimish [3] reported that vertically mounted bifacial HJT arrays increased morning and evening generation, improving annual yield compared with corresponding monofacial modules in a year-long deployment in the UK. Likewise, Iturralde Carrera et al. [4] observed that east–west-oriented rooftop installations in hotel applications enhanced performance ratios and rooftop utilization, producing notable techno-economic and environmental gains on a per-site basis. These results reinforce that PV selection extends beyond nominal efficiency values, representing a coupling of configuration, technology, and site conditions, which the present component-level analytical framework explicitly accounts for.

Module reliability is inherently dictated by material composition and encapsulation design. A 1.2 MWp field investigation by Buerhop et al. [5] associated potential-induced degradation (PID) with specific EVA/backsheet combinations and high Na⁺ ingress, identifying polymer selection as a first-order design variable influencing degradation. In the silicon family, Wu et al. [6] demonstrated that TOPCon cell reliability has been significantly enhanced through laser-integrated contact firing, which enables near-zero aluminum pastes and markedly reduces damp-heat degradation relative to baseline cells. On the tandem frontier, Castriotta et al. [7] reported that spin-free, four-terminal (4T) perovskite/silicon submodules achieved ISOS-L1 stability with T₈₀ ≈ 1459 h on a 225 cm² aperture, indicating credible pathways for industrial scale-up. These findings correct earlier overestimations of efficiency and form the empirical basis for a durability-aware ranking of PV technologies within the present analysis.

Technology selection in PV systems is increasingly governed by sustainability and material circularity considerations. Cheema et al. [8] emphasized that PV recycling plays a vital role in recovering Ag, In, and Te, mitigating polymer–metal interface risks, and advancing techno-economic feasibility via delamination and hydrometallurgical pathways. Konstantinavičiūtė et al. [9] demonstrated that glass–glass modules exhibit lower life-cycle greenhouse gas (GHG) emissions during manufacturing while maintaining a comparable levelized cost of energy levelized cost of electricity (LCOE) in cold climates relative to standard glass-backsheet panels. Future projections by Xu et al. [1] show considerable variability in Ag and In demand by 2050, contingent upon the technology mix (e.g., HJT or 4T tandems) and recycling penetration levels, with closed-loop recycling potentially reducing primary material demand by 10–30%. Consequently, energy payback time (EPBT), CO₂ intensity, and circularity readiness are treated here as core evaluation criteria rather than secondary considerations. Simultaneously, innovations in automated quality assurance, such as lightweight CNN-based electroluminescence (EL) defect classification [10], further enhance early-life diagnostics and reduce the probability of latent defects. Advanced optimization methods for integrated energy systems have also been increasingly explored in recent studies. A physics-informed deep reinforcement learning-based bi-level programming framework has been proposed for microgrid scheduling, enabling optimal reconciliation of the interests of both operators and users while leveraging demand response and load flexibility. This approach has been shown to significantly outperform conventional optimization techniques in terms of both economic benefits and computational efficiency [11]. Moreover, multi-objective optimization has become indispensable in modern energy system planning, as it is essential to integrate cost considerations with environmental impact, reliability, and stability in decision-making. Recent survey studies have emphasized the critical role of MCDM techniques in effectively selecting among Pareto-optimal solutions and thereby facilitating sustainable system design [12].

Converter selection inherently involves application-specific trade-offs beyond nominal efficiency ratings. Under irradiance variability, techniques such as voltage-lift/multilevel inverter (MLI) and evolutionary-optimized asymmetric CHB-MLI topologies have effectively minimized total harmonic distortion (THD) [13]. Isolated CHB-MLC architectures with high-frequency transformers can mitigate DC-link ripple and reduce capacitor volume while ensuring compliance with medium-voltage grid codes. For grid-support applications, phase-shifted phase-locked loop (PLL) controls are employed to maximize power factor through PV-STATCOM strategies [14]. Moreover, hybrid IGBT stacks have improved inverter reliability under realistic mission profiles, with bifacial array compatibility demonstrated for the first time [15]. These studies collectively establish the converter topology and control-criteria foundation used in our MCDM-based evaluation, while more detailed discussions on modular multilevel converter (MMC) advances from 2022–2024 are retained within the Literature Review Section for brevity.

Recent advances in converter modulation and fault-tolerant design have further strengthened the case for modular topologies in hybrid renewable architectures. Li et al. [16] proposed a reduced-ripple overmodulation scheme for indirect matrix converters, achieving superior harmonic suppression without additional sensors. Li et al. [17,18] optimized single- and dual-bridge converter modulation to enhance light-load efficiency and dynamic performance. Complementary to these control-level strategies, Zeng et al. [19] and Gao et al. [20] introduced adaptive reconfiguration and hybrid integration techniques that improved reliability and simulation precision for MMC systems. Jing et al. [21] expanded modular converter applications to offshore HVDC systems through a hybrid DC-chopper configuration, underscoring the scalability potential of multilevel architectures for next-generation HRESs.

Taken together, the evidence on performance, reliability, and circularity emphasizes the need for a component-level decision framework. PV families, storage technologies, and converter topologies are comparatively evaluated under a unified rubric encompassing technical, environmental, economic, and reliability dimensions. Conventional HRES-level assessments are insufficient to capture the variability introduced by encapsulant chemistry, metallization processes, mission profiles, and grid-service requirements. Hence, a transparent, weighted trade-off methodology as developed in this study is essential for establishing a competitive and evidence-based hierarchy among emerging energy technologies.

Table 1 summarizes recent exemplary applications of Multi-Criteria Decision-Making (MCDM) methods in the context of renewable and hybrid energy systems. It emphasizes methodological purpose, contextual application, and remaining limitations, thereby establishing a thematically analytical foundation for this work. Studies such as Aljaghoub et al. [22] and AlMallahi et al. [23] represent early contributions to sustainability-oriented MCDM applications. Their analyses demonstrated how operational parameters such as water use, cost, safety, and energy savings can be quantified as measurable decision metrics rather than treated qualitatively.

More recent studies, such as Ajithkumar and Poongavanam [24], employed an ensemble of six MCDM models, TOPSIS, EDAS, MARCOS, VIKOR, CoCoSo, and MOORA, in conjunction with Spearman rank correlation and Borda aggregation to assess ranking stability. Their multi-method triangulation approach minimized subjectivity and improved the accuracy of optimal phase-change material identification for solar drying systems. It also reported the validation power of cross-method analysis, and similar principle explicitly adopted in the present study’s correlation and sensitivity analyses.

In parallel, [25,26,27] applied MCDM to hybrid-system configuration and site-selection problems, often integrating simulation tools such as HOMER Pro with ranking frameworks like CoCoSo, LBWA, and entropy weighting. Their results revealed that the relative significance of economic, technical, and environmental criteria is highly context-dependent, varying with geographical, demographic, and resource-based factors. These findings emphasize the necessity for adaptive weighting schemes over static, expert-assigned judgments. However, these system-level investigations provide limited guidance regarding component-level characteristics, such as converter topologies or battery chemistries.

Emerging studies have begun to address materials and technology-specific applications. Tajik et al. [28] proposed hybrid MCDM–DEA models for ranking cathode materials in lithium-ion batteries using integrated AHP–CILOS–MARCOS–CoCoSo approaches. Sadeghitabar, et al. [29] employed an entropy-based TOPSIS framework for optimizing renewable energy generation in greenhouse microgrids. These works exemplify a growing trend toward objective weighting techniques (e.g., Entropy, MEREC, CILOS) and the explicit treatment of uncertainty via fuzzy, neutrosophic, or probabilistic extensions. Similarly, Li et al. [30] combined entropy weighting with TOPSIS or CODAS to evaluate hydrogen-production-integrated off-grid hybrid systems, confirming the centrality of reliability indicators such as the Loss of Power Supply Probability (LPSP). Nevertheless, these methodological advances seldom extend beyond system-level architectures. The specific component choices, for example, thin-film versus HJT PV modules, hybrid supercapacitors versus Li-ion batteries, or modular versus neutral-point-clamped converters, remain largely unexplored.

From a methodological perspective, as depicted in Table 1, most studies rely on a single weighting technique, either subjective (AHP) or objective (Entropy), without reconciling the two. Only a few studies, such as [31,32], have attempted to integrate entropy weighting with sensitivity or Monte Carlo simulation to evaluate rank stability. Although hybrid algorithms like MARCOS, CoCoSo, and VIKOR have gained traction in producing compromise solutions, comprehensive inter-method correlation and sensitivity analyses remain rare. Earlier MCDM applications, though methodologically insightful, exhibited fragmentation and lacked bidirectional validation.

The present study consolidates these methodological developments within a unified analytical framework. Through a structured decomposition, it proposes a Bonferroni-fused objective weighting ensemble comprising Entropy, CRITIC, MEREC, CILOS, and Standard Deviation coupled with the MARCOS method to balance robustness and objectivity. More importantly, the framework evaluates the PV–battery–converter triad at the component level using shared technical, environmental, economic, and reliability metrics. The incorporation of sensitivity and correlation analyses, following the approaches of [24,32], further enhances the credibility and stability of the results.

Overall, as illustrated in Table 1, MCDM methodologies have become integral to energy decision-making at the system level, yet their component-level implementation remains nascent. Through a holistic and reproducible framework that ensures both robustness and accuracy, this study transforms the literature from a qualitative overview into a quantitative assessment of technology trade-offs. By aligning the methodological heterogeneity observed in prior studies, it establishes a transparent and context-sensitive foundation for optimizing future HRESs.

Table 1. Comparative synthesis of related studies versus the present work.

Focus and Context	Criteria/Methods	Core Findings	Limitations/Gaps	How This Paper Advances Knowledge	Ref.
PCM selection for sustainable solar drying	Six MCDM methods (TOPSIS, EDAS, MOORA, MARCOS, CoCoSo, VIKOR) + Borda aggregation + Spearman rank correlation + sensitivity	validated stability of ranking	Limited to thermal energy applications	Demonstrates multi-model and robustness analysis strategy adopted here for PV–Battery–Converter selection	[24]
Hybrid energy systems for EV charging by demographic groups	HOMER optimization + MCDM ranking of on/off-grid scenarios by economic, technical, environmental criteria	PV/Wind/Battery/Converter system best for daytime users; PV/Wind/Grid best for night users	Sector-specific; no component-level prioritization	Reinforces the need for context-sensitive criteria weighting in MCDM frameworks	[26]
Off-grid HRESs for wetland areas (WIL-CoCoSo)	HOMER simulation + LBWA for weights + Wins-in-League (CoCoSo) ranking + sensitivity	PV/WT/DG/HKT + Li-ion battery optimal; flood-resilient; CO2 ↓96% vs. DG-only	Region-specific; no component-type analysis	Adds resilience and environmental criteria to decision matrix for HRES design	[25]
PV cleaning techniques and SDG alignment	18 criteria linked to SDGs (6–13); TOPSIS ranking	Manual cleaning is best for energy and water SDGs	Operation-phase only; not tech-family selection	Illustrates integration of sustainability indices into MCDM framework	[22]
PV panel cleaning in UAE (sustainability)	TOPSIS + entropy + stochastic dominance + sensitivity analysis	Robot water-based cleaning method top-ranked (0.65–0.75)	Survey-based; local bias	Validates objective–subjective weight fusion approach similar to ours (Bonferroni fusion)	[23]
Site selection of on-grid HRESs	GIS spatial filters + MCDM (technical, economic, environmental, climatic criteria) + TEA	Identified Izadkhast as optimal site with 60% RE fraction	Location-focused; no component ranking	Demonstrates structured criteria taxonomy, later adapted for PV/Battery/Converter evaluation	[27]
PV/Wind/Storage for RO desalination	FAHP (weighting) + Fuzzy-VIKOR (ranking)	Fully renewable design (100% RES); NPC 0.091 $/kWh	Water-sector application	Extends fuzzy-AHP/VIKOR combination to energy technology selection	[33]
Solar-strategy prioritization at neighborhood scale	Expert survey + adoption-score MCDM tool	Framework for passive/active solar integration in urban design	Planning domain; not component-level	Validates stakeholder-driven criteria weighting adopted for our fused objective weights	[34]
Industrial HRES sustainability optimization (4E)	Multi-objective Pareto front + MCDM post-ranking of configurations by 4E indices	Solar + Wind most sustainable (SI = 0.89)	Macro-level system view	Translates economic, environmental and technical indices into component criteria	[35]
HRESs with hydrogen & battery (PMS optimization)	Entropy weight method (EWM) + CODAS ranking (9 criteria)	PV/WT/Battery/H2 configurations evaluated; LPSP highest weight	System-level focus only	Adds PMS/aging/reliability dimensions to converter and storage criteria	[30]
Island microgrid planning (China)	HOMER + reference-point MCDM + uncertainty analysis	PV-WT-DG-Battery mix best under resilience criteria	Case-study only	Embeds resilience and uncertainty handling in criteria design	[36]
Off-grid HRESs with green hydrogen production	TOPSIS ranking of six configurations using techno-economic and environmental metrics	PV-WT-BG-Battery-H2 system top rank (RC = 0.817)	No intra-technology comparison	Establishes multi-criteria trade-off structure for PV/Battery/Converter families	[37]
MCDM evaluation of renewable systems with hysteresis control	Hybrid Entropy + CODAS approach; 9 technical/economic/env. indicators	Quantified weights for PV, WT, Battery, H2 systems; LPSP dominant	Energy-system scale; no component family distinction	Demonstrates objective entropy weighting replicated in our Bonferroni-fusion framework	[30]
DSS for PV module and battery selection	AHP + TOPSIS vs. Entropy + TOPSIS comparative decision support	Li-ion + CIGS pair optimal across methods; rank consistency verified	Limited criteria breadth	Confirms robustness of objective-weight fusion and sensitivity testing applied in this study	[38]

↓ Decreasing.

