Critical Factors Affecting Hybrid Renewable Energy Integration in Rural China: A Stakeholder-Oriented DEMATEL-ISM Analysis

Abstract

Hybrid renewable energy integration (HREI) in rural regions has received limited scholarly attention due to heterogeneous resource endowments, complex development conditions, and multiple coordination challenges. To better understand the factors affecting HREI implementation in rural China, this study develops a stakeholder-oriented analytical framework and applies an integrated DEMATEL–ISM approach. This study identifies 13 critical factors and groups them into four dimensions: complexity of the system, benefit coordination, efficiency coordination, and information coupling. An integrated DEMATEL-ISM approach is employed to examine the causal relationships among these factors and to construct their hierarchical structure. The analysis is informed by a six-member expert panel and a four-round Delphi process. The results show that five factors belong to the cause group and seven to the effect group, while one factor remains balanced. In terms of relative importance, the three highest-weighted factors are synergy degree among multiple agents (CS1, 0.111), information coupling mechanism (IC1, 0.096), and coordinated management of key resources (EC3, 0.093). In terms of net causal influence, complicated rural environment (CS4, R − C = 1.00) is the strongest driving factor, whereas construction and O&M costs (BC3, R − C = −0.77) is the most dependent effect factor. The proposed five-level hierarchical model further indicates that the complicated rural environment, the sustainability of government subsidy policies, and the supervision and service constitute the foundational layer of HREI development. This study provides stakeholder-oriented insights for understanding and promoting HREI in rural China.

1. Introduction

Amid China’s pursuit of carbon neutrality and energy structure transformation, the establishment of a clean, low-carbon, and multi-energy integrated rural energy system has emerged as a cornerstone in national energy security strategy and a vital step toward fulfilling the commitments of the Paris Agreement. Over the past decade, China has contributed to nearly two-thirds of global oil consumption growth and has retained dominance in the global coal market, intensifying its carbon emission reduction pressures [1]. In this context, transitioning to renewable energy is both a strategic imperative and a formidable challenge, particularly in rural areas where 79.3% of energy consumption relies on direct-burning biomass sources such as straw and coal [2].

The main application objectives of renewable energy sources used in rural China are heating and cooking [3]. Although rural areas contribute a considerable share of carbon emissions in many countries, academic and policy discussions on reducing these emissions through renewable energy integration remain insufficient [4]. This gap highlights the need for a more systematic understanding of rural energy systems and their potential role in supporting national decarbonization objectives. Against this background, improving energy utilization efficiency and accelerating renewable energy adoption in remote and resource-constrained rural areas has become increasingly urgent. Yet the link between renewable energy development and rural socio-economic progress has often been assumed rather than systematically conceptualized or effectively fostered in practice [5]. As rural areas become increasingly central to the implementation of China’s energy transition agenda, research on rural renewable energy and hybrid renewable energy has gained growing importance [6,7]. This, in turn, calls for the identification and analysis of the key factors shaping renewable energy system implementation through a stakeholder-oriented analytical framework, so that policy and development strategies can be both theoretically grounded and practically applicable.

To advance renewable energy integration while fostering a sustainable and resilient development environment, a systemic and synergistic approach across multiple sectors has increasingly been regarded as a viable way to address both energy security and environmental concerns [8,9]. In this context, the promotion of hybrid renewable energy in rural areas has attracted growing attention, not only because of its ecological and sustainability benefits, particularly its ability to reduce greenhouse gas emissions compared with conventional fossil-based energy sources [10], but also because of its economic and technical advantages over single-source renewable energy systems. On the one hand, hybrid renewable energy solutions offer greater flexibility, higher reliability, and better efficiency than single-source systems, thereby improving energy access in off-grid and remote rural areas [11,12]. On the other hand, the cost of renewable energy technologies has declined substantially in recent years, enhancing their market competitiveness; for example, solar power has been offered at bid prices as low as 0.0104 US$/kWh [13].

Recognizing this strategic potential, China’s policy framework has introduced comprehensive guidelines that promote hybrid renewable energy integration (HREI) as a promising and feasible solution. National policy points out that China will construct a clean and low-carbon modern rural energy system through the development of HREI [14]. HREI emphasizes the coordinated integration of multiple renewable energy sources, including wind, solar, biomass, and hydrogen, in order to improve efficiency, enhance the continuity of power supply, and reduce system uncertainty [15,16,17]. More importantly, HREI involves interactions across multiple stakeholder domains, including policy support, energy development, infrastructure operation, and end-user adaptation [13,18]. Rather than focusing on a single actor, its implementation depends on how technical, institutional, and organizational conditions are coordinated across these domains. HREI adopts a ‘system of systems’ approach, fostering synergistic interactions and coupling mechanisms among stakeholders by considering interactions within the systems [19], thereby improving the overall system performance.

Recent international studies have increasingly shown that the development and integration of renewable energy systems are shaped not only by technical performance, but also by broader institutional, financial, infrastructural, and policy conditions. Across different national contexts, the success or failure of hybrid renewable development often depends on how these non-technical factors interact with local implementation environments. For example, recent evidence from the United States highlights the importance of coordinated subsidy schemes and regulatory authorization in supporting renewable energy development [20]. In South Africa, the continued dominance of coal and the slow uptake of renewables have been linked not only to cost considerations, but also to policy, technical, and infrastructural constraints [21]. Studies on ASEAN countries similarly reveal strong cross-country disparities, showing that renewable deployment is facilitated where innovation capacity, financing access, and infrastructure are stronger, but constrained where these conditions remain weak [22]. In Denmark, recent research points to the misalignment between national energy planning and regional grid capacities, underscoring the importance of institutional coordination between energy development and infrastructure planning [23]. Taken together, these studies suggest that renewable energy integration is highly context-sensitive and that non-technical conditions play a decisive role in shaping implementation outcomes.

Compared with this growing international literature, China’s rural renewable trajectory is strongly shaped by government-led policy frameworks, rural revitalization initiatives, and public investment in hybrid energy systems. Meanwhile, the existing literature on HREI remains largely technically oriented and contextually fragmented, and several important gaps remain evident. First, research on rural HREI in China has focused predominantly on technical performance [5,24], optimization [25,26], and techno-economic feasibility [13], while insufficient attention has been paid to non-technical factors such as governance arrangements, coordination mechanisms, and institutional constraints. Second, although stakeholder theory-based approaches have been applied in related fields such as transport electrification and infrastructure [27,28], energy transition planning [29,30], and alternative technology selection [31], their use as a rigorous analytical perspective for systematically identifying, categorizing, and evaluating the factors affecting rural HREI in China remains limited. Third, existing studies pay insufficient attention to China-specific institutional bottlenecks that shape rural HREI deployment, including system performance and feasibility [32,33,34], environmental and ecological trade-offs [35,36,37], etc. Consequently, the literature still lacks a systematic understanding of which factors are most critical for rural HREI in rural China, how these factors interact, and which of them act as deeper driving forces.

Moreover, existing studies have made substantial progress in the technical domain. In rural contexts, prior research has examined benefits, barriers, environmental impacts, and electrification strategies, often through techno-economic modeling, optimization techniques, and scenario-based analyses. However, the emphasis remains largely on operational efficiency, cost minimization, and performance metrics. By contrast, non-technical dimensions, especially factor identification and stakeholder-related interactions in rural HREI development remain underexplored.

To address these gaps, this study adopts a stakeholder-oriented analytical framework and applies an integrated DEMATEL–ISM approach to investigate the critical factors affecting HREI in rural China by conceptualizing rural HREI as a complex and interdependent system. Specifically, this study addresses the following research questions:

(1)

Q1: What are the critical factors affecting HREI application in rural China within a stakeholder-oriented analytical framework?

(2)

Q2: How can these critical factors be ranked and prioritized according to their causal relationships and structural positions?

(3)

Q3: What practical implications can be derived from these factors to promote HREI adoption in rural China?

