Modeling Critical Success Factors for Green Energy Integration in Data Centers

Abstract

The rising energy demands of data centers combined with increasing sustainability requirements require the integration of green energy solutions. This study identifies and analyzes the Critical Success Factors (CSFs) that affect the effective adoption of green energy in data centers, addressing both technical and organizational challenges. Through a systematic literature review and expert validation, eleven key CSFs were identified, encompassing aspects such as energy efficiency, renewable energy integration, carbon neutrality strategies, regulatory compliance, and investment strategies. This study employs the Decision-Making Trial and Evaluation Laboratory (DEMATEL) methodology to explore the interdependencies among these factors, mapping causal relationships within the CSF network. The findings indicate that governance and organizational culture, along with investment strategy, are the most influential CSFs, driving the effectiveness of energy transition efforts. Carbon neutrality strategies serve as a crucial mediating factor, linking governance and financial commitment to operational sustainability outcomes. Additionally, renewable energy integration and regulatory compliance act as structural mediators, ensuring that governance-led sustainability initiatives result in practical, measurable impacts. These insights provide a strategic roadmap for data center operators, policymakers, and industry stakeholders to prioritize interventions that promote a resilient and sustainable energy transition. By clarifying the cause-and-effect dynamics within the green energy adoption framework, this study contributes to the development of more effective policies and investment strategies aimed at reducing the carbon footprint of data centers and enhancing their long-term sustainability.

1. Introduction

The rapid expansion of the data center industry, driven by the proliferation of cloud computing, artificial intelligence (AI), data-intensive applications, and digital services, has positioned data centers as one of the most energy-intensive infrastructures globally [1,2,3]. The global data center market, projected to grow from USD 242.72 billion in 2024 to USD 584.86 billion by 2032 [4], raises pressing concerns regarding energy sustainability. Data centers currently consume between 2% and 3% of global electricity [1,5], with projections indicating that their electricity consumption could surpass 20% of global demand by 2030 [6]. While this growing energy footprint raises sustainability concerns, part of the increased electricity consumption can be attributed to the shift from decentralized, often less efficient, local computing toward centralized cloud-based processing in data centers [7,8].

Global environmental concerns and regulatory pressures have accelerated the push toward greener data centers. International frameworks, including the International Energy Agency’s Net Zero by 2050 roadmap and the European Union’s Climate Neutral Data Centre Pact, emphasize the urgency of decarbonizing data center operations [9]. Leading technology companies, such as Google, Meta, and Apple, have responded by integrating renewable energy sources and enhancing energy efficiency in their data centers [6]. The industry’s shift toward sustainability is further driven by investor expectations, customer demands, and corporate social responsibility commitments [9].

Technological innovations are pivotal in advancing sustainable data center operations. Emerging solutions, such as modular data centers, AI-driven energy optimization, energy-aware computing, and advanced cooling systems, offer pathways to improve energy efficiency and reduce environmental impact [3,9,10]. Additionally, the integration of renewable energy sources like solar and wind power is gaining momentum [2,11,12]. However, this transition faces persistent challenges, including intermittency, high initial investment costs, and infrastructural limitations [1,2]. Addressing these obstacles requires comprehensive strategies involving advanced forecasting, demand-side flexibility, and energy storage optimization [1,2,13]. Beyond these technical hurdles, non-technical challenges further complicate the successful integration of green energy into data centers. These include the need for standardized carbon emission reduction guidelines, fragmented decision-making processes, procurement complexities, and organizational reluctance to adopt new technologies [10,14].

Achieving sustainable data center operations requires a holistic approach that integrates energy-efficient technologies, renewable energy utilization, intelligent infrastructure design, and strategic operational management [1,3,15]. Such an approach must balance economic, environmental, regulatory, and operational considerations while accommodating the industry’s evolving demands [1,16]. Comprehensive models that incorporate various sustainability components—including energy sources, cooling systems, hardware efficiency, and network architecture—are essential for assessing the overall environmental impact and guiding data center operators toward best practices [13].