Table 1. Comparative synthesis of related studies versus the present work.

Focus and Context	Criteria/Methods	Core Findings	Limitations/Gaps	How This Paper Advances Knowledge	Ref.
PCM selection for sustainable solar drying	Six MCDM methods (TOPSIS, EDAS, MOORA, MARCOS, CoCoSo, VIKOR) + Borda aggregation + Spearman rank correlation + sensitivity	validated stability of ranking	Limited to thermal energy applications	Demonstrates multi-model and robustness analysis strategy adopted here for PV–Battery–Converter selection	[24]
Hybrid energy systems for EV charging by demographic groups	HOMER optimization + MCDM ranking of on/off-grid scenarios by economic, technical, environmental criteria	PV/Wind/Battery/Converter system best for daytime users; PV/Wind/Grid best for night users	Sector-specific; no component-level prioritization	Reinforces the need for context-sensitive criteria weighting in MCDM frameworks	[26]
Off-grid HRESs for wetland areas (WIL-CoCoSo)	HOMER simulation + LBWA for weights + Wins-in-League (CoCoSo) ranking + sensitivity	PV/WT/DG/HKT + Li-ion battery optimal; flood-resilient; CO2 ↓96% vs. DG-only	Region-specific; no component-type analysis	Adds resilience and environmental criteria to decision matrix for HRES design	[25]
PV cleaning techniques and SDG alignment	18 criteria linked to SDGs (6–13); TOPSIS ranking	Manual cleaning is best for energy and water SDGs	Operation-phase only; not tech-family selection	Illustrates integration of sustainability indices into MCDM framework	[22]
PV panel cleaning in UAE (sustainability)	TOPSIS + entropy + stochastic dominance + sensitivity analysis	Robot water-based cleaning method top-ranked (0.65–0.75)	Survey-based; local bias	Validates objective–subjective weight fusion approach similar to ours (Bonferroni fusion)	[23]
Site selection of on-grid HRESs	GIS spatial filters + MCDM (technical, economic, environmental, climatic criteria) + TEA	Identified Izadkhast as optimal site with 60% RE fraction	Location-focused; no component ranking	Demonstrates structured criteria taxonomy, later adapted for PV/Battery/Converter evaluation	[27]
PV/Wind/Storage for RO desalination	FAHP (weighting) + Fuzzy-VIKOR (ranking)	Fully renewable design (100% RES); NPC 0.091 $/kWh	Water-sector application	Extends fuzzy-AHP/VIKOR combination to energy technology selection	[33]
Solar-strategy prioritization at neighborhood scale	Expert survey + adoption-score MCDM tool	Framework for passive/active solar integration in urban design	Planning domain; not component-level	Validates stakeholder-driven criteria weighting adopted for our fused objective weights	[34]
Industrial HRES sustainability optimization (4E)	Multi-objective Pareto front + MCDM post-ranking of configurations by 4E indices	Solar + Wind most sustainable (SI = 0.89)	Macro-level system view	Translates economic, environmental and technical indices into component criteria	[35]
HRESs with hydrogen & battery (PMS optimization)	Entropy weight method (EWM) + CODAS ranking (9 criteria)	PV/WT/Battery/H2 configurations evaluated; LPSP highest weight	System-level focus only	Adds PMS/aging/reliability dimensions to converter and storage criteria	[30]
Island microgrid planning (China)	HOMER + reference-point MCDM + uncertainty analysis	PV-WT-DG-Battery mix best under resilience criteria	Case-study only	Embeds resilience and uncertainty handling in criteria design	[36]
Off-grid HRESs with green hydrogen production	TOPSIS ranking of six configurations using techno-economic and environmental metrics	PV-WT-BG-Battery-H2 system top rank (RC = 0.817)	No intra-technology comparison	Establishes multi-criteria trade-off structure for PV/Battery/Converter families	[37]
MCDM evaluation of renewable systems with hysteresis control	Hybrid Entropy + CODAS approach; 9 technical/economic/env. indicators	Quantified weights for PV, WT, Battery, H2 systems; LPSP dominant	Energy-system scale; no component family distinction	Demonstrates objective entropy weighting replicated in our Bonferroni-fusion framework	[30]
DSS for PV module and battery selection	AHP + TOPSIS vs. Entropy + TOPSIS comparative decision support	Li-ion + CIGS pair optimal across methods; rank consistency verified	Limited criteria breadth	Confirms robustness of objective-weight fusion and sensitivity testing applied in this study	[38]

↓ Decreasing.

Research Gaps, Questions, Objectives, and Significance

Two key research gaps emerge from the reviewed literature. First, although MCDM methods are widely adopted in renewable energy planning, their application to component-level prioritisation remains underexplored. Second, existing studies typically evaluate PV modules, storage batteries, and converters independently, often lacking an integrated perspective.

• Research Question

The central research question guiding this study is: How can multi-criteria decision-making techniques be employed to prioritise photovoltaic, battery, and converter technologies, thereby enabling the optimal design of solar-based HRESs?

To address these gaps, the present study develops a multi-domain decision matrix for solar-based HRESs. It applies MARCOS with Bonferroni-fused objective weights, thereby enabling the robust prioritization of PV, battery, and converter technologies, which leads to the achievement of the research objectives.

• Research Objectives

Based on the identified gaps, the specific objectives of this study are:

• To critically evaluate photovoltaic, battery, and converter technologies relevant to solar-based HRESs while considering technical, economic, environmental, and reliability criteria.
• To apply an advanced MCDM framework (MARCOS with fused weighting) for systematic prioritization of component technologies.
• Conduct correlation and sensitivity analyses to assess the consistency and robustness of ranking outcomes across diverse decision scenarios.
• Significance of the Study

Although the renewable energy literature has increasingly adopted system-level optimisation approaches (e.g., for sizing, siting, or policy formulation), this study deliberately redirects attention towards the constituent-level, focusing on the prioritisation of critical components. This micro-level perspective provides a valuable complement to macro-level optimisation studies.

By integrating MCDM techniques, the study develops a transparent and reproducible framework for the optimal selection of PV, battery, and converter technologies, minimising trade-offs among efficiency, cost, and sustainability. The findings are expected to provide actionable insights for policymakers, engineers, and energy planners, ultimately contributing to the improved design of solar-based HRESs and advancing global objectives in clean energy transitions, energy security, and climate change mitigation.

2. Materials and Methods

2.1. MCDM Methodology: Prioritizing PV, Battery, and Converter Technologies

The framework integrates complementary MCDM techniques to establish a comprehensive approach for prioritizing PV, battery, and converter technologies. So, the MCDM method used is the Measurement of Alternatives and Ranking according to Compromise Solution (MARCOS) [39,40]. As illustrated in Figure 1, the methodology begins with the construction of a decision matrix (Equation (1)), where criteria are classified as either benefits or costs. The normalization process (Equations (2) and (3)), combined with the inclusion of ideal and anti-ideal solutions (Equation (4)), ensures comparability across criteria expressed on different scales and establishes consistent performance baselines.

Weight assignment plays a critical role in MCDM, as it reflects the relative importance of each criterion. In this study, five objective weighting schemes were employed: Entropy (Equations (5)–(7)), Standard Deviation (Equations (8) and (9)), CRITIC (Equations (10)–(12)), MEREC (Equations (13)–(15)), and CILOS (Equations (16) and (17)). Each method is grounded in distinct principles: Entropy emphasizes informational diversity, Standard Deviation captures variability, CRITIC accounts for contrast intensity and inter-criterion correlation, MEREC evaluates performance projections with and without individual criteria, and CILOS highlights the performance dip caused by the removal of specific alternatives [41,42].

To avoid dependence on any single methodological philosophy, these weights were integrated using the Bonferroni operator (Equation (18)), thereby achieving a balanced compromise. The MARCOS method (Equations (19)–(25)) was selected as it simultaneously references both ideal and anti-ideal solutions, offering greater robustness than approaches that rely solely on a single benchmark. For validation, additional MCDM methods, ARAS, COCOSO, COPRAS, EDAS, TOPSIS, VIKOR, WASPAS, and PROMETHEE-II, were applied using the same fused weights [43,44,45,46]. This ensured that the observed rankings were not artifacts of a single methodological assumption.

Furthermore, correlation analysis (Equations (26) and (27)) was employed to quantify the level of agreement among different methods, while sensitivity analysis (Equations (28)–(35)) assessed the stability of rankings under weight perturbations. By combining these approaches, the proposed hybrid framework enhances transparency, minimizes methodological bias, and delivers more reliable and accurate decision outcomes for sustainable HRES technology selection.

• Step 1: Construct the initial decision matrix; PV, battery, and converter technologies

Let the decision matrix be as per Equation (1):

$$D={\left[d_{ij}\right]}_{m\times n}$$

(1)

where $d_{ij}$ is the value of alternative i under criterion j; m: number of alternatives; and n: number of criteria.

• Step 2: Determine the criteria types of PV, battery, and converter technologies

For each criterion j:

$$\mathrm{If} j \mathrm{is} \mathrm{benefit-type}: t_j=1$$

$$\mathrm{If} j \mathrm{is} \mathrm{cost-type}: t_j=-1$$

• Step 3: MARCOS method

The MARCOS technique has been selected as a primary tool because it incorporates both ideal and anti-ideal solutions [47]. This two-anchor comparison provides greater robustness than conventional methods that rely solely on a single reference point. For instance, when all alternatives exhibit similar rankings across criteria, approaches such as TOPSIS or VIKOR may struggle to yield meaningful distinctions. By evaluating performance relative to both extremes, MARCOS enhances differentiation among alternatives and mitigates the rank reversal phenomenon. Its validity was further confirmed through comparative ranking against eight established MCDM methods and correlation testing, which demonstrated high consistency.

Normalize the decision matrix of PV, battery, and converter technologies, for benefit-type criteria use Equation (2) and for cost-type criteria use Equation (3).

$$n_{ij}={\displaystyle \frac{d_{ij}}{d_j^{AI}}}$$

(2)

$$n_{ij}={\displaystyle \frac{d_j^{AI}}{d_{ij}}}$$

(3)

where $d_j^{AI }$ is the best value among alternatives for the criterion $j$

• Step 4: Add anti-ideal and ideal alternatives as let as per Equation (4).

$$d_{i^{-}}=\left[\min d_{ij}\right] \mathrm{and} d_{i^{+}}=\left[\max d_{ij}\right]$$

(4)

These are appended to the matrix to form an extended matrix.

• Step 5: Subjective weights computation: Application of weighting methods (Entropy, standard deviation, CRITIC, MEREC, and CILOS). Each method calculates weights $w_j$ for each criterion:
• (a). Entropy Weight: Normalize the decision matrix by Equation (5), calculate the entropy of each criterion by Equation (6), and determine weights by Equation (7) [29].

$$p_{ij}={\displaystyle \frac{d_{ij}}{{\sum }_{i=1}^m{d}_{ij}}}$$

(5)

$$e_j=-k\sum _{i=1}^m{p}_{ij}\cdot \ln \left(p_{ij}\right)$$

(6)

$$w_j={\displaystyle \frac{1-e_j}{{\sum }_{j=1}^n\left(1-e_j\right)}}$$

(7)

• (b). Standard Deviation Weight: Find the standard deviation of criterion j by Equation (8), and the weight of criterion j using Equation (9) [48].

$${\mathrm{\sigma }}_j=\sqrt{{\displaystyle \frac{1}{m}}\sum _{i=1}^m{\left(d_{ij}-\overline{d_j}\right)}^2}$$

(8)

$$w_j={\displaystyle \frac{{\mathrm{\sigma }}_j}{{\sum }_{j=1}^n{\mathrm{\sigma }}_j}}$$

(9)

• (c). CRITIC Weight: Compute correlation coefficients between criteria using Equation (10), compute information content of criterion j by Equation (11), and normalize to obtain weights utilizing Equation (12) [49].

$$r_{jk}=\mathrm{Correlation} \mathrm{between} j \mathrm{and} k$$

(10)

$$C_j={\mathrm{\sigma }}_j\cdot \sum _{k=1}^n\left(1-r_{jk}\right)$$

(11)

$$w_j={\displaystyle \frac{C_j}{{\sum }_{j=1}^n{C}_j}}$$

(12)

• (d). MEREC Weight: Compute the overall score of each alternative by Equation (13), then remove criterion j, recalculate scores by finding the absolute error, measure the error caused by removing criterion j by Equation (14), and normalize to get the final weights, Equation (15) [50].

$$S_i=\sum _{j=1}^n{w}_j\cdot n_{ij}$$

(13)

$$E_j=\sum _{i=1}^m\left|S_i-S_{i\left(j\right)}\right|$$

(14)

$$w_j={\displaystyle \frac{E_j}{{\sum }_{j=1}^n{E}_j}}$$

(15)

• (e). CILOS Weight: Compute reciprocal impact for each criterion j using Equation (16) and normalize to determine weights by Equation (17) [51].

$$R_j=\sum _{i=1}^m{\displaystyle \frac{1}{d_{ij}}}$$

(16)

$$w_j={\displaystyle \frac{R_j}{{\sum }_{j=1}^n{R}_j}}$$

(17)

• (f). Bonferroni Operator Fused Weights

Existing objective weighting schemes often emphasize different aspects. For example, entropy tends to prioritize variability-driven measures, CRITIC highlights feature with high correlation and contrast intensity, while MEREC focuses on marginal performance contributors. Such methodological variations can result in divergent rankings when considered independently. The Bonferroni operator addresses this issue by enabling a balanced trade-off, thereby preventing extremely high or low weights from disproportionately influencing the distribution [52].