DEMATEL, first proposed in 1971, provides a structured approach for analyzing interrelationships and assessing the relative importance of factors within complex and uncertain systems [38]. By combining matrix operations with digraph theory, DEMATEL transforms qualitative judgments into quantitative influence measures, thereby facilitating the identification of both direct and indirect relationships among system factors [39]. Existing studies have applied DEMATEL, either independently or in combination with other analytical tools, to investigate influencing factors and barriers in diverse domains, such as the construction industry [40], coal supply [41], clean energy infrastructure [42], and distributed PV systems [43]. These applications demonstrate the method’s strength in identifying causal prominence and prioritizing complex interdependencies. However, most of these studies remain concentrated in technical, industrial, or infrastructure-oriented contexts, and they provide limited insight into rural HREI under context-sensitive socio-institutional conditions.

ISM, in contrast, is effective in translating expert knowledge and practical experience into a hierarchical structure among multiple sub-dimensions, offering a clear graphical representation of their interconnections [32]. It has been widely used in combination with other decision-support methods, including DEMATEL-ISM [38], ISM-TOPSIS [44], and ISM-MICMAC [33], to reveal structural relationships in complex systems. Nevertheless, ISM has inherent limitations. Its strong reliance on expert judgment may introduce subjectivity, and it cannot quantify the magnitude of influence among factors, which restricts its ability to distinguish the relative strength of inter-factor relationships [44]. For this reason, ISM is often integrated with complementary weighting or causal analysis methods to improve analytical rigor.

In this regard, the integrated DEMATEL–ISM approach is particularly appropriate for the present study because it combines the causal quantification strength of DEMATEL with the structural hierarchy-building capacity of ISM. Moreover, the integrated DEMATEL-ISM approach is particularly effective for investigating interdependencies because it simultaneously considers both direct and indirect factor relationships. This makes it well-suited for systematic investigation of integration mechanisms underlying HREI, and it performs effectively even with small sample sizes, a benefit highlighted in its diverse applications [41]. More importantly, compared with prior DEMATEL–ISM applications in related fields, its use in rural HREI remains limited, especially in studies that seek to account for non-technical factors, stakeholder-oriented interactions, and China-specific institutional bottlenecks. Therefore, the value of DEMATEL–ISM in this study lies not only in its methodological suitability for analyzing interdependent factors, but also in its ability to address a context that has been insufficiently examined in the existing literature.

Deploying HREI in rural China involves a highly complex interplay of political, economic, social, and technical factors. Many of these factors influence one another in both direct and mediated ways, and stakeholder perspectives may differ regarding their influence paths. In this context, the hybrid DEMATEL-ISM is particularly appropriate for addressing the stated research questions in several ways: (1) DEMATEL quantifies both direct and indirect influence among factors, allowing us to distinguish driving vs. driven variables; (2) ISM structures those relations into a multi-level hierarchy, clarifying root causes versus dependent ones; (3) together, they yield a causally annotated structural map that aligns with our research aims: identifying (Q1) and ranking (Q2) the factors, and translating that structure into actionable guidance (Q3) for stakeholders and policymakers. By analyzing the critical factors and their interrelationships, the study delivers informed policy recommendations to enhance stakeholder collaboration and strategic prioritization. The findings hold significant implications for advancing hybrid renewable energy integration in rural China, fostering a sustainable rural energy ecosystem that accommodates multiple stakeholder interests.

Compared with existing studies on rural renewable energy and HREI, this study contributes in three respects. First, whereas much of the current literature focuses primarily on technical design, optimization, electrification strategies, or techno-economic performance, this study shifts attention to a broader set of socio-institutional and coordination-related factors that shape HREI implementation in rural China. On this basis, it identifies 13 critical factors and organizes them into four interrelated dimensions: complexity of the system, benefit coordination, efficiency coordination, and information coupling. Second, although stakeholder-related issues are frequently acknowledged in the literature, they are less often adopted in a structured analytical framework for factor identification and prioritization in the rural HREI context. This study therefore develops a stakeholder-oriented analytical framework considering the practical and developmental conditions of rural China. Third, while DEMATEL-ISM has been applied in related fields such as infrastructure, supply chains, and energy systems, its use in the context of rural HREI remains limited. By applying an integrated DEMATEL-ISM approach to this underexplored setting, the study reveals both the causal structure and hierarchical relationships among critical factors and provides policy-relevant priorities for promoting rural HREI development in China.

The organization of the paper is as follows: The factors identification and methodology are conducted in Section 2. Applications of the proposed method and the results are presented in Section 3. Section 4 provides an in-depth discussion about correlation analysis, comparative, sensitivity analysis, and implications. Finally, conclusions are summarized in Section 5.

2. Factors and Methodology

2.1. Factors Affecting HREI in Rural China

A comprehensive and systematic identification of the factors affecting HREI in rural China is conducted to identity the critical factors (see details in Section 4.1). Given the multi-stakeholder nature of rural HREI and its reliance on synergistic effects, the influencing factors are categorized into four key dimensions: (a) The complexity of the system, which addresses the multifaceted nature of rural HREI, including stakeholder collaboration and environmental challenges; (b) benefit coordination, which focuses on balancing economic and financial considerations among stakeholders while mitigating costs and potential losses; (c) efficiency coordination, which emphasizes optimizing resource utilization and operational efficiency across the energy value chain; (d) information coupling, which strengthens data sharing, communication, and transparency to facilitate coordinated decision-making. Based on factors identified in the reviewed literature and the factor identification procedure, we construct a critical factor system, as shown in

Table 1. Critical factors affecting HREI in rural China.

Dimensions	Factors	Nomenclature
Complexity of the system (CS)	Synergy degree among multiple agents	CS1
	Life-cycle cost–benefit performance	CS2
	Technological collaboration barriers among stakeholders	CS3
	Complicated rural environment	CS4
Benefit coordination (BC)	Sustainability of government subsidy policies	BC1
	Financing access	BC2
	Construction and O&M costs	BC3
Efficiency coordination (EC)	Energy management and integration efficiency	EC1
	Rural power grid upgrade	EC2
	Coordinated management of key resources	EC3
Information coupling (IC)	Information coupling mechanism	IC1
	Supervision and service	IC2
	Professional talents	IC3

.

2.1.1. Complexity of the System (CS) Dimension

Synergy degree among multiple agents (CS1): It reflects the extent to which various stakeholders coordinate their activities, share resources, and align goals to achieve collaborative benefits and enhance overall system efficiency [34,45]. A higher degree of synergy indicates more effective stakeholder interactions, leading to improved system performance.

Life-cycle cost–benefit performance (CS2): This factor evaluates the long-term balance between economic costs and social–environmental benefits across the entire project life cycle. Conducting a comprehensive cost–benefit analysis is essential for assessing the economic feasibility [46]. Such analysis improves the effectiveness of decision-making by solving complex system problems with many dimensions [47,48].

Technological collaboration barriers among stakeholders (CS3): [42,49,50]: Technological collaboration barriers arise due to intermittency and variability of different energy sources (including solar, wind, energy storage, and grid connection), the complexity of integrating diverse system components (power sources, converters, and controllers), and the need to maintain grid stability during multi-energy coupling.

Complicated rural environment (CS4): Achieving HREI deployment in rural regions is inherently complex, creating formidable challenges for renewable integration. These challenges stem from underdeveloped infrastructure and low network resilience, which reduce system stability and restrict load flexibility [51,52]; heterogeneous energy demand patterns caused by variations in living habits and willingness to pay for energy [51]; and imbalance in policy development and weak linkages among stakeholders [53].

2.1.2. Benefit Coordination (BC) Dimension

Sustainability of government subsidy policies (BC1): Stable and well-designed subsidy policies play a pivotal role in incentivizing investment, ensuring the long-term viability, maintaining the economic attractiveness and operational continuity of rural HREI projects [42,43].

Financing access (BC2): Improving access to credit, attracting social and commercial investment, and establishing innovative financing models (e.g., green bonds, public–private partnerships) are crucial for overcoming funding gaps and enabling project replication [54,55].

Construction and O&M costs (BC3): High upfront construction costs and significant operation and maintenance (O&M) expenses present substantial financial burdens for rural renewable energy projects [56].

2.1.3. Efficiency Coordination (EC) Dimension

Energy management and integration efficiency (EC1): This factor refers to the capacity to effectively coordinate multiple energy sources through advanced management strategies and technologies, thereby improving system reliability and operational performance [13].

Rural power grid upgrade (EC2): Upgrading the existing rural power grid is essential for accommodating distributed renewable energy sources and ensuring the secure and stable operation of hybrid systems [57].