This integration enhances fairness in multicriteria contexts, particularly where trade-offs exist between economic indicators (e.g., cost per kWh) and sustainability metrics (e.g., carbon footprint). Furthermore, to ensure that the fused weights were not artefacts of the operator, correlation and sensitivity analyses were performed. In both cases, the rankings demonstrated invariance to perturbations, such as weight fluctuations, thereby confirming the robustness of the Bonferroni-based weighting integration approach.

When multiple weights exist for the same criterion from different weight methods, use Equation (18) to find $w_j$ which is a balanced compromise across weight methods.

$$w_j={\displaystyle \frac{1}{m}}\sum _{k=1}^m\left(w_{jk}+{\displaystyle \frac{w_{jk}\cdot w_{jl}}{2}}\right)$$

(18)

where $w_{jk}$ and $w_{jl}$ are weights from two different methods.

• Step 6: Construct weighted normalized matrix (weights by fused with Bonferroni Operator), use Equation (19), calculate overall utility score for each alternative by Equation (20), determine ideal (AI) and anti-ideal scores (AAI) using Equations (21) and Equation (22), respectively, relative utility of each alternative, relative to ideal by Equation (23) and relative to anti-ideal by Equation (24), and then compute utility functions by Equation (25). Rank the alternatives in descending order based on $U_i$.

$$v_{ij}=n_{ij}\cdot w_j$$

(19)

$$S_i=\sum _{j=1}^n{v}_{ij}$$

(20)

$$S_{AI}=\underset{i}{\max }\left(S_i\right)$$

(21)

$$S_{AAI}=\underset{i}{\min }\left(S_i\right)$$

(22)

$$K_i^{+}={\displaystyle \frac{S_i}{S_{AI}}}$$

(23)

$$K_i^{-}={\displaystyle \frac{S_i}{S_{AAI}}}$$

(24)

$$U_i={\displaystyle \frac{K_i^{+}}{K_i^{-}}}$$

(25)

• Step 7: Other MCDM Methods (Applied for Comparison)

ARAS, COCOSO, COPRAS, EDAS, TOPSIS, VIKOR, WASPAS, PROMETHEE-II, and applied with the same Bonferroni weights [53,54,55].

• Step 8: Ranks by different MCDM methods: Correlation module

To measure the agreement between rankings, correlation coefficients are applied. Let $R_{ki}$ and $R_{li}$ be ranking vectors from two MCDM methods k and l. Spearman’s Correlation using Equation (26), and Kendall Rank Correlation using Equation (27), where C: concordant pairs, D: discordant pairs.

$${\mathrm{\rho }}_{kl}=1-{\displaystyle \frac{6{\sum }_{i=1}^n{\left(R_{ki}-R_{li}\right)}^2}{n\left(n^2-1\right)}}$$

(26)

$$\mathrm{\tau }={\displaystyle \frac{C-D}{\frac{1}{2}n\left(n-1\right)}}$$

(27)

• Step 9: Sensitivity Analysis in MCDM (with MARCOS)

It is applied to evaluate how the ranking of alternatives responds to changes in criterion weights, typically to test the stability and robustness of the decision [56].

• Baseline Weight Vector

Let the initial weight vector be:

$$w=\left[w_1,w_2,\dots ,w_n\right]$$

where $w_j$ is the weight of criterion j, and ${\sum }_{j=1}^n{w}_j=1$

• Perturbation of Weight(s)

Perturb one criterion weight $w_k$ by a factor δ, while proportionally adjusting the others, Equations (28)–(30), for all j ≠ k:

$$w_k^{\mathrm{new}}=w_k+\mathrm{\delta }$$

(28)

$$w_j^{\mathrm{new}}=w_j-{\displaystyle \frac{\mathrm{\delta }\cdot w_j}{{\sum }_{j\ne k}{w}_j}}$$

(29)

$$\sum _{j=1}^n{w}_j^{\mathrm{new}}=1$$

(30)

• Recalculate Weighted Matrix

Using the normalized matrix $n_{ij}$, compute the updated weighted matrix, Equation (31).

$$v_{ij}^{\mathrm{new}}=n_{ij}\cdot w_j^{\mathrm{new}}$$

(31)

• Update Scores

Recompute the weighted sums, Equation (32), then compute updated utility functions, Equations (33)–(35).

$$S_i^{\mathrm{new}}=\sum _{j=1}^n{v}_{ij}^{\mathrm{new}}$$

(32)

$$K_i^{+}={\displaystyle \frac{S_i^{\mathrm{new}}}{S_{AI}^{\mathrm{new}}}}$$

(33)

$$K_i^{-}={\displaystyle \frac{S_i^{\mathrm{new}}}{S_{AAI}^{\mathrm{new}}}}$$

(34)

$$U_i^{\mathrm{new}}={\displaystyle \frac{K_i^{+}}{K_i^{-}}}$$

(35)

• Re-rank Alternatives: Sort $U_i^{\mathrm{new}}$ in descending order to obtain the new ranking:

$$\mathrm{Rank}\left(A_i\right)=\mathrm{Position} \mathrm{of} U_i^{\mathrm{new}} \mathrm{in} \mathrm{descending} \mathrm{list}$$

2.2. Multi-Domain Decision Matrix Framework for Component Prioritization in Solar-Based HRESs

A schematic diagram of a solar-based HRESs is presented in Figure 2, illustrating the typical configuration modeled in tools such as HOMER Pro. Within such platforms, the user must select and size the principal system components, including PV modules, batteries, converters, and optional dispatchable sources such as biomass or hydrogen generators. This modularity in system design necessitates informed component-level decisions, as each choice directly influences cost, reliability, and sustainability. While simulation tools are capable of optimizing system performance with respect to technical and economic parameters, there is limited integrated analysis of the trade-offs among competing technologies.

Embedding an MCDM framework at the component selection stage, therefore, adds significant value. It provides structured prioritization of alternatives and design pathways across environmental, technical, and operational dimensions, thereby improving decision quality, reducing design uncertainty, and enhancing alignment of system configurations with broader energy access and sustainability objectives. This approach ensures that component-level decisions support long-term dependability, techno-economic feasibility, and ecological outcomes, objectives of increasing importance in distributed renewable energy planning.

To enable the systematic ranking of critical components within a solar-based HRES, three independent decision matrices were developed and analyzed. These matrices pertain to three essential technological domains: PV technology, battery storage technology, and converter technology. The selection of evaluation criteria for each matrix was guided by considerations of performance characteristics, operational metrics, environmental impact, and lifecycle attributes specific to the corresponding component category [57].

2.2.1. Decision Matrix: PV Technologies for Solar-Based HRESs

The selection and prioritisation of PV technologies in the design of HRESs is of critical importance, as each technology possesses distinct performance, cost, and reliability attributes that directly affect overall system efficiency and long-term sustainability. MCDM techniques are indispensable in this process, as they enable systematic consideration of trade-offs among multiple, and often conflicting, evaluation criteria. MCDM techniques ensure that the prioritisation of PV technologies is aligned with the broader objectives of HRESs, such as minimising environmental impacts while maintaining economic feasibility and technical reliability. Moreover, they reduce the influence of subjectivity in decision-making, which might otherwise occur when designers select PV technologies based solely on personal experience or random choice [58,59]

Table 2 presents the performance characteristics of the PV technologies applied in a solar-based HRES. It has ten criteria that collectively capture environmental, economic, technical, and performance-related aspects. These criteria include carbon footprint (CO₂), energy payback time (EPBT), cost-per-watt (Cst/W), degradation constant (Dc), warranty (War), lifespan (Lfs), efficiency (Eff), temperature coefficient (Tcof), mass (Wgt), and low-irradiance performance (LIP) [60,61]. To preserve methodological consistency, each criterion is classified according to its orientation: cost criteria are assigned a value of −1 (indicators with undesired outcomes where lower values are preferable), while benefit criteria are assigned a value of +1 (indicators with desired outcomes where higher values are preferable). Each criterion in the decision matrix fulfils a distinct role in evaluating PV performance. Environmental sustainability is represented by CO₂ emissions and EPBT; economic and reliability dimensions are captured by Cst/W, degradation rate, and warranty; long-term energy yield is reflected in lifespan and efficiency; resilience under high-heat conditions is measured by the temperature coefficient; ease of installation and structural implications are associated with module weight; and adaptability under shaded or diffuse light conditions is assessed through LIP [62,63]. The combined evaluation of these indicators offers a holistic representation of how each PV technology would perform in real-world HRES applications.

Different PV technologies exhibit unique strengths. Monocrystalline silicon (PV-Mcs) is valued for its high efficiency and extended longevity. In contrast, polycrystalline silicon (PV-Pcs) offers moderate efficiency alongside a favourable cost-to-performance ratio, making it a viable alternative. Emerging technologies such as PERC and TOPCon improve energy capture through enhanced passivation, thereby providing superior performance under low-light conditions. Thin-film technologies (CdTe/CIGS) offer advantages of reduced weight and lower CO₂ footprints, rendering them well-suited for large-scale and weight-sensitive applications. Heterojunction (HJT) and bifacial modules demonstrate exceptional efficiency and adaptability under variable irradiance, positioning them as promising candidates for innovative, high-output solar-based HRES designs [64]. In recent field investigations, silicon heterojunction PV modules have been shown to undergo degradation at an earlier stage, particularly under harsh desert climates where encapsulant aging plays a critical role in determining performance. Compared with PERC and PERT modules, HJT devices exhibited more severe power loss (up to −7.3%), underscoring the importance of encapsulant material selection in ensuring the long-term reliability of advanced PV technologies deployed in extreme hot-climate regions [65].

PV panel alternatives have been selected to encompass a range of well-established and emerging technologies appropriate for solar-based HRESs. With efficiencies approaching 99% and extended operational lifespans, PV-Mcs represent an optimal solution for space-constrained, high-energy applications. PV-Pcs, though slightly less efficient, provide a more cost-effective alternative when spatial limitations are not restrictive and larger installations are feasible. The next-generation passivation UTG-integrated technologies (PV-Prc and PV-Tpc) employ advanced passivation layers, enhancing internal quantum efficiency and thereby improving long-term energy yield. PV-Hjt, characterized by its hybrid crystalline–amorphous structure, demonstrates excellent performance under high-temperature and low-irradiation conditions, while PV-Thf offers lightweight modules with significantly lower embodied carbon. PV-Bfmc, through backside illumination reflection, improves annual energy output in high-albedo environments such as snow-covered landscapes or white-pavement surfaces.

A comprehensive evaluation of these alternatives requires ten criteria encompassing environmental, economic, and operational dimensions, each essential for informed component selection in HRESs. Upstream sustainability metrics are captured through CO₂ emissions and energy payback time. The combined cost–reliability profile of each technology is assessed using cost per watt, durability, and warranty lifespan. Efficiency ensures long-duration energy delivery potential, while temperature coefficient becomes particularly critical in hot climates, where performance stability is paramount. Weight influences structural design requirements, and low-irradiance performance reflects system effectiveness under diffuse light or partial shading.

The diversity of available options emphasizes the necessity of systematic selection methods to achieve near-optimal system design. The integration of MCDM approaches with well-defined performance characteristics enables a structured and robust prioritisation process, thereby maximising both operational performance and sustainability. Ultimately, this ensures that solar-powered HRESs achieve ecological and economic objectives in a balanced and reliable manner.

Empirical studies continue to validate the contextual dependence of PV performance. Duan et al. [66] demonstrated a PV-Trombe wall that simultaneously enhances indoor air quality and heat recovery, highlighting multifunctional integration potential. Wei et al. [67] provided quantitative insights for city-scale rooftop PV deployment, establishing how structural constraints and local irradiance gradients influence optimal module choice. At the forecasting level, Yang et al. [68] developed a spatiotemporal graph-attention model to predict PV cluster power with sub-hour precision, while Zhang et al. [69] statistically optimized battery capacity in PV systems, reinforcing the coupling between PV selection and storage sizing adopted in the present study.

Table 2. Performance characteristics of selected PV technologies used for HRES prioritization.

PV Technology	CO2 (kg)	EPBT (yr)	($) (Cst/W)	(%/yr) Drt	(yr) (War)	(yr) (Lfs)	(%) (Eff)	(%/°C) (Tcof)	Weight (kg/m2)	(LIP)	Ref.
PV-Mcs	600	1.1	1.25	0.5	27.5	27.5	21	0.35	13	3	[70,71,72,73,74,75,76]
PV-Pcs	600	1.1	0.85	0.7	27.5	27.5	16	0.45	13	2	[70,71,72,73,74,75,76]
PV-Prc	600	1.1	0.485	0.4	25	27.5	21	0.35	13	3	[70,71,72,73,74,75,76,77]
PV-Thf	300	0.6	0.6	0.2	22.5	22.5	12.5	0.28	14	4	[71,72,73,74,75,76,78]
PV-Hjt	600	0.94	1.2	0.3	30	27.5	24	0.25	17	5	[70,72,73,74,75,76,79]
PV-Tpc	600	1.1	0.33	0.4	25	25	23	0.3	15	5	[70,71,72,73,74,75,76]
PV-Bfmc	650	1.1	1.35	0.5	30	25	21	0.35	20	5	[71,72,73,74,76]
	−1	−1	−1	−1	1	1	1	−1	−1	1

−1 = cost criterion, +1 = benefit criterion.

Table 2. Performance characteristics of selected PV technologies used for HRES prioritization.

PV Technology	CO2 (kg)	EPBT (yr)	($) (Cst/W)	(%/yr) Drt	(yr) (War)	(yr) (Lfs)	(%) (Eff)	(%/°C) (Tcof)	Weight (kg/m2)	(LIP)	Ref.
PV-Mcs	600	1.1	1.25	0.5	27.5	27.5	21	0.35	13	3	[70,71,72,73,74,75,76]
PV-Pcs	600	1.1	0.85	0.7	27.5	27.5	16	0.45	13	2	[70,71,72,73,74,75,76]
PV-Prc	600	1.1	0.485	0.4	25	27.5	21	0.35	13	3	[70,71,72,73,74,75,76,77]
PV-Thf	300	0.6	0.6	0.2	22.5	22.5	12.5	0.28	14	4	[71,72,73,74,75,76,78]
PV-Hjt	600	0.94	1.2	0.3	30	27.5	24	0.25	17	5	[70,72,73,74,75,76,79]
PV-Tpc	600	1.1	0.33	0.4	25	25	23	0.3	15	5	[70,71,72,73,74,75,76]
PV-Bfmc	650	1.1	1.35	0.5	30	25	21	0.35	20	5	[71,72,73,74,76]
	−1	−1	−1	−1	1	1	1	−1	−1	1

−1 = cost criterion, +1 = benefit criterion.

2.2.2. Decision Matrix: Battery Technologies for Solar-Based HRESs

Energy storage technologies play a pivotal role in solar-based HRESs, contributing not only to grid stability through the fulfilment of energy demand but also to the more effective utilisation of intermittent solar resources. A structured evaluation is therefore essential, as the selection of an appropriate battery has a direct influence on system performance, operational reliability, lifecycle costs, and overall sustainability.

Table 3 presents the decision matrix for batteries, incorporating fourteen attributes categorised across technical, economic, environmental, and operational perspectives. Cycle life represents the number of charge–discharge cycles until significant degradation occurs, while depth of discharge (DoD) indicates the usable proportion of stored energy. Round-trip efficiency quantifies the effectiveness of energy conversion, whereas specific energy (Wh/kg) and volumetric energy density (Wh/L) measure storage capacity relative to weight and volume, respectively. The C-rate reflects the charging and discharging speed. Economic and environmental metrics, including cost per kWh, maintenance requirements, toxicity, and carbon footprint, emphasise cost-effectiveness, maintenance needs, and ecological impact. Long-term reliability and sustainability are further addressed through attributes with system-level implications, such as recyclability or second-life potential, safety and risk, battery management system (BMS) integration capability, and state-of-charge (SoC) retention [80].

Different battery technologies present distinct advantages and limitations. Lithium-ion variants (LFP, NMC, and LTO) provide a favourable balance between efficiency and cycle life, while lead–acid batteries remain attractive due to low capital expenditure despite limited durability. Emerging battery technologies such as sodium-ion and aluminium-ion offer resource abundance and improved environmental performance, whereas flow batteries (e.g., VRFB, zinc–bromine) stand out for their scalability and durability [81]. Next-generation technologies, including solid-state batteries and hybrid supercapacitors, promise enhanced safety and ultra-long endurance. Recent progress has been consolidated on nano-structured composite polymer electrolytes, emphasizing their role in enhancing ionic transport through Li⁺-dynamic interfaces with halloysite nanotubes. The construction of enriched LiF interphases and a >2000% improvement in mechanical toughness are highlighted as key outcomes [82]. A comprehensive overview of cathode binder strategies for solid-state batteries, illustrating their capacity to improve cycle life, mechanical integrity, and thermal safety across LiFePO₄ and transition metal oxide systems. Comparative analysis highlights the distinctive advantages and drawbacks of various binder chemistries in different cathodes [83]. On the solid-state electrolyte front for next-generation lithium-metal batteries, flexible Li_6.5La₃Zr_1.5Ta_0.5O₁₂ (LLZTO) thin films, inspired by eggshell membrane design, have demonstrated outstanding mechanical resilience, cycling stability up to 300 h, and high Coulombic efficiency (~99.7%). Such compressible garnet electrolytes provide a promising pathway toward the realization of high-energy-density thin-film batteries with enhanced safety and long-term cycling performance [84].

The integration of MCDM techniques with these criteria enables a systematic and transparent pre-selection process, facilitating the identification of the most appropriate battery technology for solar-driven HRES applications.

The selection of battery alternatives comprises a comprehensive portfolio of established and emerging chemistries suitable for both short-duration and long-duration energy storage in HRESs. Lithium iron phosphate batteries are widely adopted due to their long cycle life, high safety profile, and moderate cost, rendering them a balanced option. Nickel manganese cobalt batteries provide high energy density and efficient operation, albeit with a comparatively shorter lifespan. Sustainable material approaches are emerging in lithium-ion battery development. A study synthesized porous carbon–silicon composites from low-cost fly ash cenospheres via magnesiothermic reduction, achieving high reversible capacity (526 mAh·g⁻¹) and stable cycling. This work highlights the potential of waste-derived silicon materials as cost-effective, high-performance anodes for next-generation Li-ion batteries [85].

Lead–acid batteries remain cost-effective but are constrained by shallow depth of discharge and limited cycle life. By contrast, sodium-ion and aluminum-ion batteries represent promising next-generation chemistries, offering resource abundance, enhanced safety, and competitive energy densities. Advanced chemistries such as lithium titanate and hybrid supercapacitors provide exceptional cycle life and rapid charging capabilities, making them suitable for applications requiring high-frequency cycling. Flow-based technologies such as vanadium redox flow batteries and zinc–bromine batteries, while comparatively bulkier, are highly suited to stationary HRES installations where extreme durability and modular scalability are prioritized. The inclusion of second-life electric vehicle batteries further supports circular economy objectives by repurposing end-of-life EV cells into stationary storage applications.

The criteria cycle life and depth of discharge determine long-term durability and usable capacity. Round-trip efficiency, specific energy, and volumetric energy density capture essential efficiency and density parameters for system sizing. The C-rate indicates compatibility with dynamic load conditions. Economic viability is reflected in cost per kilowatt-hour and maintenance requirements. Environmental dimensions are assessed through toxicity and carbon footprint. Additionally, recyclability, safety, battery management systems, and real-time state-of-charge monitoring provide critical insights into maturity, integration, and operational safety.

Collectively, these criteria form a robust foundation for evaluating battery suitability under diverse use-case scenarios. Embedding these indicators within an MCDM framework ensures that selection strategies prioritize lifecycle performance, safety, and sustainability over cost or availability alone, thereby aligning storage system design with the overarching objectives of HRES deployment.

Beyond electrochemical storage, complementary approaches have emerged to enhance hybrid system flexibility. Li F. et al. [86] assessed the feasibility of gravity-based energy storage systems as sustainable complements to battery banks. Niu et al. [87] analyzed supercritical CO₂ Brayton cycles for solar tower applications, offering thermodynamic insights relevant to high-temperature energy coupling in advanced HRESs. Meanwhile, Meng et al. [88] optimized residential PV microgrids using adaptive storage control, confirming that integrated management of battery charge dynamics markedly improves system efficiency and reliability.

Table 3. Performance characteristics of selected Battery technologies used for HRES prioritization.

Battery Type	CL (No. of Cycles)	DoD (%)	RTE (%)	SE (Wh/kg)	VED (Wh/L)	C-Rate	CpkWh ($/kWh)	Maint (0–5 Scale)	Tox (0–5 Scale)	CF (kg CO2/kWh)	Recycle (0–5 Scale)	Safety (0–5 Scale)	BMS (0–5 Scale)	SoC_RT (%)	Ref.
LFP	6000	100	95	160	333	1.5	110	5	5	60	3	4	5	98	[89,90,91,92]
LA	500	50	80	40	75	0.2	100	2	1	150	5	4	2	80	[89,90,92]
NMC	2000	80	90	220	580	1	140	5	3	100	4	3	5	96	[89,90]
Na-Ion	1500	100	90	150	275	3	90	5	5	80	3	4	4	95	[89,90,93]
2ndEV	1200	80	85	120	400	1	70	5	3	40	4	3	3	90	[89,90,92,93]
VRFB	15,000	100	75	25	30	0.25	350	4	2	180	5	5	3	100	[89,90,93]
Zn-Br	5000	100	75	85	65	0.25	400	3	1	180	5	5	2	100	[89,90,92,93]
NiFe	2500	100	65	25	125	0.5	500	1	3	160	4	4	1	90	[89,90,93]
NiZn	800	80	85	100	280	2	450	5	4	16	5	5	3	92	[89]
Zn-Air	1000	100	60	300	1000	0.25	250	3	5	60	5	5	2	100	[89,90,92,93]
Al-Ion	2000	100	90	400	450	5	200	5	5	60	5	5	4	98	[89,90,93]
LTO	10,000	100	90	100	177	10	600	5	3	150	3	5	5	98	[89,90,92]
HSC	100,000	100	95	30	20	100	1000	5	5	70	3	5	5	100	[89,90,93,94]
SSB	5000	100	90	350	500	4	200	5	3	100	4	5	5	98	[89,90,92]
	1	1	1	1	1	1	−1	−1	−1	−1	1	1	1	1

−1 = cost criterion, +1 = benefit criterion.

Table 3. Performance characteristics of selected Battery technologies used for HRES prioritization.

Battery Type	CL (No. of Cycles)	DoD (%)	RTE (%)	SE (Wh/kg)	VED (Wh/L)	C-Rate	CpkWh ($/kWh)	Maint (0–5 Scale)	Tox (0–5 Scale)	CF (kg CO2/kWh)	Recycle (0–5 Scale)	Safety (0–5 Scale)	BMS (0–5 Scale)	SoC_RT (%)	Ref.
LFP	6000	100	95	160	333	1.5	110	5	5	60	3	4	5	98	[89,90,91,92]
LA	500	50	80	40	75	0.2	100	2	1	150	5	4	2	80	[89,90,92]
NMC	2000	80	90	220	580	1	140	5	3	100	4	3	5	96	[89,90]
Na-Ion	1500	100	90	150	275	3	90	5	5	80	3	4	4	95	[89,90,93]
2ndEV	1200	80	85	120	400	1	70	5	3	40	4	3	3	90	[89,90,92,93]
VRFB	15,000	100	75	25	30	0.25	350	4	2	180	5	5	3	100	[89,90,93]
Zn-Br	5000	100	75	85	65	0.25	400	3	1	180	5	5	2	100	[89,90,92,93]
NiFe	2500	100	65	25	125	0.5	500	1	3	160	4	4	1	90	[89,90,93]
NiZn	800	80	85	100	280	2	450	5	4	16	5	5	3	92	[89]
Zn-Air	1000	100	60	300	1000	0.25	250	3	5	60	5	5	2	100	[89,90,92,93]
Al-Ion	2000	100	90	400	450	5	200	5	5	60	5	5	4	98	[89,90,93]
LTO	10,000	100	90	100	177	10	600	5	3	150	3	5	5	98	[89,90,92]
HSC	100,000	100	95	30	20	100	1000	5	5	70	3	5	5	100	[89,90,93,94]
SSB	5000	100	90	350	500	4	200	5	3	100	4	5	5	98	[89,90,92]
	1	1	1	1	1	1	−1	−1	−1	−1	1	1	1	1

−1 = cost criterion, +1 = benefit criterion.

2.2.3. Decision Matrix: Converter Technologies for Solar-Based HRESs

In solar-powered HRESs, converters function as the critical power interface, enabling the stable conversion, conditioning, and delivery of electrical energy from PV modules and storage units to the grid or load. A comparative assessment of various energy conversion systems is an indispensable stage in the design of HRESs, as the choice of converter significantly influences the efficiency, reliability, economic feasibility, and safety of the overall system.