Coordinated management of key resources (EC3): The integration of energy, land, grid, storage, and other critical resources through coordinated management can significantly improve the efficiency and sustainability of rural energy systems [58,59].

2.1.4. Information Coupling (IC) Dimension

Information coupling mechanism (IC1): This factor refers to the establishment of standardized platforms, protocols, and digital infrastructures that enable stakeholders to share real-time data on system operations, energy supply, and demand fluctuations [60,61].

Supervision and service (IC2): Regulatory oversight and technical service mechanisms are critical for ensuring that information coupling is accurate, reliable, and aligned with system goals. Strong supervisory frameworks promote compliance by stakeholders with technical and policy requirements [62,63].

Professional talents (IC3): The availability of qualified professional talents with diverse knowledge, skills, and cultural backgrounds in areas such as renewable energy, green production, technological innovation, and digital governance plays a vital role in supporting information flow, knowledge exchange, and technical services, strengthening the capacity of rural regions to implement HREI decision-making and to sustain long-term system resilience [64,65].

It is worth mentioning that the synergy degree emphasizes functional and operational collaboration among stakeholders. In contrast, information coupling focuses on data flow and integration of information using factors including information coupling mechanism (IC1), supervision and service (IC2), and professional talents (IC3) rather than the actual outcomes of cooperative action. For example, in a biomass–solar microgrid project, if the project implements a shared data system that gives both parties access to each other’s generation data but they operate independently without scheduling generation to complement each other, the system would exhibit high information coupling but low synergy degree. Thus, the factor synergy degree among multiple agents (CS1) is different from the factor dimension information coupling (IC).

2.2. Methods

The application of the integrated DEMATEL-ISM proceeds in three main stages: critical factors identification, cause-and-effect relationships determination and influence weight assignment with DEMATEL, establishment the hierarchical structure of factors with ISM method. The methodology is illustrated in Figure 1.

2.2.1. Factor Identification

We conducted an extensive literature review to compile a preliminary set of factors affecting HREI in rural China. To refine and validate this list, we used a Delphi method. Experts were selected purposively according to three criteria: (1) demonstrated academic or professional expertise in renewable energy systems, hybrid energy integration, energy management, or related fields; (2) sustained research or practical engagement relevant to rural energy systems or sustainable power supply; and (3) willingness to participate in multiple Delphi rounds and provide justification for proposed revisions. In addition, efforts were made to include experts from different institutional backgrounds, including universities, research institutions, and enterprises. Detailed information about the experts is presented in

Table 2. Expert panel profile.

NO.	Area	Affiliation	Degree	Job	Gender
E1	Renewable energy systems	University	PhD	Researcher	Female
E2	Hybrid energy integration	University	PhD	Professor	Female
E3	Energy management	Enterprise	PhD	Senior research scientist	Male
E4	Sustainable power supply	University	Master’s	PhD student	Female
E5	Power supply and demand	Research institution	PhD	Senior research scientist	Male
E6	Energy management strategy	University	Master’s	PhD student	Male

. Although the panel covered several knowledge domains relevant to HREI, it did not include direct representation from all stakeholder groups involved in rural implementation, such as local government agencies, utility providers, and rural communities. Therefore, the stakeholder dimension of this study should be understood primarily as an analytical framing rather than a fully representative stakeholder sample. Accordingly, the emphasis in this study is placed on the relevance and expertise of the respondents rather than on statistical representativeness. The present study should therefore be understood as an exploratory, expert-based analysis of critical factors affecting rural HREI in China, rather than as a nationally representative survey.

At the end, 13 key factors remained, categorized under the four dimensions. The factor identification process involved the following steps:

Step 1: Information gathering and preliminary factors identification. Based on the reviewed literature, relevant factors were compiled from reports, the academic literature, policy documents, and recent developments. Repetitive or overlapping items were removed, yielding a preliminary list.

Step 2: Expert review and refinement. The preliminary factor list was distributed to each expert, who was asked to review and refine factors and provide justifications. The refined results were collected and distributed to experts for a second round. Experts were encouraged to compare their opinions and revise their evaluations as appropriate. This iterative process continued for four rounds. Convergence was assessed based on the stabilization of the factor list and the reduction in substantive revision requests across rounds. The process was terminated after the fourth round because no new factors were introduced, no further mergers or deletions were proposed, and all experts indicated that no further substantive revisions were necessary.

Step 3: Brainstorm session for final factor list. First, the researchers presented the consolidated list of factors after the four-round evaluation, highlighting those with unresolved disagreements. Second, experts discussed the factors they believed should be excluded and stated their reasons. Third, each controversial factor was put to a vote, with experts rating with “agree” or “not agree”. A consensus threshold of 60% was applied: any factor receiving more than 60% “not agree” votes was removed from the list. Finally, the finalized set of factors was obtained (see Table 1). It is noteworthy that there is no universally accepted threshold for defining consensus. Previous methodological reviews indicate that consensus thresholds vary widely across Delphi research, typically ranging from 50% to 97% agreement, depending on the research objectives and disciplinary context [66]. Several Delphi applications have adopted agreement levels around 60% as the minimum threshold for consensus, particularly in exploratory studies or studies involving relatively small expert panels. For example, Sforzini et al. defined weak consensus as 51–60% agreement in a Delphi consensus guideline study [67]. Considering the exploratory nature of this research and the limited size of the expert panel, a 60% agreement threshold was adopted as the minimum criterion for consensus.

2.2.2. Data Collection for Application

Questionnaires and expert interviews were conducted from November 2024 to December 2024 to collect data for applying both the ISM and DEMATEL approaches. An expert panel comprising six members was formed for data collection. The panel members possessed expertise in renewable energy systems, hybrid energy integration, energy management, sustainable power supply, power supply and demand, and energy management strategies. They included professors, researchers, and PhD students from universities, research institutions, and enterprises. Several considerations guided the determination of panel size. First, the research design draws on and aligns with precedent in similar studies. For instance, Xu and Zou (2020) suggested that a panel of four to eight qualified experts is sufficient for a reliable respondent pool [68], whereas Novakowski and Wellar (2009) proposed that a minimum of five experts constitutes an adequate sample size [69]. Second, prior DEMATEL-ISM studies provide grounded reference points: Guo et al. (2024) argued that five experts were adequate to achieve data saturation and capture a diversity of perspectives [38]. Wu et al. (2022) used a five-member expert committee to explore critical barriers to rural distributed PV adoption, achieving full data collection [43]. Third, in applications, the quality of respondents is often deemed more important than panel size [70], with some scholars indicating that the method can operate effectively with as few as two experts [71]. Last, all experts in this study were selected for their substantial experience and knowledge in relevant domains.

The objectives of data collection are (1) to establish the direct-relation matrix based on expert opinions, construct the cause-and-effect relationships among factors, and evaluating their relative significance; and (2) to develop the structural self-interaction matrix (SSIM), determine the initial and the final reachability matrices, and establish the hierarchical structure reflecting the relationships among the identified factors. Following the identification of key factors, data collection was conducted using two separate questionnaires (one for DEMATEL and one for ISM). The questionnaires were sent individually to each expert on the panel separately and all participants completed them, resulting in a 100% response rate.

2.2.3. DEMATEL Methodology

The following outlines the specific steps involved in this technique:

(1) Identify critical factors. The identified factors set can be represented as follows:

$$F={(f_1,f_2,\cdots ,f_n)}^T; n=13$$

(1)

(2) Establish the direct-relation matrix $A$. Collect the data on the influential effects between the factors through pairwise comparisons, utilizing expert opinions gathered via a four-level rating scale ranging from 0 to 3 (see

Table 3. Linguistic expressions and corresponding scales.

Scale	0	1	2	3
Linguistic expression	No influence	Low influence	Moderate influence	High influence

). Compute the arithmetic mean and develop the direct-relation matrix $A$, as represented in Equation (2).

$$A={\left[\begin{array}{ccccc}{a}_{11}& \cdots & a_{1j}& \cdots & a_{1n}\\ ⋮& ⋮& ⋮& ⋮& ⋮\\ a_{i1}& \cdots & a_{ij}& \cdots & a_{in}\\ ⋮& ⋮& ⋮& ⋮& ⋮\\ a_{n1}& \cdots & a_{nj}& \cdots & a_{nn}\end{array}\right]}_{n\times n};i,j\in \left\{1,\cdots ,n\right\}$$

(2)

where $a_{ij}$ represents the arithmetic mean of influence of factor $i$ on $j$.