Table 4 depicts converters across twelve attributes spanning technical, economic, environmental, and operational dimensions. Energy transfer characteristics are primarily measured through conversion efficiency and power factor, where the conversion efficiency indicates the effectiveness of energy transfer and the power factor ensures the quality of power supplied to the grid. Waveform quality is assessed via total harmonic distortion (THD), while reliability and durability are quantified by the MTBF. Economic and sustainability-related indicators include cost per kW, maintenance requirements, toxicity or environmental hazards, and carbon footprint. At the system level, recyclability, safety and risk, grid compatibility, and redundancy measure ease of integration, operational resilience, and long-term sustainability [95].

Different types of converters exhibit distinct advantages. String inverters are generally simpler and more cost-effective in large quantities, whereas microinverters enhance reliability at the module level. Power optimizer–central inverter hybrids maximise power point tracking. Modular Multilevel Converters (MMC) and Neutral-Point Clamped (NPC) converters provide high efficiency, scalability, and improved grid stability, while bidirectional DC–DC converters enable flexible energy flow for storage applications. The transformer-coupled cascaded six-leg bridge converter offers modularity, inherent isolation, and simplified control over MMCs and CHBs, maintaining comparable voltage, power, and harmonic performance for efficient medium- and high-voltage renewable integration [96]. The resonance behavior of MMC-HVDC systems in interaction with AC networks has also been extensively investigated [97].

Conversely, flyback converters, though inexpensive and suitable for small-scale systems, suffer from comparatively low reliability [98]. Water-PV systems optimized through PSO-based floating PV reconfiguration and integrated with multi-level converters enhance power yield, cut mismatch losses by 47%, and reduce water evaporation by 75%, supporting scalable energy–water sustainability [99].

Converter alternatives have been selected to capture the diversity of power electronic architectures employed in solar-based HRESs, encompassing centralized systems as well as highly modular and bidirectional platforms. Centralized converters are cost-effective and readily scalable for large-scale installations, though they exhibit limited fault tolerance. A novel chain-loop iteration-based fault analysis method for MMC grid-forming HVDC systems improves convergence and accuracy, enhancing reliability and protection in renewable-dominated grids [100].

Module converters provide enhanced reliability at the module level along with maximum power point tracking control, thereby improving system granularity and safety. Power optimizers and string converters offer a balance between cost-effectiveness and energy optimization by incorporating string-level optimization. Modular multilevel converters and neutral-point clamped inverters exemplify multilevel inverter architectures that achieve high efficiency, strong fault tolerance, and grid-code compliance, rendering them suitable for critical-load applications. Bidirectional converters facilitate two-way energy flow between PV arrays, batteries, and load centers an increasingly essential capability in hybrid system configurations. Flyback converters, while compact and inexpensive, are more appropriate for low-capacity applications given their limited efficiency and shorter lifespan.

The twelve evaluation criteria encompass dimensions of efficiency, quality, reliability, and sustainability. Conversion efficiency and power factor assess fundamental performance and utilization. Total harmonic distortion quantifies waveform purity, while mean time between failures reflects operational reliability and uptime expectations. Cost per kilowatt, maintenance requirements, and toxicity contribute to lifecycle cost analysis and environmental risk assessment. Carbon footprint and recyclability capture sustainability and end-of-life attributes. Finally, safety, grid compliance, and redundancy (Redund) measure compliance, resilience, and integration into broader energy networks.

By applying MCDM techniques, these diverse converter options can be systematically prioritised, enabling an evidence-based evaluation of trade-offs across performance, safety, cost, and sustainability. This structured approach supports the optimal selection of converters for solar-powered components within HRESs.

Table 4. Performance characteristics of selected Converter technologies used for HRES prioritization.

	ConvEff (%)	PF	THD (%)	MTBF (h)	CpkW ($/kW)	Maint (0–5 Scale)	Tox (0–5 Scale)	CF (kg CO2/kW)	Recycle (0–5 Scale)	Safety (0–5 Scale)	GridComp (0–5 Scale)	Redund (0–5 Scale)	Ref.
SC	97	0.99	3	80,000	120	3	4	35	3	4	5	3	[74,91,101,102,103]
MC	96	0.98	2	90,000	160	2	4	30	4	5	5	5	[74,101,102,103]
PC	98	0.99	2.5	85,000	150	2	4	32	3	5	5	4	[74,101,102,103]
MMC	98.5	0.99	2	95,000	200	3	5	40	4	5	5	5	[74,101,102,103]
BDC	97.5	0.98	2.5	92,000	180	3	4	38	4	5	5	4	[74,101,102,103]
NPC	98	0.99	2.2	94,000	210	4	5	42	4	5	5	5	[74,101,102,103,104]
FC	95	0.97	3.5	75,000	90	2	3	28	3	4	3	2	[74,101,102,103]
	1	1	−1	1	−1	−1	−1	−1	1	1	1	1

−1 = cost criterion, +1 = benefit criterion.

Table 4. Performance characteristics of selected Converter technologies used for HRES prioritization.

	ConvEff (%)	PF	THD (%)	MTBF (h)	CpkW ($/kW)	Maint (0–5 Scale)	Tox (0–5 Scale)	CF (kg CO2/kW)	Recycle (0–5 Scale)	Safety (0–5 Scale)	GridComp (0–5 Scale)	Redund (0–5 Scale)	Ref.
SC	97	0.99	3	80,000	120	3	4	35	3	4	5	3	[74,91,101,102,103]
MC	96	0.98	2	90,000	160	2	4	30	4	5	5	5	[74,101,102,103]
PC	98	0.99	2.5	85,000	150	2	4	32	3	5	5	4	[74,101,102,103]
MMC	98.5	0.99	2	95,000	200	3	5	40	4	5	5	5	[74,101,102,103]
BDC	97.5	0.98	2.5	92,000	180	3	4	38	4	5	5	4	[74,101,102,103]
NPC	98	0.99	2.2	94,000	210	4	5	42	4	5	5	5	[74,101,102,103,104]
FC	95	0.97	3.5	75,000	90	2	3	28	3	4	3	2	[74,101,102,103]
	1	1	−1	1	−1	−1	−1	−1	1	1	1	1

−1 = cost criterion, +1 = benefit criterion.

2.3. Data Collection and Validation

Data for the decision matrices about PV modules, battery storage systems, and power converters were obtained from peer-reviewed literature (2020–2024), manufacturer technical datasheets, and validated experimental reports. The objective was to compile a comprehensive, real-time dataset that accurately reflects the current technological performance trends and cost metrics relevant to solar HRES technologies.

A comprehensive screening of both published and industrial sources was conducted for each technology category to identify data corresponding to the technical, environmental, economic, and reliability criteria used in the MCDM framework. In cases where multiple credible sources reported similar values, for example, efficiency, degradation rate, or LCOE, average or representative mid-range values were adopted to ensure internal consistency. The PV dataset integrates recent field and laboratory data for crystalline silicon, heterojunction, passivated-emitter-rear-cell, tunnel-oxide-passivated-contact, and thin-film technologies. Drawing upon field performance data, industry assessments, and life-cycle analysis databases, parameters were compiled for efficiency, energy payback time, degradation rate, temperature coefficient, and embodied carbon. These datasets represent actual module behavior under varied meteorological and operational conditions, providing a realistic basis for comparative evaluation.

For battery technologies, data were collected from manufacturer specifications (e.g., Panasonic NCR, CATL, Maxwell) and recent scientific literature encompassing lithium-ion, lithium-iron-phosphate, sodium-ion, solid-state battery, flow battery, and hybrid supercapacitor chemistries. For cross-comparative analysis, all data were normalized to standard test conditions (25 °C, rated C-rate), including cycle life, depth of discharge, efficiency, self-discharge rate, and energy density.

The converter dataset comprises specifications and performance data for multilevel inverter and modular multilevel converter topologies, sourced from simulation-based experimental studies and manufacturer documentation. Indicators were selected based on their direct influence on hybrid system operation, including total harmonic distortion, power factor, efficiency, voltage stress, and reliability metrics.

This rigorous data collection and validation protocol ensures that all input criteria are traceable, credible, and representative of contemporary technology performance. Consequently, the datasets used in this study reinforce the transparency, reproducibility, and reliability of the MARCOS-based MCDM evaluation presented herein.

2.4. Operational Context and System Specifications for Component Evaluation

Effective HRES component selection should be grounded in precise definitions of subsequent system specifications. Across applications such as cold storage, rural microgrids, agro-processing, and institutional loads, these systems are typically deployed within a capacity range of 50–500 kW. They may operate in on-grid, standalone, or hybrid modes, supporting daily energy demands of 300–2000 kWh and accommodating load profiles characterized by steady base loads combined with variable peak demands.

Component performance is significantly influenced by environmental factors, including solar irradiance (4.5–6.5 kWh/m²/day) and ambient temperature (5–45 °C). PV modules must sustain efficiency under elevated temperatures; batteries must demonstrate stable depth of discharge, long cycle life, and robustness to temperature fluctuations; and converters must achieve low harmonic distortion, high conversion efficiency, and compliance with grid requirements.

These contextual specifications ensure that the MCDM framework remains grounded in practical deployment considerations. Selecting components in isolation, without regard to the operational environment, risks suboptimal outcomes. By embedding system-level assumptions, the MCDM model enables defensible and application-specific prioritization across technical, economic, and sustainability dimensions.

3. Results and Discussion

The decision matrix, as per Equation (1), is depicted in Table 2, Table 3 and Table 4 for PV, battery, and converter technologies for solar-based HRESs.

3.1. Objective Weights Computation

Multiple weighting methods were applied to the normalized decision matrices corresponding to each component in order to assign weights and rank PV, battery, and converter technology alternatives for solar-based HRESs as a whole. Criterion weights derived from five objective methods Entropy (Equations (5)–(7)), Standard Deviation (Equation (8) to (9)), CRITIC (Equations (10)–(12)), MEREC (Equations (13)–(15)), and CILOS (Equations (16) and (17)) are presented in

Table 5. Normalized criteria weights obtained using five objective methods and Bonferroni fusion for ranking PV, battery, and converter technologies in solar-based HRESs.

PV Technologies for Solar-Based HRESs
	CILOS	ENTROPY	MEREC	CRITIC	STD_DEV	Bonferroni
CO2	0.1614	0.1548	0.1064	0.0889	0.0921	0.1223
EPBT	0.1399	0.3614	0.0678	0.0943	0.1030	0.1469
Cst/W	0.1911	0.0979	0.1983	0.1106	0.1091	0.1428
Drt	0.1725	0.0544	0.1327	0.0684	0.0876	0.1030
War	0.0398	0.0627	0.0904	0.1236	0.1013	0.0840
Lfs	0.0193	0.0514	0.1039	0.1296	0.1074	0.0815
Eff	0.0653	0.0532	0.1017	0.0996	0.0969	0.0846
Tcof	0.0837	0.0506	0.0410	0.0683	0.0887	0.0673
Wgt	0.0420	0.0492	0.0797	0.1163	0.1032	0.0784
LIP	0.0850	0.0644	0.0782	0.1004	0.1106	0.0893
Battery Technologies for Solar-based HRESs
CL	0.2943	0.9712	0.0568	0.3461	0.0714	0.3432
DoD	0.0031	0.0006	0.0607	0.0001	0.0714	0.0247
RTE	0.0019	0.0004	0.0680	0.0006	0.0714	0.0258
SE	0.0709	0.0046	0.0637	0.0228	0.0714	0.0501
VED	0.0807	0.0101	0.0575	0.0853	0.0714	0.0668
C-rate	0.3883	0.0010	0.0570	0.3551	0.0714	0.1733
CpkWh	0.0636	0.0097	0.0687	0.0412	0.0714	0.0556
Maint	0.0141	0.0001	0.0918	0.0512	0.0714	0.0476
Tox	0.0219	0.0001	0.0980	0.0372	0.0714	0.0474
CF	0.0335	0.0020	0.0706	0.0547	0.0714	0.0500
Recycle	0.0049	0.0000	0.1032	0.0037	0.0714	0.0335
Safety	0.0033	0.0000	0.0781	0.0006	0.0714	0.0279
BMS	0.0190	0.0001	0.0714	0.0011	0.0714	0.0318
SoC_RT	0.0004	0.0002	0.0544	0.0003	0.0714	0.0224
Converter Technologies for Solar-based HRESs
ConvEff	0.0033	0.0004	0.0068	0.0692	0.0753	0.0275
PF	0.3269	0.0001	0.0028	0.0792	0.0830	0.0827
THD	0.1321	0.0942	0.0702	0.0602	0.0778	0.0910
MTBF	0.0000	0.0178	0.0328	0.0666	0.0798	0.0386
CpkW	0.0022	0.2343	0.3567	0.1110	0.0756	0.1511
Maint	0.1264	0.1669	0.1031	0.0938	0.0798	0.1195
Tox	0.0797	0.0728	0.1684	0.1035	0.0728	0.1035
CF	0.0094	0.0541	0.0901	0.1026	0.0793	0.0688
Recycle	0.0919	0.0531	0.0441	0.1027	0.1128	0.0844
Safety	0.0689	0.0258	0.0138	0.0811	0.1030	0.0593
GridComp	0.0705	0.0674	0.0187	0.0660	0.0798	0.0630
Redund	0.0889	0.2131	0.0923	0.0642	0.0812	0.1107

. Each method was applied independently to the individual sets of criteria (PV, battery, and converter technologies), and the results were subsequently combined using the Bonferroni aggregation operator (Equation (18)) to ensure robustness and consensus.