(3) Obtain the normalized direct-relation matrix $O$:

$$O={\displaystyle \frac{1}{\underset{1\le i\le n}{\max }{\displaystyle \sum _{j=1}^n{a}_{ij}}}}A$$

(3)

(4) Calculate the total-relation matrix $T$ using Equation (4):

$$T={\left[t_{ij}\right]}_{n\times n}=O{(I-O)}^{-1}$$

(4)

where $t_{ij}$ represents the relation magnitude of factor $i$ on $j$ and $I$ is the identity matrix.

(5) Establish the cause-and-effect diagram. For each row and each column, calculate their sums respectively and obtain the row sum and column sum values $R={[r_i]}_{n\times 1}$ and $C={[c_i]}_{1\times n}$:

$$r_i={\displaystyle \sum _{j=1}^n{t}_{ij}}$$

(5)

$$c_i={\displaystyle \sum _{i=1}^n{t}_{ij}}$$

(6)

For each factor $i$, calculate its centrality degree $r_i+c_i$ and the cause degree $r_i-c_i$.

(6) Construct the cause–effect relationship diagram and analyze the results. Using the centrality degree $r_i+c_i$, interpret the influence of each factor within the system. For the cause degree $r_i-c_i$ [72]: if $r_i-c_i>0$, $i$ is primarily a cause/driving factor, exerting direct impacts on other factors; if $r_i-c_i<0$, $i$ is primarily an effect/outcome factor that has small impacts on other factors but can be influenced by other factors. In this basis, targeted measures and decision-making strategies can be developed to address the influence dynamics within the system.

(7) Weights determination. The prominence value of factor was calculated as $R+C$. The relative weight $w_i$ of each factor $i$ was then derived by normalizing its prominence value:

$$w_i={\displaystyle \frac{r_i+c_i}{{\displaystyle \sum _{i=1}^n(r_i+c_i)}}}$$

(7)

where $w_i$ represents the weight of factor $i$, and $n$ is the total number of factors. This procedure allows the factor weights reported in the Results section to be directly traced to the DEMATEL outputs.

(8) Threshold setting and inner dependence matrix. An inner dependence matrix helps to provide detailed information about the influence power of the certain factor [73]. This inner dependence matrix $\tilde{T}$ is obtained by removing elements with minor influences from the matrix $T$ with the threshold $\beta$ [40].

Firstly, a threshold value $\beta$ is set with Equation (8):

$$\beta ={\displaystyle \frac{{\displaystyle \sum _{j=1}^{n}su{m}_j}}{n\times n}}$$

(8)

where $\beta$ is the threshold and $su{m}_j$ is determined by:

$$su{m}_j={\displaystyle \sum _{i=1}^n{t}_{ij}}$$

(9)

Secondly, subtract the remaining elements in matrix $\tilde{T}$ into three sections with a same size ranging from the lowest to highest values, resulting in three categories of interaction influences, namely Category (a; [0.19, 0.25); weak influence; gray), Category (b; [0.25, 0.32); moderate influence; blue), Category (c; [0.32, 0.39]; strong influence; purple).

After that, the obtained results can be used for robustness analysis, referring to the discussion about the causal relationships among factors (see Section 4.1).

2.2.4. ISM Methodology

The ISM application process involves several key steps that structure the factors involved and represent their relationships clearly. These steps are as displayed as follows [39,72,74].

(1) Identify a contextual relationship and construct the structural self-interaction matrix (SSIM). Experts provide inputs on the influences between factors $i$ and $j$ with symbolic notations $O$, $V$, $A$, $X$ to describe the relationships. Where $O$ reflect no relation between factors $i$ and $j$, $V$ reflect that $i$ affects $j$ but not in both directions, $A$ represents that $j$ affects $i$ but not in both directions, and $X$ represents that $i$ and $j$ affect each other in both directions.

(2) Obtain a reachability matrix based on the SSIM constructed. The symbolic notations $O$, $V$, $A$, $X$ in SSIM should be converted into binary values (1 or 0) with the following rules:

$$O: \mathrm{Set} (i,j)=0 \mathrm{and} (j,i)=0$$

(10)

$$V: \mathrm{Set} (i,j)=1 \mathrm{and} (j,i)=0$$

(11)

$$A: \mathrm{Set} (i,j)=0 \mathrm{and} (j,i)=1$$

(12)

$$X: \mathrm{Set} (i,j)=1 \mathrm{and} (j,i)=1$$

(13)

Thus, the initial reachability matrix $M$ can be constructed, and it should satisfy conditions $M+I=M$ and $M^2=M$, where $I$ is the identity matrix [75]. Meanwhile, the final reachability matrix should be checked for transitivity, namely if factor $A$ affects factor $B$ and $B$ affects $C$, then $A$ affects $C$ necessarily.

(3) Use the reachability matrix to perform level partitions. First, identity the reachability set (RS) and the antecedent set (AS) for each factor. For factor $i$, if its reachability $R{S}_i$ equals to its intersection $R{S}_i\cap A{S}_i$, then $i$ is assigned the top level and removed from further iterations. Repeat until a complete hierarchical structure is built.

(4) Draw the ISM model and review the hierarchical structure. Once the reachability matrix is analyzed and level partitions are established, the hierarchical structure can be drawn. Noteworthy, modifications are necessarily for conceptual inconsistency and logical errors [39].

2.2.5. Robustness Procedures

To enhance methodological transparency, this study employed three robustness checks after the baseline DEMATEL-ISM analysis.

First, to further discuss the inner dependence among factors, a correction analysis is conducted for robustness analysis based on threshold setting and inner dependence matrix procedures in Section 2.2.3. The results are as discussed in Section 4.1.

Second, a comparative analysis was conducted by comparing the proposed integrated DEMATEL-ISM results with the study whose data input of ISM were derived from the evaluation result of DEMATEL, in order to assess whether the combined framework provided consistent yet more structured insights into the causal relationships among factors. The details and results are as illustrated in Section 4.2.

Third, a sensitivity analysis was performed by varying the weights assigned to individual experts and re-running the DEMATEL procedure under multiple scenarios. The resulting changes in factor weights and cause–effect rankings were then compared with the baseline case to evaluate the stability of the findings. The details and results are as shown in Section 4.3.

3. Results

3.1. Causal Relationships Among Factors Using DEMATEL

This section determines the role of each factor and classify them into the cause-and-effect groups. First, the experts assessed the pairwise interactions among factors through structured questionnaires. The evaluation was guided by the linguistic terms and corresponding scales presented in

Table 4. Direct-relation matrix.

	CS1	CS2	CS3	CS4	BC1	BC2	BC3	EC1	EC2	EC3	IC1	IC2	IC3
CS1	0	2.17	1.67	0.83	1.33	1.5	1.33	2.5	1	2.5	2.33	0.83	0.67
CS2	1.67	0	0.83	0.33	1.67	1.83	2.17	0.83	0.83	0.5	0.5	0.67	0.17
CS3	2.33	1.83	0	0.33	0.67	1.17	1.67	1.17	1.17	1.83	1.83	0.33	1
CS4	1.83	0.83	0.33	0	0.83	1.33	1.17	0.83	1	1.17	1	0.67	0.5
BC1	1.33	1	0.5	0.67	0	1.67	0.67	1	1.17	0.83	0.17	0.33	1
BC2	1.33	1.33	0.67	0.67	1.33	0	1.33	1	0.67	1	0.67	0.33	1.33
BC3	0.83	2	1	0.17	0.67	1.33	0	1.17	0.83	1	0.83	0.67	1.17
EC1	2	1.17	0.5	0.17	0.83	1.5	1.5	0	1.33	1.67	1.33	0.83	0.67
EC2	0.83	0.5	1	0.5	0.83	1.17	0.83	1.17	0	0.33	0.17	0.5	0.5
EC3	2.5	0.5	1	0.83	0.67	1	1.5	1.67	0.33	0	2.17	0.83	0.83
IC1	2.67	0.83	1.5	0.83	0.33	1	1.5	1.67	0.5	2.67	0	2	1.5
IC2	0.83	0.67	0.33	0.17	0.33	0.5	0.67	0.83	0.17	1.17	2.17	0	0.67
IC3	0.83	1	1.17	0.33	0.33	1.17	1.5	1.83	1	1.83	1.83	0.83	0

. The average scores were then calculated to construct the direct-relation matrix $A$ (see Table 4). After that, the matrix was normalized to obtain the normalized matrix $O$ and the total-relation matrix $T$ (see

Table 5. Total-relation matrix.