The methods assign importance to criteria under varying mathematical assumptions:

• CILOS: Focuses on inverse values, primarily low-performing metrics, emphasizing performance from the perspective of loss.
• Entropy: Captures the degree of disorder within the criterion values, with greater variability indicating higher importance.
• MEREC: Evaluates the direct contribution of each criterion to the overall performance.
• CRITIC: Incorporates both the standard deviation of each criterion and the degree of correlation among criteria.
• STD_DEV: A dispersion-based method that assigns greater weight to more variable criteria.

To mitigate bias and differences among weighting schemes, the Bonferroni operator was employed as a fusion strategy, Equation (18). Relying solely on direct weights is often inadequate, as individual weighting methods may emphasize only direct importance or include interaction effects when paired with other methods. The Bonferroni operator balances direct and interactive effects, ensuring that the final weights accurately represent the collective input from all methods.

As shown in Table 5, most methods assign higher weights to PV technology criteria (first section). In particular, CO₂ emissions, energy payback time, and cost per watt consistently receive relatively greater importance, indicating their decisive role in PV technology selection. Similarly, for batteries, cycle life and C-rate achieve the highest weights under CILOS and CRITIC, respectively, highlighting their technical significance. For converter technologies, parameters such as operational voltage range and capacity per kilowatt are consistently identified as critical across methods.

The Pareto-based distribution of the Bonferroni-weighted criteria for technology selection in solar-based HRESs is presented in Figure 3. Figure 3a highlights the dominant influence of energy payback time and cost per watt on PV technology selection, while Figure 3b illustrates the substantial contributions of cycle life and C-rate to battery prioritization. Figure 3c presents the critical parameters for converter technologies, namely measurement redundancy, operational voltage range, and cost per kilowatt. These visualizations collectively facilitate a clearer understanding of the parameters exerting the greatest impact on HRES components, thereby supporting more informed decision-making.

The resulting Bonferroni weights are subsequently applied within the MARCOS method (and other MCDM techniques for comparison) to rank alternative technologies. This enables the systematic selection of the most suitable technology under multi-criteria conditions.

3.2. Prioritizing Benchmark by MARCOS: PV, Battery, and Converter Technologies for HRESs

The MARCOS approach was employed to evaluate solar-based HRES technologies, encompassing PV, battery, and converter alternatives. Following the normalization of decision matrices (Equations (2) and (3)) and the incorporation of ideal and anti-ideal reference points (Equation (4)), weighted normalized values were obtained using Bonferroni-fused weights (Equation (18)), refer to Table 5, weight values used. Subsequently, the ideal and anti-ideal values (Equations (21) and (22)) were compared with the aggregated scores of each alternative (Equation (20)), yielding relative utilities (Equations (23) and (24)) and final utility scores (Equation (25)).

The ranking outcomes highlight distinct technology leaders across domains. Among PV technologies, thin-film performed best with a utility score of 0.7108, closely followed by tunnel oxide passivated contact (TOPCon; PV-Tpc) and heterojunction (PV-Hjt) modules, which demonstrated strong trade-offs between high efficiency, long durability, and favorable carbon performance. In terms of energy storage, hybrid supercapacitors (HSC; 0.6990) significantly outperformed other options, including solid-state and aluminum-ion batteries, owing to their exceptional cycle life and intrinsic safety characteristics. For converters, modular converters achieved the highest performance (1.1812), with power converters and flyback converters following closely, reflecting their robustness and high compatibility with grid integration requirements. The findings are consolidated in

Table 6. MARCOS-based prioritization results for PV, battery, and converter technologies in solar-based HRESs.

PV Technologies for Solar-based HRESs
PV Technology	Sum of Weighted Values	Utility vs. Ideal	Utility vs. Anti-Ideal	Final Utility Score	Rank
PV-Mcs	0.6327	0.6327	1.2687	0.5426	6
PV-Pcs	0.5925	0.5925	1.1881	0.5082	7
PV-Prc	0.6954	0.6954	1.3946	0.5965	4
PV-Thf	0.8287	0.8287	1.6619	0.7108	1
PV-Hjt	0.7294	0.7294	1.4628	0.6256	3
PV-Tpc	0.7740	0.7740	1.5521	0.6638	2
PV-Bfmc	0.6331	0.6331	1.2695	0.5430	5
Battery Technologies for Solar-based HRESs
HSC	0.7085	0.7085	5.7318	0.6990	1
SSB	0.3116	0.3116	2.5208	0.3074	2
Al-Ion	0.3053	0.3053	2.4699	0.3012	3
Zn-Air	0.2999	0.2999	2.4263	0.2959	4
LFP	0.2798	0.2798	2.2636	0.2761	5
NMC	0.2770	0.2770	2.2407	0.2733	6
2ndEV	0.2741	0.2741	2.2176	0.2705	7
LTO	0.2640	0.2640	2.1354	0.2604	8
NiZn	0.2614	0.2614	2.1143	0.2579	9
Na-Ion	0.2580	0.2580	2.0871	0.2545	10
VRFB	0.2561	0.2561	2.0720	0.2527	11
Zn-Br	0.2516	0.2516	2.0353	0.2482	12
LA	0.2481	0.2481	2.0067	0.2447	13
NiFe	0.2151	0.2151	1.7398	0.2122	14
Converter Technologies for Solar-based HRESs
SC	0.7687	0.7687	1.2494	1.009	7
MC	0.8999	0.8999	1.4626	1.1812	1
PC	0.8395	0.8395	1.3644	1.1019	2
MMC	0.815	0.815	1.3247	1.0698	4
BDC	0.798	0.798	1.297	1.0475	5
NPC	0.7807	0.7807	1.269	1.0249	6
FC	0.8257	0.8257	1.3421	1.0839	3

, which establishes clear technology benchmarks for the sustainable and reliable design of solar-based HRESs.

In the next step, the study extends the analysis by verifying the consistency of the MARCOS-based ranking with those derived from alternative MCDM methods and conducting sensitivity and correlation tests to ensure robustness. Additionally, the results are contextualized within broader sustainability and techno-economic perspectives to elucidate their practical and policy implications. The extended, normalized, and weighted matrices for PV, battery, and converter technologies using Bonferroni-fused weights are provided in Supplementary Tables S1–S3.

3.3. Comparative Rankings and Correlation Analysis of Technologies

For comparing PV, battery, and converter technologies across nine MCDM techniques, MARCOS, ARAS, COCOSO, COPRAS, EDAS, TOPSIS, VIKOR, WASPAS, and PROMETHEE II, the robustness of the prioritization results was validated. As illustrated in Figure 4 PV-Thf, HSCs, and MCs consistently performed above average within their respective categories, whereas PV-Pcs, lead–acid converters, and SCs performed significantly below average. This stability across MCDM methodologies accentuates the reliability of both the best and worst performers.

The emphasis on reliability in converter and storage ranking aligns with recent resilience analyses. Xiahou et al. [105] introduced a quantitative resilience assessment framework for hybrid AC/DC cyber-physical power systems, demonstrating that cascading-failure modeling is essential to ensure robust operation under converter or communication disturbances parameters directly considered in our MCDM reliability criteria.

To further assess consistency, Kendall’s rank correlation and Spearman’s rank correlation were calculated for each technology domain, as shown in Equations (26) and (27). As presented in

Table 7. Spearman and Kendall correlation matrices for PV, battery, and converter technologies across MCDM methods.

PV Technologies: Spearman Correlation Matrix
Method	MARCOS	ARAS	COCOSO	COPRAS	EDAS	TOPSIS	VIKOR	WASPAS	PROMETHEE II
MARCOS	1	1.0000	0.8571	0.8571	0.9286	0.8929	−0.1429	0.9643	0.7857
ARAS		1	0.8571	0.8571	0.9286	0.8929	−0.1429	0.9643	0.7857
COCOSO			1	0.7500	0.7857	0.8571	0.1429	0.8929	0.9643
COPRAS				1	0.9643	0.9643	−0.4643	0.9286	0.7857
EDAS					1	0.9286	−0.3214	0.9643	0.7500
TOPSIS						1	−0.3214	0.9643	0.8929
VIKOR							1	−0.1786	0.0000
WASPAS								1	0.8571
PROMETHEE II									1
PV Technologies: Kendall Rank Correlation Matrix
Method	MARCOS	ARAS	COCOSO	COPRAS	EDAS	TOPSIS	VIKOR	WASPAS	PROMETHEE II
MARCOS	1
ARAS	1.0000	1
COCOSO	0.7143	0.7143	1
COPRAS	0.7143	0.7143	0.6190	1
EDAS	0.8095	0.8095	0.7143	0.9048	1
TOPSIS	0.8095	0.8095	0.7143	0.9048	0.8095	1
VIKOR	−0.0476	−0.0476	0.0476	−0.3333	−0.2381	−0.2381	1
WASPAS	0.9048	0.9048	0.8095	0.8095	0.9048	0.9048	−0.1429	1
PROMETHEE II	0.6190	0.6190	0.9048	0.7143	0.6190	0.8095	−0.0476	0.7143	1
Battery Technologies: Spearman Correlation Matrix
Method	MARCOS	ARAS	COCOSO	COPRAS	EDAS	TOPSIS	VIKOR	WASPAS	PROMETHEE II
MARCOS	1	0.8198	0.5956	0.6747	0.6659	0.7495	0.6308	0.8857	0.6220
ARAS		1	0.7011	0.9121	0.8945	0.7055	0.8989	0.9429	0.7934
COCOSO			1	0.5121	0.5956	0.4813	0.6703	0.6791	0.6527
COPRAS				1	0.8637	0.6615	0.8022	0.8374	0.7582
EDAS					1	0.5209	0.9209	0.9121	0.9385
TOPSIS						1	0.6835	0.6747	0.4989
VIKOR							1	0.8549	0.8593
WASPAS								1	0.8593
PROMETHEE II									1
Battery Technologies: Kendall Rank Correlation Matrix
Method	MARCOS	ARAS	COCOSO	COPRAS	EDAS	TOPSIS	VIKOR	WASPAS	PROMETHEE II
MARCOS	1
ARAS	0.7363	1
COCOSO	0.4286	0.5165	1
COPRAS	0.4945	0.7582	0.3626	1
EDAS	0.5165	0.7363	0.4286	0.7143	1
TOPSIS	0.6264	0.5824	0.3187	0.5165	0.4066	1
VIKOR	0.4945	0.7143	0.4945	0.6923	0.8022	0.5165	1
WASPAS	0.7582	0.8462	0.5385	0.6923	0.7582	0.5165	0.6923	1
PROMETHEE II	0.4725	0.6484	0.4725	0.5824	0.8242	0.4066	0.7143	0.7143	1
Converter Technologies: Spearman Correlation Matrix
Method	MARCOS	ARAS	COCOSO	COPRAS	EDAS	TOPSIS	VIKOR	WASPAS	PROMETHEE II
MARCOS	1	0.9640	0.9640	−0.7500	0.9640	0.8930	−0.2860	0.9640	0.8930
ARAS		1	0.9290	−0.6430	0.9290	0.8570	−0.2140	0.9290	0.8570
COCOSO			1	−0.8210	1.0000	0.8930	−0.3570	1.0000	0.8210
COPRAS				1	−0.8210	−0.7860	0.7140	−0.8210	−0.5710
EDAS					1	0.8930	−0.3570	1.0000	0.8210
TOPSIS						1	−0.5000	0.8930	0.7500
VIKOR							1	−0.3570	−0.0710
WASPAS								1	0.8210
PROM-II									1
Converter Technologies: Kendall Rank Correlation Matrix
Method	MARCOS	ARAS	COCOSO	COPRAS	EDAS	TOPSIS	VIKOR	WASPAS	PROMETHEE II
MARCOS	1
ARAS	0.8570	1
COCOSO	0.9050	0.8100	1
COPRAS	−0.6670	−0.5240	−0.7620	1
EDAS	0.9050	0.8100	1.0000	−0.7620	1
TOPSIS	0.8100	0.7140	0.8100	−0.7140	0.8100	1
VIKOR	−0.1430	−0.0950	−0.1900	0.7140	−0.1900	−0.2860	1
WASPAS	0.9050	0.8100	1.0000	−0.7620	1.0000	0.8100	−0.1900	1
PROM-II	0.7620	0.6670	0.6670	−0.4290	0.6670	0.5710	0.0480	0.6670	1

, the results indicate strong agreement among most methods, particularly for PV and converter technologies. For PV, MARCOS, ARAS, EDAS, and WASPAS demonstrated strong positive correlations (ρ > 0.90), while VIKOR diverged, showing overall negative correlations with others. Battery technologies exhibited moderate consensus (average ρ ≈ 0.74, Kendall’s W ≈ 0.69), with MARCOS, ARAS, and WASPAS forming a cluster, whereas COCOSO and TOPSIS showed weaker alignment. The highest concordance was observed for converter technologies, where MARCOS, ARAS, COCOSO, EDAS, and WASPAS clustered strongly (ρ ≈ 0.90–0.96), in contrast with COPRAS and VIKOR, which exhibited negative concordance.

Overall, the comparative rankings and correlation-based analyses confirm that the leading technologies (PV-Thf, HSCs, and MCs) consistently demonstrate superior performance across diverse decision-making approaches, thereby strengthening confidence in their selection for sustainable solar-based HRES design.

This comparative validation further highlights the superior stability of MARCOS. In contrast to single-reference methods such as TOPSIS or VIKOR, which exhibited rank reversals under correlated cases, MARCOS consistently produced stable outcomes by incorporating both ideal and anti-ideal solutions. This strength was also substantiated through correlation analysis, where MARCOS demonstrated strong associations with nearly all MCDM methods, confirming its suitability as a reliable ranking tool.

By integrating all evaluation criteria, the best-performing alternatives are identified as PV–Thf (thin-film), HSC (hybrid supercapacitor), and MC (module converter). These bottom-up subcomponents exhibit a strong alignment with the techno-economic and sustainability requirements of solar-based HRESs, particularly in mid-scale applications.

4. Sensitivity Analysis

In line with Step 9 of the proposed methodology, a sensitivity analysis was conducted based on perturbations of the criterion weights, with particular attention to the stability of the MARCOS-based prioritization. The baseline weight vector $w=\left[w_1,w_2,\dots ,w_n\right],\sum w_j=1$ was perturbed one criterion at a time, ensuring that the total weight remained normalized. Following each perturbation, the weighted normalized decision matrix was recalculated (Equation (31)), updated utility scores were obtained (Equation (32)), and final utility functions were derived (Equations (33)–(35)). Subsequently, the alternatives were re-ranked to assess the robustness of the prioritization under shifting preferences.

Perturbations of $\delta =\pm 5\mathrm{\%},\pm 10\mathrm{\%},\pm 15\mathrm{\%}$ were applied individually to key parameters within each main category: degradation rate, CO₂ emissions, and efficiency for PV technologies; cycle life, specific energy, and cost per kWh for batteries; and cost per kW, maintenance, and redundancy for converters. This yielded 18 scenarios per technology domain, as summarized in

Table 8. Sensitivity analysis of PV, battery, and converter technologies under criteria weight perturbations using the MARCOS method.

PV Technologies
Scenario	Target	Delta	CO2	EPBT	Cst/W	Drt	War	Lfs	Eff	Tcof	Wgt	LIP
Sc01	S-4 (Drt)	−0.15	0.1244	0.1494	0.1452	0.0875	0.0854	0.0829	0.0860	0.0685	0.0797	0.0908
Sc02	S-4 (Drt)	−0.1	0.1237	0.1486	0.1444	0.0927	0.0850	0.0824	0.0856	0.0681	0.0793	0.0903
Sc03	S-4 (Drt)	−0.05	0.1230	0.1477	0.1436	0.0978	0.0845	0.0820	0.0851	0.0677	0.0788	0.0898
Sc04	S-4 (Drt)	0.05	0.1216	0.1460	0.1420	0.1081	0.0835	0.0810	0.0841	0.0669	0.0779	0.0888
Sc05	S-4 (Drt)	0.1	0.1209	0.1452	0.1411	0.1133	0.0830	0.0806	0.0836	0.0665	0.0775	0.0883
Sc06	S-4 (Drt)	0.15	0.1202	0.1444	0.1403	0.1184	0.0825	0.0801	0.0831	0.0661	0.0770	0.0878
Sc07	S-1 (CO2)	−0.15	0.1039	0.1500	0.1458	0.1051	0.0857	0.0832	0.0864	0.0687	0.0800	0.0912
Sc08	S-1 (CO2)	−0.1	0.1101	0.1489	0.1448	0.1044	0.0852	0.0826	0.0858	0.0682	0.0795	0.0905
Sc09	S-1 (CO2)	−0.05	0.1162	0.1479	0.1438	0.1037	0.0846	0.0821	0.0852	0.0678	0.0789	0.0899
Sc10	S-1 (CO2)	0.05	0.1284	0.1459	0.1418	0.1023	0.0834	0.0809	0.0840	0.0668	0.0778	0.0887
Sc11	S-1 (CO2)	0.1	0.1345	0.1448	0.1408	0.1016	0.0828	0.0804	0.0834	0.0664	0.0773	0.0880
Sc12	S-1 (CO2)	0.15	0.1406	0.1438	0.1398	0.1008	0.0822	0.0798	0.0828	0.0659	0.0768	0.0874
Sc13	S-7 (Eff)	−0.15	0.1240	0.1489	0.1448	0.1044	0.0852	0.0826	0.0719	0.0682	0.0795	0.0905
Sc14	S-7 (Eff)	−0.1	0.1234	0.1482	0.1441	0.1039	0.0848	0.0822	0.0761	0.0679	0.0791	0.0901
Sc15	S-7 (Eff)	−0.05	0.1229	0.1476	0.1434	0.1035	0.0844	0.0819	0.0804	0.0676	0.0788	0.0897
Sc16	S-7 (Eff)	0.05	0.1217	0.1462	0.1421	0.1025	0.0836	0.0811	0.0888	0.0670	0.0780	0.0889
Sc17	S-7 (Eff)	0.1	0.1212	0.1455	0.1415	0.1020	0.0832	0.0807	0.0931	0.0667	0.0777	0.0885
Sc18	S-7 (Eff)	0.15	0.1206	0.1448	0.1408	0.1016	0.0828	0.0804	0.0973	0.0664	0.0773	0.0881
Battery Technologies
Scen	Target	Δ	CL	DoD	RTE	SE	VED	C-rate	CpkWh	Maint	Tox	CF	Recycle	Safety	BMS	SoC_RT
Sc01	S-4 (SE)	−0.15	0.3459	0.0249	0.0260	0.0425	0.0673	0.1747	0.0561	0.0479	0.0478	0.0504	0.0337	0.0281	0.0320	0.0226
Sc02	S-4 (SE)	−0.1	0.3450	0.0248	0.0259	0.0450	0.0671	0.1742	0.0559	0.0478	0.0477	0.0503	0.0336	0.0280	0.0319	0.0226
Sc03	S-4 (SE)	−0.05	0.3441	0.0248	0.0258	0.0476	0.0669	0.1738	0.0558	0.0477	0.0476	0.0502	0.0336	0.0280	0.0318	0.0225
Sc04	S-4 (SE)	0.05	0.3423	0.0246	0.0257	0.0526	0.0666	0.1728	0.0555	0.0474	0.0473	0.0499	0.0334	0.0278	0.0317	0.0224
Sc05	S-4 (SE)	0.1	0.3414	0.0246	0.0256	0.0551	0.0664	0.1724	0.0554	0.0473	0.0472	0.0498	0.0333	0.0277	0.0316	0.0223
Sc06	S-4 (SE)	0.15	0.3405	0.0245	0.0256	0.0576	0.0662	0.1719	0.0552	0.0472	0.0471	0.0496	0.0332	0.0277	0.0315	0.0223
Sc07	S-1 (CL)	−0.15	0.2917	0.0267	0.0278	0.0540	0.0720	0.1869	0.0600	0.0513	0.0511	0.0540	0.0361	0.0301	0.0342	0.0242
Sc08	S-1 (CL)	−0.1	0.3089	0.0260	0.0271	0.0527	0.0703	0.1824	0.0586	0.0500	0.0499	0.0526	0.0352	0.0294	0.0334	0.0236
Sc09	S-1 (CL)	−0.05	0.3260	0.0254	0.0264	0.0514	0.0685	0.1778	0.0571	0.0488	0.0487	0.0513	0.0343	0.0286	0.0326	0.0230
Sc10	S-1 (CL)	0.05	0.3603	0.0241	0.0251	0.0487	0.0650	0.1688	0.0542	0.0463	0.0462	0.0487	0.0326	0.0272	0.0309	0.0219
Sc11	S-1 (CL)	0.1	0.3775	0.0234	0.0244	0.0474	0.0633	0.1642	0.0527	0.0451	0.0449	0.0474	0.0317	0.0264	0.0301	0.0213
Sc12	S-1 (CL)	0.15	0.3947	0.0228	0.0237	0.0461	0.0615	0.1597	0.0513	0.0438	0.0437	0.0461	0.0308	0.0257	0.0293	0.0207
Sc13	S-7 (CpkWh)	−0.15	0.3462	0.0249	0.0260	0.0505	0.0674	0.1748	0.0473	0.0480	0.0478	0.0505	0.0338	0.0281	0.0320	0.0226
Sc14	S-7 (CpkWh)	−0.1	0.3452	0.0249	0.0259	0.0504	0.0672	0.1743	0.0501	0.0478	0.0477	0.0503	0.0337	0.0281	0.0319	0.0226
Sc15	S-7 (CpkWh)	−0.05	0.3442	0.0248	0.0258	0.0502	0.0670	0.1738	0.0529	0.0477	0.0476	0.0502	0.0336	0.0280	0.0318	0.0225
Sc16	S-7 (CpkWh)	0.05	0.3422	0.0246	0.0257	0.0499	0.0666	0.1728	0.0584	0.0474	0.0473	0.0499	0.0334	0.0278	0.0317	0.0224
Sc17	S-7 (CpkWh)	0.1	0.3412	0.0246	0.0256	0.0498	0.0664	0.1723	0.0612	0.0473	0.0471	0.0497	0.0333	0.0277	0.0316	0.0223
Sc18	S-7 (CpkWh)	0.15	0.3402	0.0245	0.0255	0.0496	0.0662	0.1718	0.0640	0.0471	0.0470	0.0496	0.0332	0.0276	0.0315	0.0222
Converter Technologies
	Criterion	ConvEff	PF	THD	MTBF	CpkW	Maint	Tox	CF	Recycle	Safety	GridComp	Redund
Sc01	CpkW −15%	0.0282	0.0849	0.0934	0.0396	0.1285	0.1227	0.1062	0.0706	0.0866	0.0609	0.0647	0.1136
Sc02	CpkW −10%	0.0280	0.0841	0.0926	0.0393	0.1360	0.1217	0.1053	0.0700	0.0859	0.0604	0.0641	0.1126
Sc03	CpkW −5%	0.0277	0.0834	0.0918	0.0390	0.1436	0.1206	0.1044	0.0694	0.0851	0.0598	0.0635	0.1116
Sc04	CpkW +5%	0.0273	0.0819	0.0902	0.0383	0.1587	0.1185	0.1025	0.0682	0.0836	0.0588	0.0624	0.1097
Sc05	CpkW +10%	0.0270	0.0812	0.0894	0.0379	0.1662	0.1174	0.1016	0.0676	0.0829	0.0583	0.0618	0.1087
Sc06	CpkW +15%	0.0268	0.0804	0.0885	0.0376	0.1738	0.1163	0.1007	0.0670	0.0821	0.0577	0.0613	0.1077
Sc07	Maint −15%	0.0281	0.0843	0.0928	0.0394	0.1542	0.1016	0.1056	0.0702	0.0861	0.0605	0.0643	0.1129
Sc08	Maint −10%	0.0279	0.0838	0.0922	0.0391	0.1532	0.1076	0.1049	0.0697	0.0855	0.0601	0.0638	0.1122
Sc09	Maint −5%	0.0277	0.0832	0.0916	0.0389	0.1522	0.1136	0.1042	0.0693	0.0850	0.0597	0.0634	0.1114
Sc10	Maint +5%	0.0273	0.0821	0.0904	0.0383	0.1501	0.1255	0.1028	0.0683	0.0838	0.0589	0.0625	0.1099
Sc11	Maint +10%	0.0271	0.0815	0.0897	0.0381	0.1491	0.1315	0.1021	0.0679	0.0832	0.0585	0.0621	0.1092
Sc12	Maint +15%	0.0269	0.0810	0.0891	0.0378	0.1481	0.1375	0.1014	0.0674	0.0827	0.0581	0.0617	0.1084
Sc13	Redund −15%	0.0280	0.0842	0.0927	0.0393	0.1539	0.1218	0.1054	0.0701	0.0860	0.0604	0.0641	0.0941
Sc14	Redund −10%	0.0278	0.0837	0.0921	0.0391	0.1530	0.1210	0.1048	0.0697	0.0854	0.0601	0.0638	0.0996
Sc15	Redund −5%	0.0277	0.0832	0.0915	0.0388	0.1521	0.1203	0.1041	0.0692	0.0849	0.0597	0.0634	0.1051
Sc16	Redund +5%	0.0273	0.0821	0.0904	0.0384	0.1502	0.1188	0.1028	0.0684	0.0839	0.0589	0.0626	0.1162
Sc17	Redund +10%	0.0272	0.0816	0.0898	0.0381	0.1492	0.1180	0.1022	0.0680	0.0833	0.0586	0.0622	0.1217
Sc18	Redund +15%	0.0270	0.0811	0.0893	0.0379	0.1483	0.1173	0.1015	0.0675	0.0828	0.0582	0.0618	0.1273

.