	CS1	CS2	CS3	CS4	BC1	BC2	BC3	EC1	EC2	EC3	IC1	IC2	IC3	${\mathit{r}}_{\mathit{i}}$
CS1	0.28	0.31	0.24	0.13	0.22	0.29	0.30	0.35	0.19	0.37	0.33	0.18	0.18	3.38
CS2	0.25	0.14	0.14	0.07	0.18	0.23	0.25	0.19	0.14	0.18	0.16	0.12	0.11	2.15
CS3	0.35	0.27	0.14	0.10	0.16	0.25	0.28	0.26	0.18	0.30	0.28	0.13	0.18	2.87
CS4	0.26	0.17	0.12	0.06	0.14	0.20	0.20	0.19	0.14	0.21	0.19	0.12	0.12	2.12
BC1	0.21	0.17	0.12	0.08	0.09	0.21	0.16	0.18	0.14	0.17	0.13	0.09	0.13	1.87
BC2	0.24	0.20	0.14	0.09	0.16	0.14	0.21	0.20	0.13	0.20	0.17	0.10	0.16	2.14
BC3	0.21	0.23	0.15	0.07	0.13	0.21	0.15	0.20	0.13	0.20	0.18	0.12	0.15	2.12
EC1	0.30	0.21	0.15	0.08	0.15	0.23	0.24	0.17	0.17	0.26	0.23	0.14	0.14	2.48
EC2	0.16	0.12	0.12	0.07	0.11	0.16	0.15	0.16	0.07	0.13	0.11	0.08	0.09	1.52
EC3	0.34	0.19	0.18	0.11	0.15	0.22	0.26	0.27	0.13	0.20	0.29	0.15	0.16	2.66
IC1	0.39	0.24	0.23	0.13	0.15	0.25	0.29	0.31	0.16	0.36	0.22	0.23	0.21	3.16
IC2	0.18	0.14	0.10	0.06	0.09	0.13	0.15	0.16	0.08	0.19	0.22	0.07	0.11	1.67
IC3	0.25	0.20	0.18	0.08	0.13	0.22	0.25	0.27	0.16	0.27	0.26	0.15	0.11	2.52
$c_i$	3.415	2.59	2.01	1.12	1.85	2.74	2.90	2.91	1.81	3.04	2.76	1.67	1.86

). For the total-relation matrix $T$, the sum of each row is denoted by $r_i$ representing the total influence dispatched by factor $i$, whereas the sum of each column is denoted by $c_i$, representing the total influence received by factor $i$. Based on the above calculations, the centrality degree $r_i+c_i$ and cause degree $r_i-c_i$ were derived for each factor. These results were then used to construct the cause–effect relationship map, as presented in

Table 6. Cause–effect relationship degree.

Factors	$\mathit{R}$	$\mathit{C}$	$\mathit{R}+\mathit{C}$	$\mathit{R}-\mathit{C}$	Identity
CS1	3.39	3.42	6.80	−0.03	Effect
CS2	2.15	2.58	4.74	−0.43	Effect
CS3	2.87	2.01	4.88	0.86	Cause
CS4	2.12	1.12	3.24	1.00	Cause
BC1	1.87	1.85	3.72	0.02	Cause
BC2	2.14	2.74	4.87	−0.60	Effect
BC3	2.12	2.89	5.02	−0.77	Effect
EC1	2.48	2.91	5.38	−0.43	Effect
EC2	1.52	1.81	3.33	−0.29	Effect
EC3	2.66	3.04	5.69	−0.38	Effect
IC1	3.16	2.75	5.91	0.40	Cause
IC2	1.67	1.67	3.34	0.00	Cause–effect balanced
IC3	2.52	1.87	4.39	0.66	Cause

and Figure 2. In addition, the factors were ranked according to their relative influence, with the results shown in Figure 3.

(1)

Cause group

Among 13 factors affecting HREI development in rural China, five are recognized as cause factors: technological collaboration barriers among stakeholders (CS3), complicated rural environment (CS4), sustainability of government subsidy policies (BC1), information coupling mechanism (IC1), professional talents (IC3).

Among these, CS4 has the highest $r_i-c_i$ value, indicating that the complicated rural environment exerts the most significant influence on other factors in deploying renewable energy integration. The rural environment in China presents unique challenges for energy integration due to heterogeneous energy resource availability, underdeveloped infrastructure, prevalent energy poverty, and so on [52]. Many rural areas are located in mountainous, desert, or coastal zones where access roads, transmission lines, and maintenance services are difficult to establish. Harsh weather conditions and the diversity of available renewable resources (e.g., biomass in northeast regions, solar in the northwest, small hydropower in the southwest) increase stakeholder’s concerns. This highlights the critical role of carefully addressing environmental complexities in early stages of developing a HREI project.

The next most influential factor is technological collaboration barriers among stakeholders (CS3). Rural HREI projects inherently require coordinated technological efforts among diverse actors, including government agencies, renewable energy developers, grid operators, equipment suppliers, and local communities. Successful implementation of such a complex system depends on the active involvement of all stakeholders while simultaneously addressing a wide range of technical considerations [76]. For example, renewable energy developers may adopt proprietary technologies that are incompatible with local grid infrastructure, leading to costly retrofits or delays. Thus, knowledge-sharing mechanisms are also underdeveloped, which restricts the transfer of best practices between pilot projects and new deployments, making technological collaboration essential for system integration and efficiency.

Similarly, professional talents (IC3) under the information coupling dimension demonstrates considerable influence, emphasizing the need for skilled professionals to manage complex technical and operational aspects of HREI projects. Many rural projects depend on technicians and engineers from urban centers, this talent gap extends to project management, where limited local expertise in financial planning, stakeholder engagement, and regulatory compliance can impede successful deployment. Building local capacity through targeted vocational training, university–industry partnerships, and incentives for skilled professionals to work in rural areas is therefore critical to ensuring long-term operational stability and local economic benefits.

Efficient integration of diverse renewable energy sources relies heavily on timely and accurate information coupling mechanism (IC1) among stakeholders. In rural China, digital infrastructure is often underdeveloped compared with urban areas, resulting in poor real-time monitoring and inadequate predictive maintenance. Fragmented data ownership, wherein local governments and enterprises and other entities maintain separate datasets, further limits the capacity for integrated energy management. Therefore, strengthening the information coupling mechanism through shared platforms, interoperable monitoring systems, and standardized data protocols could significantly improve coordination, reduce operational inefficiencies.

Lastly, BC1 (Sustainability of government subsidy policies) stands out as a vital factor supporting the broader HREI framework. Government subsidies have historically been a primary driver of renewable energy adoption in China’s rural areas. However, uncertainty in subsidy duration, inconsistent regional implementation, and abrupt phase-outs (often referred to as “policy cliffs”) threaten long-term project viability. This instability is particularly critical in rural HREI, where project payback periods tend to be longer due to higher logistics and maintenance costs. Policy sustainability is therefore essential not only for investor confidence but also for ensuring continuous community benefits and avoiding stranded assets.

(2)

Effect group

Based on negative $r_i-c_i$ values, seven factors are identified as effect factors, meaning they are primarily affected by the cause factors affecting HREI in rural China. These factors include synergy degree among multiple agents (CS1), life-cycle cost–benefit performance (CS2), financing access (BC2), construction and O&M costs (BC3), energy management and integration efficiency (EC1), rural power grid upgrade (EC2), coordinated management of key resources (EC3). Among them, BC3 has the lowest value (−0.771), indicating the highest dependency on cause factors, followed by BC2 (−0.602). The remaining factors, in descending absolute values, are CS2 (−0.433), EC1 (−0.426), EC3 (−0.381), EC2(−0.293), and CS1(−0.029).

The results highlight several important insights. First, construction and O&M costs (BC3) remain the most sensitive barrier for rural HREI projects, reflecting the higher logistical, labor, and maintenance expenses in dispersed and remote areas compared to urban settings. These costs are strongly influenced by upstream factors such as technological collaboration barriers and availability of skilled personnel. Second, financing access (BC2) is the second most affected factor, indicating that rural projects often struggle to secure sufficient capital due to long payback periods, policy uncertainties, and higher perceived investment risks. Innovative and available financing pathways should be developed for HREI in rural context, such as community-based energy cooperatives and blended finance schemes. Third, life-cycle cost–benefit performance (CS2) and energy management and integration efficiency (EC1) in rural HREI projects face more significant challenges compared with HREI deployment projects in urban areas. Fourth, coordinated management of key resources (EC3) depends heavily on the coordination degree among multiple stakeholders. As rural HREI systems typically combine biomass, solar, wind, and small hydropower, the effective allocation and scheduling of these resources requires both technical coordination and institutional alignment. Fifth, while rural power grid upgrade (EC2) is strongly constrained by environmental and infrastructural constraints. Many Chinese rural regions lack adequate transmission and distribution capacity to accommodate variable renewable generation, which can result in curtailment and inefficiencies. Finally, synergy degree among multiple agents (CS1) shows relatively low dependency in absolute terms, it remains a foundational enabler for all other effect factors. High synergy levels can mitigate financing barriers, improve cost–benefit outcomes, and enhance integration efficiency. Notably, supervision and service (IC2) demonstrates a balanced influence, with its total influence degree equal to the extent it is influenced by other factors. This suggests a dual role in the system, acting both as a significant influencer and a receiver of influence from others.

(3)

Weights of factors

The weight assignment results reveal that CS1 (synergy degree among multiple agents) is the most critical factor, with a weight of 0.111, indicating its substantial impact on hybrid renewable energy integration in rural China. This highlights the essential role of fostering collaboration among various stakeholders. The second and third most significant factors are IC1 (information coupling mechanism) and EC3 (coordinated management of key resources), with weights of 0.096 and 0.093, respectively. This finding suggests that enhancing information sharing and resource coordination should be prioritized to ensure smooth HREI implementation. Following these are energy management and integration efficiency (EC1) with a weight of 0.088 and construction and O&M costs (BC3) at 0.082. On the one hand, stakeholders must consider construction and operational costs when making decisions about HREI. On the other hand, efficient energy management and system integration directly influence project deployment. The remaining factors and their corresponding weights are CS3 (technological collaboration barriers among stakeholders, 0.08), BC2 (financing access, 0.079), CS2 (life-cycle cost–benefit performance, 0.077), IC3 (professional talents, 0.072), BC1(sustainability of government subsidy policies, 0.061), IC2 (supervision and service, 0.054), CS4 (complicated rural environment, 0.053). These results underscore the multidimensional nature of HREI development. Factors related to stakeholder collaboration, resource management, and financial considerations should be prioritized to achieve efficient and sustainable integration of renewable energy in rural China.

3.2. Hierarchical Structure of Factors Responsible Using ISM

The ISM approach systematically establishes the direct or indirect hierarchical structure of complex factors, incorporating multiple stakeholders’ concerns. Following the calculation process of ISM approach, the approach was applied to address the identified factors influencing HREI in rural China. Experts’ opinions on the relationship categories between two factors were collected with letters $O$, $V$, $A$, $X$ (indicating uncorrelation, direct correlation, or intercorrelation) as the basis for constructing the SSIM matrix, as

Table 7. Structural self-interaction matrix (SSIM).

	CS1	CS2	CS3	CS4	BC1	BC2	BC3	EC1	EC2	EC3	IC1	IC2	IC3
CS1	O	O	A	A	O	V	O	O	A	A	A	O
CS2	X	O	O	O	A	X	A	X	A	A	O
CS3	O	O	V	V	O	V	X	X	O	O
CS4	O	O	V	V	V	O	V	V	O
BC1	O	O	O	O	V	V	O	V
BC2	O	O	O	X	O	X	O
BC3	O	O	A	A	A	A
EC1	A	O	A	X	O
EC2	O	A	O	O
EC3	A	O	X
IC1	A	O
IC2	O
IC3

Note: V indicates that factor i influences factor j but not in both directions; A indicates that factor j influences factor i but not in both directions; X indicates mutual influence between factors i and j in both directions; O indicates no relation between the two factors. Only the upper triangular relationships are reported. These symbolic relations were converted into binary values (1/0) according to Equations (10)–(13) to obtain the reachability matrix.

illustrates. Derived from the SSIM matrix, Equation (10)–(13) outlined in the methodology section were used to convert qualitative descriptions into corresponding binary system form (1 s or 0 s), resulting in the initial reachability matrix. Transitivity analysis was then applied to generate the final reachability matrix (see

Table 8. Final reachability matrix.

	CS1	CS2	CS3	CS4	BC1	BC2	BC3	EC1	EC2	EC3	IC1	IC2	IC3
CS1	1	1	0	0	0	1 *	1 *	1	0	1 *	0	0	1 *
CS2	1	1	1 *	0	0	1	1 *	1	0	1 *	1 *	0	1
CS3	1	1 *	1	0	0	1	1	1	0	1	1	0	0
CS4	1	1	1 *	1	0	1	1	1 *	1	1	1	0	1 *
BC1	1	1	1 *	0	1	1	1 *	1	1	1 *	0	0	1 *
BC2	1 *	1	1	0	0	1	1 *	1	0	1	1 *	0	1 *
BC3	1 *	1	1	0	0	1 *	1	1 *	0	1 *	1 *	0	1 *
EC1	1 *	1	1 *	0	0	1	1	1	0	1	1 *	0	1 *
EC2	1 *	1	1 *	0	0	1 *	1	1 *	1	0	0	0	1 *
EC3	1	1 *	1 *	0	0	1	1	1	0	1	1	0	0
IC1	1	1 *	1 *	0	0	1 *	1	1	0	1	1	0	0
IC2	0	1 *	0	0	0	0	1 *	0	1	0	0	1	0
IC3	1 *	1	0	0	0	1 *	1 *	1	0	1	1	0	1

1 This table is obtained by processing transitivity analysis on initial reachability matrix, with * entries indicating reachability.

), wherein 1* represents the addition by transitivity to reduce the biases and errors [74].

After constructing the final reachability matrix, level partitioning is carried out to define the reachability set (RS) and the antecedent set (AS). The reachability set for a given factor includes the factor itself and all other factors it influences. The antecedent set contains the factor along with those factors that influence it. The intersection of these two sets is then calculated to determine and rank the level of each factor. Level partitioning continues until all factors are assigned a level. The level partitioning results are presented in

Table 9. Final level partitioning.

Factors	Reachability Set	Antecedents Set	Intersection Set	Level
CS1	1, 2, 6, 7, 8, 10, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13	1, 2, 6, 7, 8, 10, 13	I
CS2	1, 2, 3, 6, 7, 8, 10, 11, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 6, 7, 8, 10, 11, 13	I
CS3	1, 2, 3, 6, 7, 8, 10, 11	2, 3, 4, 5, 6, 7, 8, 9, 10, 11	2, 3, 6, 7, 8, 10, 11	II
CS4	1, 2, 3, 4, 6, 7, 8, 9, 10, 11, 13	4	4	V
BC1	1, 2, 3, 5, 6, 7, 8, 9, 10, 13	5	5	V
BC2	1, 2, 3, 6, 7, 8, 10, 11, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13	1, 2, 3, 6, 7, 8, 10, 11, 13	I
BC3	1, 2, 3, 6, 7, 8, 10, 11, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 6, 7, 8, 10, 11, 13	I
EC1	1, 2, 3, 6, 7, 8, 10, 11, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 13	1, 2, 3, 6, 7, 8, 10, 11, 13	I
EC2	1, 2, 3, 6, 7, 8, 9, 13	4, 5, 9, 12	9	IV
EC3	1, 2, 3, 6, 7, 8, 10, 11	1, 2, 3, 4, 5, 6, 7, 8, 10, 11, 13	1, 2, 3, 6, 7, 8, 10, 11	I
IC1	1, 2, 3, 6, 7, 8, 10, 11	2, 3, 4, 6, 7, 8, 10, 11, 13	2, 3, 6, 7, 8, 10, 11	II
IC2	2, 7, 9, 12	12	12	V
IC3	1, 2, 6, 7, 8, 10, 11, 13	1, 2, 4, 5, 6, 7, 8, 9, 13	1, 2, 6, 7, 8, 13	III

, reveal a five-level hierarchy for the 13 factors affecting HREI in rural China. Figure 4 illustrates the hierarchical structure and relationships.