Figure 5 illustrates the outcomes of the sensitivity analysis across the 18 perturbation scenarios. For PV technologies, PV-Thf consistently retained the top position, followed by PV-Tpc and PV-Hjt, with only minor positional changes observed between PV-Mcs and PV-Bfmc under higher perturbation levels. Within the battery domain, HSCs maintained a decisive lead, with SSB and Al–Ion ranking next, while conventional chemistries such as LA and NiFe persistently occupied the lowest positions. For converter technologies, MCs and PCs consistently occupied the leading ranks, whereas SC remained the lowest across all scenarios.

Spearman correlation values with respect to the baseline rankings were nearly unity ($\rho \ge 0.964$), accentuating a high degree of stability. At higher perturbation levels, only mid-tier technologies (e.g., PV-Mcs vs. PV-Bfmc, NMC vs. 2ndEV) exhibited positional shifts. In the case of redundancy and cost variations, minor exchanges were observed between FCs and MMCs, but these occurred exclusively among the lowest-ranked options. For converters, shifts occurred between FCs and MMCs under perturbations in redundancy (−15%) or cost per kW (+15%).

The overall robustness of the leading alternatives across all perturbation scenarios reaffirms the methodological soundness of the MARCOS approach. This strengthens confidence that the prioritized technologies are not artifacts of specific weighting schemes but rather represent stable, reliable choices for sustainable HRES design and planning.

5. Discussion of Top-Ranked Technologies and Implications of the Study

5.1. Discussion of Top-Ranked Technologies

The ranked results identified the most promising technologies for solar-based HRESs, highlighting thin-film photovoltaic (PV-Thf) modules, hybrid supercapacitors (HSCs), and modular converters (MCs). While these outcomes emphasize significant technical, environmental, and economic advantages, it is essential to contextualize them against practical implementation realities. Accordingly, the discussion of these leading technologies should be regarded not as prescriptive recommendations, but rather as indicative signposts of potential trends in the future development of sustainability-oriented HRESs.

Thin-film PV modules were prioritized over crystalline technologies primarily due to their short energy payback time, low embodied carbon, and lightweight construction, making them highly suitable for rooftop and other weight-constrained applications. Their strong performance under diffuse irradiance and high-temperature conditions further reinforces their suitability across varied climatic environments. However, thin-film modules generally exhibit lower efficiency (11–15%) compared to monocrystalline and heterojunction modules (20–24%), and concerns regarding long-term durability persist, particularly with respect to encapsulant performance and degradation over extended field deployment. These issues continue to represent barriers to large-scale commercial adoption [106].

Hybrid supercapacitors emerged as the optimal storage technology, offering longer cycle life (>100,000 cycles) [94], higher power density, rapid charge–discharge rates, and superior safety relative to lithium-ion batteries (LIBs) and other storage systems. These attributes make HSCs particularly well-suited for high-frequency cycling applications, such as solar intermittency smoothing and grid stabilization. Nevertheless, their lower energy density (~20–30 Wh/kg compared to 150–250 Wh/kg for LIBs), limited scalability, higher costs, and immature supply chains for raw materials have constrained their widespread deployment. Thus, HSCs may be considered a forward-looking technology with strong near-term potential as a complement to lithium-ion systems, while holding longer-term promise for sustainable energy storage solutions.

Within the converter domain, modular converters ranked highest, underscoring their scalability, redundancy, and superior harmonic performance. MCs are capable of achieving conversion efficiencies exceeding 97% while maintaining low total harmonic distortion (<2–3%) without requiring extensive filtering, making them highly grid-compliant and reliable for multi-source HRES applications. Their modular design, which facilitates fault isolation and system expansion, provides distinct advantages over traditional string and central inverters. However, these benefits are tempered by higher capital costs, system complexity, and evolving challenges in meeting strict grid-code requirements, particularly in comparison with other renewable energy converter models. Although improved fault tolerance and reduced downtime across the system lifecycle may offset these drawbacks, comprehensive cost–benefit analyses remain necessary [107].

Table 9. Comparative analysis of top-ranked PV, battery, and converter technologies in solar-based HRESs.

Technology	Key Benefits (Ranking Justification)	Limitations/Real-World Constraints	Practical Relevance and Outlook
PV-Thf (Thin-Film PV)	Short EPBT, low CO2 footprint, lightweight, good performance under high-temp and diffuse light	Lower efficiency vs. crystalline, shorter lifespan, reliability issues in harsh climates	Strong option for sustainability-focused and rooftop/large-scale projects; long-term reliability needs improvement
HSC (Hybrid Supercapacitors)	>100,000 cycles, high power density, fast charge–discharge, high safety	Limited production scale, high costs, uncertain raw material supply	Future-oriented option; complements Li-ion for high cycling applications; promising for next-gen HRESs
MC (Modular Converter)	Scalability, redundancy, high efficiency, low THD, grid compatibility	Higher capital cost, complex design, compliance challenges	Commercially feasible for multi-source HRESs; lifecycle reliability offsets upfront cost

provides a comparative summary of the merits, limitations, and practical viability of these top-ranked technologies. Overall, the findings demonstrate that while thin-film PV modules and modular converters are already commercially attractive and increasingly deployed, hybrid supercapacitors remain an emerging technology. These results highlight that prioritization outcomes should be interpreted as reflective snapshots of current technological strengths and as indicators of future directions for innovation, representing the most promising pathways for advancing solar-based HRESs.

The ranking results further reinforce the synergistic rationale explaining why these three technologies—thin-film PV, hybrid supercapacitors, and modular converters—demonstrate distinct advantages over their counterparts within hybrid renewable energy configurations, thereby enriching the interpretation of these findings. Collectively, these technologies generate system-level synergies that produce a multiplicative impact on efficiency, reliability, and sustainability, advancing the integrated performance of solar-based hybrid systems.

Beyond their conventional advantages of lower environmental footprint and reduced equipment stress, thin-film PV technologies exhibit remarkable thermal tolerance and consistent performance under partial shading and low-irradiance conditions, characteristics that are especially critical for tropical and semi-arid installations. Their lightweight construction contributes to reduced balance-of-system requirements, lowering installation costs and minimizing structural loads in rooftop and floating PV applications. These attributes collectively explain their strong performance across environmental and operational criteria, despite relatively lower nominal conversion efficiencies [108].

Hybrid supercapacitors achieved high rankings due to their exceptional cycling stability, rapid charge–discharge kinetics, and inherent safety, all of which are crucial for mitigating solar intermittency in hybrid renewable systems. By delivering high power outputs over short durations with negligible losses, HSCs effectively smooth transient fluctuations and support frequency regulation. Although their limited energy density restricts long-duration storage, their integration as fast-response buffers alongside higher-capacity batteries represents a practical and scalable approach to enhancing overall system stability and dynamic performance [94].

Modular converters, particularly modular multilevel converters, are distinguished by their outstanding scalability, fault tolerance, and redundancy, ensuring continuous operation even under partial module failures. Their capability to maintain grid synchronization, improve grid-side voltage quality, and minimize total harmonic distortion enables superior grid compliance and enhances power delivery reliability. Moreover, their modular design facilitates ease of maintenance, fault isolation, and high system availability, positioning them as a cornerstone technology for multi-source renewable architectures and future high-voltage DC (HVDC) grid integration [109].

Overall, these insights demonstrate that the top-ranked technologies not only excel in discrete performance metrics but also provide context-specific advantages that align closely with the evolving design priorities of solar-based HRESs, namely, resilience, adaptability, and lifecycle sustainability.

5.2. Implications of the Study

This study carries both theoretical and practical significance, deriving implications from the findings on HRESs. By integrating novel energy system elements within the broader concept of hybrid renewables, which are less dependent on fossil fuels, it contributes to advancing sustainable energy design. Through the integration of the MARCOS method with multiple objective weighting schemes and comparative MCDM techniques, this work proposes a rigorous and transparent approach for ranking enabling technologies.

• Theoretical Implications

This study advances decision science by demonstrating how methodological biases associated with individual weighting or ranking techniques can be mitigated through fusion strategies such as the Bonferroni operator. The comparative assessment of nine MCDM methods against MARCOS strengthens the consistency of prioritization outcomes and reinforces confidence in multi-method frameworks. Furthermore, the incorporation of sensitivity analysis highlights the importance of robustness testing in decision modeling, ensuring that technology selection remains stable underweight perturbations. Such methodological rigor offers valuable insights for future applications of MCDM in energy, materials, and sustainability research domains.

• Practical Implications

The results establish clear benchmarks for HRES design from a technology adoption perspective. Thin-film PV modules, hybrid supercapacitors, and modular converters consistently emerged as superior technologies under both baseline and perturbed conditions. These findings provide actionable insights for system designers, policymakers, and investors seeking strategies for risk reduction, improved reliability, and long-term efficiency. Importantly, the results also suggest that less competitive technologies may become viable under evolving market or policy contexts, such as declining costs of aluminium-ion batteries or improved redundancy in converters.

• Policy and Sustainability Implications

The study offers a transparent and reproducible framework for sustainability-oriented technology assessment, aligning with global energy transition and climate mitigation goals. This model can guide policymakers in shaping procurement policies, harmonizing evaluation protocols, and supporting the development of resilient energy infrastructures, including cold-chain and agro-industrial systems. By incorporating life-cycle and environmental performance dimensions, the framework effectively bridges technical optimization with sustainability assessment, thereby encouraging responsible and forward-looking innovation.

In summary, this study provides a decision-support roadmap that is academically rigorous yet practically relevant, facilitating the implementation of reliable, decarbonized HRES technologies.

6. Conclusions

This study presents a comprehensive MCDM framework for the prioritization of PV, battery, and converter technologies within solar-based HRESs. The framework ensures unbiased and balanced determination of criteria weights by integrating five objective weighting methods: Entropy, Standard Deviation, CRITIC, MEREC, and CILOS, consolidated through the Bonferroni operator. The MARCOS method was employed as the principal ranking technique, complemented by comparative assessments using eight additional MCDM methods, along with correlation and sensitivity analyses to verify robustness.

The findings reveal that thin-film (PV-Thf), tunnel oxide passivated contact (PV-Tpc), and heterojunction (PV-Hjt) PV technologies demonstrated superior performance in terms of efficiency, durability, and carbon reduction potential, positioning them as the most sustainable options. Among energy storage technologies, hybrid supercapacitors (HSC) ranked highest, outperforming solid-state and aluminium-ion batteries due to their exceptional cycle life, safety, and operational stability. With respect to converters, modular converters (MC) attained the highest scores, with strong performances also observed for power converters (PC) and neutral-point clamped (NPC) topologies, attributable to their enhanced grid integration capability and long-term reliability.

The results of the sensitivity analysis confirmed the stability of these rankings under perturbations of up to ±15% in criteria weights, thereby reinforcing that the identified top-ranked technologies are not artifacts of particular weighting assumptions but represent robust choices across diverse scenarios. Furthermore, correlation analyses demonstrated strong alignment of MARCOS with most alternative MCDM methods, further validating the reliability of the framework.

The anticipated MCDM framework thus addresses a key research gap by shifting emphasis from macro-level HRES planning to micro-level component prioritization within solar-based HRESs. The results provide policymakers, system designers, and engineers with actionable insights to support the deployment of hybrid energy systems that are more viable, cost-effective, and sustainable. Future work could extend this approach to dynamic operational scenarios, incorporate emerging technologies such as hydrogen-based storage and advanced inverters, and integrate lifecycle-based optimization under region-specific policy and market conditions.

The proposed framework provides structured support for component procurement decisions within the diverse renewable integration strategies currently pursued in India and aligns closely with the objectives of SDG-7 on sustainable energy. This approach offers utility to both policymakers and engineers by enabling a balanced consideration of capital expenditure, lifecycle cost, and lifecycle environmental impacts.

Limitations and Future Work

Although the proposed framework provides a robust and reliable approach for prioritizing solar-based HRES technologies such as PV, battery, and converter systems, it is not without limitations. First, the analysis relies on static decision matrices derived from literature-based performance, economic, and environmental indicators. These metrics were not explicitly modeled to account for real-world variability, including regional climatic conditions, market-wide economic fluctuations, or long-term technology degradation. Second, the framework does not incorporate time-dependent operational parameters, such as stochastic solar irradiance, fluctuating load demand, or maintenance schedules, which can significantly affect component performance in practical implementation.

Furthermore, hydrogen-based storage and emerging converter architectures were not extensively considered due to the limited availability of consistent data, potentially constraining the transferability of results as new technologies continue to evolve. Future work should extend the framework to dynamic simulation models, integrate bio-circular value chain life-cycle cost assessments, and apply stochastic optimization under uncertainty to better reflect real-world conditions. Coupling the framework with techno-economic databases and region-specific policy constraints would enhance applicability, while incorporating advanced technologies such as solid oxide batteries, perovskite PV, and AI-assisted power electronics would ensure continued relevance in an increasingly dynamic renewable energy landscape.