As illustrated in Table 9 and Figure 4, the factors complicated rural environment (CS4), sustainability of government subsidy policies (BC1), supervision and service (IC2), belong to dimensions complexity of the system (CS), benefit coordination (BC), and information coupling (IC), respectively, emerge as the most substantial and independent drivers affecting HREI development in rural areas. They show impacts on the factor rural power grid upgrade (EC2) under the dimension of efficiency coordination (EC) directly. While these factors playing a driving role in the rural HREI system, they can potentially influence the system efficiency. Therefore, these factors warrant priorities to present the maximum impacts, and stakeholders should adopt corresponding strategies based on these factors while maintaining attentions on other factors they influence.

Within the hierarchy, professional talents (IC3) are positioned at the third level and technological collaboration barriers among stakeholders (CS3) and information coupling mechanism (IC1) are positioned on the second level. The introduction of professional talents who have cross-sector knowledge and rich experience perform well in addressing technological collaboration barriers among multiple components and understand the underlying information coupling rules among different sectors. Meanwhile, these two play pivotal bridging roles, exerting direct influence over multiple downstream variables, including synergy degree among multiple agents (CS1), life-cycle cost–benefit performance (CS2), financing access (BC2), construction and O&M costs (BC3), energy management and integration efficiency (EC1), coordinated management of key resources (EC3). This hierarchical positioning highlights those improvements in stakeholder collaboration and information coupling are likely to yield cascading benefits across technical, economic, and operational dimensions of rural HREI systems.

4. Discussion

4.1. Correlation Analysis Among Factors

Apply the threshold setting and robustness procedures in Section 2.2.3, the threshold value can be calculated as $\beta =0.18$. Then, the inner dependence matrix can be obtained and categorized by removing elements with the threshold, as

Table 10. Inner dependence matrix.

	CS1	CS2	CS3	CS4	BC1	BC2	BC3	EC1	EC2	EC3	IC1	IC2	IC3
CS1	0.28	0.31	0.24		0.22	0.29	0.30	0.35	0.19	0.37	0.33
CS2	0.25					0.23	0.25	0.19
CS3	0.35	0.27				0.25	0.28	0.26		0.30	0.28
CS4	0.26					0.20	0.20	0.19		0.21	0.19
BC1	0.21					0.21
BC2	0.24	0.20					0.21	0.20		0.20
BC3	0.21	0.23				0.21		0.20		0.20
EC1	0.30	0.21				0.23	0.24			0.26	0.23
EC2
EC3	0.34	0.19				0.22	0.26	0.27		0.20	0.29
IC1	0.39	0.24	0.23			0.25	0.29	0.31		0.36	0.22	0.23	0.21
IC2										0.19	0.22
IC3	0.25	0.20				0.22	0.25	0.27		0.27	0.26

1 Interaction influence categories: Category (a; [0.19, 0.25); weak influence; gray), Category (b; [0.25, 0.32); moderate influence; blue), Category (c; [0.32, 0.39]; strong influence; purple).

presents.

Considering the correlations among factors, interpretations emerge that deepen our understanding of how rural HREI deployment unfolds.

On one hand, IC1 appears to exert the largest and strongest influence over other factors. Its high prominence suggests that HREI implementation depends not only on physical energy integration, but also on the integration of information flows across actors and subsystems. In rural settings, information asymmetry between various stakeholders can delay decision-making, reduce trust, and weaken operational responsiveness, and a weak information coupling mechanism therefore affects both technical performance and stakeholder coordination since it limits real-time adjustment, undermines problem feedback, and increases the risk of fragmented management. Similarly, as discussed in the literature [61], an effective information coupling of data consist of not only measurements but also non-electrical data will revolutionize the operation of the power system.

On the other hand, CS1, namely synergy degree among multiple agents, also shows strong influence over many other factors (it affects nine others). This indicates that stakeholder coordination is not simply a complementary condition, but a core mechanism through which rural HREI becomes implementable. This finding resonates with the concept of stakeholder synergy from the study by Tantalo and Priem (2016) [77]. Rural hybrid systems usually involve multiple actors with different objectives, stronger synergy can reduce transaction costs, improve joint decision-making, and facilitate resource sharing across the project life cycle. This explains why CS1 has such high overall importance: it acts as a system-level enabler that conditions the effectiveness of many other factors.

4.2. Comparative Analysis

A comparative analysis is conducted to observe the robustness and reliability of the findings. We compare the proposed method with the methodology of He et al., (2021) [78], whose data input of ISM were derived from the evaluation result of DEMATEL. In this case, the total-relation matrix (see Table 6) is processed with further calculation to generate the reachable matrix for conducting ISM. The results of ISM are shown in

Table 11. Level partitions of the reachability matrix.

Factors	Reachability Set	Antecedents Set	Intersection Set	Level
CS1	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
CS2	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
CS3	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
CS4	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	4	4	II
BC1	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
BC2	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
BC3	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
EC1	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
EC2	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
EC3	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
IC1	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
IC2	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I
IC3	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13	1, 2, 3, 5, 6, 7, 8, 9, 10, 11, 12, 13	I

. Only two levels are obtained and only factor CS4 (complicated rural environment) is identified as the independent driver. Therefore, collecting data with a separate way for DEMATEL and ISM helps to gather more detailed information of expert evaluations and the results are more applicable and precise.

4.3. Sensitivity Analysis

Following Garg (2021) [79] and Feldmann et al. (2022) [40], this paper conducted a sensitivity analysis by assigning relatively higher weights to each of the six experts (E1–E6) in turn, while keeping the weights of the remaining experts equal within each scenario. This generated six investigation scenarios, denoted as $I{S}_i(i=1,2,\cdots ,6)$. In each scenario, one designated expert received a higher weight (2/6), and the remaining five experts shared the residual weight equally. The DEMATEL procedure was repeated for each scenario, and the resulting factor weights and cause–effect rankings were compared with the baseline results (see

Table 12. Criteria weights of sensitivity analysis.

Factors	Criteria Weights
	$\mathit{I}{\mathit{S}}_1$	$\mathit{I}{\mathit{S}}_2$	$\mathit{I}{\mathit{S}}_3$	$\mathit{I}{\mathit{S}}_4$	$\mathit{I}{\mathit{S}}_5$	$\mathit{I}{\mathit{S}}_6$
CS1	0.12	0.10	0.11	0.11	0.10	0.12
CS2	0.07	0.08	0.07	0.08	0.08	0.07
CS3	0.07	0.08	0.09	0.08	0.08	0.07
CS4	0.05	0.06	0.05	0.05	0.05	0.05
Dimension weights of CS	0.32	0.32	0.32	0.32	0.32	0.32
BC1	0.06	0.06	0.06	0.06	0.06	0.06
BC2	0.07	0.08	0.08	0.09	0.08	0.07
BC3	0.08	0.09	0.08	0.08	0.08	0.08
Dimension weights of BC	0.21	0.22	0.23	0.23	0.22	0.21
EC1	0.09	0.09	0.08	0.08	0.09	0.09
EC2	0.05	0.06	0.05	0.05	0.06	0.05
EC3	0.10	0.09	0.10	0.09	0.09	0.10
Dimension weights of EC	0.24	0.24	0.23	0.23	0.24	0.24
IC1	0.10	0.09	0.10	0.10	0.09	0.10
IC2	0.06	0.05	0.05	0.05	0.06	0.06
IC3	0.07	0.07	0.07	0.07	0.07	0.07
Dimension weights of IC	0.23	0.22	0.22	0.22	0.22	0.23

and

Table 13. Cause–effect relationship degrees (a).

Factors	$\mathit{I}{\mathit{S}}_1$		$\mathit{I}{\mathit{S}}_2$		$\mathit{I}{\mathit{S}}_3$		$\mathit{I}{\mathit{S}}_4$		$\mathit{I}{\mathit{S}}_5$		$\mathit{I}{\mathit{S}}_6$
	$\mathit{R}-\mathit{C}$	Rank	$\mathit{R}-\mathit{C}$	Rank	$\mathit{R}-\mathit{C}$	Rank	$\mathit{R}-\mathit{C}$	Rank	$\mathit{R}-\mathit{C}$	Rank	$\mathit{R}-\mathit{C}$	Rank
CS1	−0.02	7	−0.05	7	−0.03	7	−0.21	7	0.11	5	−0.02	7
CS2	−0.30	11	−0.82	12	−0.39	10	−0.38	9	−0.51	10	−0.30	11
CS3	0.60	2	0.98	2	0.78	2	1.41	1	0.96	2	0.60	2
CS4	0.69	1	1.43	1	0.92	1	1.28	2	1.19	1	0.69	1
BC1	0.01	5	−0.01	6	0.00	6	0.28	4	−0.19	7	0.01	5
BC2	−0.41	12	−0.63	10	−0.54	12	−0.96	12	−0.74	12	−0.41	12
BC3	−0.53	13	−0.98	13	−0.70	13	−1.26	13	−0.75	13	−0.53	13
EC1	−0.29	10	−0.66	11	−0.41	11	−0.47	10	−0.51	11	−0.29	10
EC2	−0.21	8	−0.41	8	−0.27	8	−0.38	8	−0.31	8	−0.21	8
EC3	−0.27	9	−0.60	9	−0.34	9	−0.55	11	−0.36	9	−0.27	9
IC1	0.27	4	0.89	3	0.37	4	0.27	5	0.46	4	0.27	4
IC2	0.00	6	0.16	5	0.01	5	−0.14	6	−0.08	6	0.00	6
IC3	0.47	3	0.71	4	0.61	3	1.09	3	0.73	3	0.47	3

). The purpose of this procedure was to examine whether moderate changes in expert influence would substantially alter the ranking structure or causal classification of the factors.

The sensitivity analysis shows that the criteria weights across different scenarios exhibit only slight fluctuations, with the overall distribution remaining consistent, suggesting that our factor model is structurally robust. The largest variation is observed for factor C1 (ranging from 0.12 to 0.10), while the smallest occurs for IC3 (remaining constant at 0.07). Notably, weights of BC1 and IC3 remain unchanged across all scenarios, indicating that the factor system is stable and robust. Similarly, the aggregated weights of the four dimensions also remain stable.

The consistent dominance of certain cause factors (CS4, CS3, IC3, IC1, BC1) across scenarios underlines their centrality in shaping HREI outcomes in rural China. The persistence of these factors underscores their central role in shaping HREI outcomes in rural China. (1) The prominence of CS4 (complicated rural environment) aligns with the existing literature [18,42,43,80], its strong causal role can be explained by its function as a root implementation condition rather than a single isolated barrier. Rural environment shapes the operating context within which many other factors function. This helps explain why CS4 appears as a foundational driver in the hierarchical structure since it does not merely create technical difficulty, but also amplifies institutional and managerial complexity across the HREI system. (2) The standing of CS3 (technological collaboration barriers among stakeholders) is consistent with green innovation theory: collaboration among heterogeneous stakeholders is critical to absorb the broad knowledge base needed for innovation, especially in green or renewable projects [81,82]. (3) The importance of professional talents (IC3) is likewise well-founded: human capital has been proven one of the significant factors that has been seen as a primary driver of economic growth for decades and improved human capital enhances operational effectiveness, enabling better management and operation of renewable energy projects [83]. (4) On the effect side, factors such as life-cycle cost–benefit performance, financing access, construction and O&M costs persistently cluster downstream as dependent outcomes. Across all scenarios, each factor remains within their respective groups, confirming that the cause–effect structure of the model is fundamentally stable. Results also indicate that variations in expert weighting have no significant impact on the interpretation of factor relationships, with only minor changes observed.

4.4. Implications

(1)

Theoretical implications

From a theoretical perspective, the research extends stakeholder-oriented analysis into the underexplored domain of rural HREI by systematically identifying and structuring critical influencing factors. Rather than claiming full empirical representation of all stakeholder groups, the study uses expert-informed judgments to examine how technical, institutional, and coordination-related factors interact within the rural HREI context. The theoretical implications of the paper are twofold. In one hand, by conceptualizing HREI as a complex, interdependent system, the study moves beyond the predominantly technical focus of the existing literature to integrate socio-institutional dimensions. On the other hand, this paper makes contributions by deepen the understanding of interactions among various influencing factors in rural HREI development. By using the integrated DEMATEL-ISM method in the field, various relationships among individual potential factors have been uncovered and discussed. This is done by further evaluating the influence degree of certain factors and addressing the “what” and “how” research questions.

(2)

Managerial implications

From a managerial perspective, rural HREI should be understood as a coordination-intensive system rather than a purely technical deployment task. The results suggest that HREI performance depends not only on technology choice, but also on how environmental conditions, policy support, stakeholder coordination, information exchange, and resource management interact in practice. In particular, the high importance of synergy among multiple agents indicates that fragmented responsibilities and weak incentive alignment can easily undermine implementation. Likewise, factors such as a complicated rural environment, subsidy sustainability, supervision and service, and rural grid upgrade function as upstream enablers because they shape multiple downstream outcomes simultaneously.

These findings imply that managerial intervention should focus on system-level mechanisms rather than isolated symptoms. Information coupling and coordinated management of key resources should be treated as core operational conditions. At the same time, downstream economic problems such as financing difficulty, low life-cycle cost–benefit performance, and high construction and O&M costs should be interpreted as signals of deeper structural constraints rather than as stand-alone issues. More broadly, because the rural environment exerts strong influence on implementation, managerial strategies should remain context-sensitive with HREI project design, support policies, and service arrangements.

Although this study is grounded in the rural Chinese context, its implications are not limited to China alone. Some factors are closely linked to China-specific conditions, such as government-led policy support, rural revitalization strategies, and subsidy sustainability. At the same time, other factors, including synergy degree among multiple agents, financing access, construction and O&M costs, energy management and integration efficiency, coordinated management of key resources, etc., are also relevant to rural renewable energy development in many other countries. Therefore, the framework proposed in this study should be understood as context-sensitive rather than China-exclusive. Although the relative importance and structural relationships of these factors may vary across countries, the framework may still provide a useful reference for identifying and adapting critical factors in other rural settings.

5. Conclusions

This study investigated the critical factors affecting hybrid renewable energy integration (HREI) in rural China by applying an integrated DEMATEL–ISM approach. A total of 13 critical factors were identified and organized into four dimensions. The findings indicate that rural HREI is shaped not only by technical conditions, but also by environmental complexity, institutional support, stakeholder coordination, and information-related mechanisms. This highlights the need to understand rural HREI as a complex system rather than a purely technical deployment task.

The study contributes by providing a structured framework for identifying and analyzing critical factors in the rural Chinese context and by demonstrating the usefulness of DEMATEL–ISM for revealing their causal and hierarchical relationships. From a practical perspective, the findings suggest that effective HREI development requires stronger upstream support, better stakeholder coordination, and more context-sensitive implementation strategies. The insights obtained offer a valuable reference for stakeholders and decision-makers in promoting HREI.

The authors acknowledge several limitations of this study. First, the expert panel was relatively small and did not fully represent all stakeholder groups directly involved in rural HREI implementation. Therefore, the findings should be interpreted as an exploratory, expert-informed structural analysis rather than as a nationally representative assessment. Second, this study adopts a cross-sectional perspective based on static expert evaluations and does not capture possible changes in factor weights, causal relationships, or structural positions over time. Third, the Delphi procedure in this study did not employ formal statistical convergence indicators, which could be incorporated in future research to improve methodological rigor. Last, some nuances in expert judgments may have been simplified during data collection.

Future research should expand the size and diversity of the expert panel, incorporate broader stakeholder participation, and examine how factor roles may evolve over time under changing policy, infrastructural, and rural conditions. In addition, more advanced uncertainty-handling methods could be introduced to further improve the robustness of the analysis.