Prior research has explored various factors influencing green energy integration in data centers, primarily focusing on technical and operational aspects, such as energy efficiency and performance optimization, real-time energy monitoring, dynamic demand response, workload optimization, smart grid integration, renewable energy integration, strategic renewable energy procurement, carbon neutrality strategies, and sustainable data center design [3,6,15,17,18]. A limited number of studies also address organizational and strategic factors, such as governance, organizational culture, leadership commitment, cross-functional collaboration, human resources skills, investment strategies, and vendors’ alignment with the organization’s green energy goals [14,19,20]. The above-mentioned studies have largely analyzed these factors in isolation, neglecting to account for their complex interconnections within a broader systemic context. As a result, there is a clear gap in studies that systematically identify and analyze cause-and-effect relationships among these factors.

To address this gap, this study conducts a comprehensive examination of the critical success factors (CSFs) for green energy integration in data centers. First, CSFs were identified through a systematic literature review and validated by industry experts to ensure their relevance and accuracy. Then, applying the Decision-Making Trial and Evaluation Laboratory (DEMATEL) methodology, this study mapped the interconnections among these CSFs and assessed their wider impacts. Thus, the primary research question guiding this study is as follows:

What are the levels of importance and relationships among the CSFs for green energy integration in data centers?

By clarifying these relationships, the findings provide policymakers and business leaders with essential insights to develop more effective, integrated, and innovative strategies. Ultimately, this contributes to advancing energy integration and promoting sustainable practices within the data center sector.

The remainder of this paper is organized as follows: Section 2 provides a comprehensive and systematic review of the existing literature on the CSFs for green energy integration in data centers. Section 3 details the DEMATEL methodology employed in this study, including information on sample selection and data collection procedures. Section 4 presents the empirical findings, emphasizing the identified cause-and-effect relationships among the CSFs. Finally, Section 5 discusses the research findings, explores managerial and societal implications, acknowledges this study’s limitations, and proposes directions for future research.

2. Literature Review

This section explores the CSFs for green energy integration in data centers based on a thorough review of the relevant literature. A structured three-stage research approach was used to build a comprehensive bibliographic dataset. The process began in early 2025 with a targeted search in the Scopus database, using the keywords “data center” AND (energy OR green OR sustainab*) in titles, along with (factors OR determinants OR drivers OR enablers OR barriers OR obstacles OR inhibitors OR challenges OR constraints) in titles, abstracts, or keywords. The inclusion of barrier-related terms was intentional, as reversing a barrier could reveal a corresponding CSF.

This initial search produced 765 articles, excluding non-English publications. Next, two co-authors independently conducted a qualitative assessment to identify studies specifically addressing CSFs for green energy integration in data centers, reducing the selection to 164 articles. In the final stage, purposive sampling was applied to focus on articles with unique contributions, resulting in a core set of 53 articles. The analysis of these articles identified 11 CSFs for green energy integration in data centers, which are discussed in the following subsections.

2.1. Operational Energy Efficiency and Monitoring

Operational energy efficiency and monitoring are important CSFs for integrating green energy into data centers [15,18,21,22]. Achieving this involves deploying high-efficiency servers and implementing advanced cooling technologies, such as liquid cooling, free cooling, and adsorption-based hybrid cooling systems, which significantly reduce energy and water consumption while optimizing thermal management [20,21,23,24]. Waste heat recovery systems further enhance cooling efficiency by repurposing excess heat for district heating or internal energy needs [16,21,22,25]. Combining these technologies with information technology (IT) virtualization and modular data center designs can improve server utilization and minimize idle energy consumption [15,17,26]. Real-time energy and thermal monitoring systems are vital for optimizing operational performance. These systems, supported by sensors and intelligent management platforms like Infrastructure Management for Data Centers and AI-driven control systems, enable dynamic workload scheduling, early fault detection, and immediate adjustments to energy consumption [3,5,9,11,18,20]. Metrics such as Power Usage Effectiveness and Water Usage Effectiveness are extensively used, although the literature notes that relying solely on these may be insufficient for comprehensive energy efficiency assessments [6,10,14]. Advanced monitoring solutions integrate data from energy flows, thermal conditions, and server utilization, facilitating continuous optimization and improved operational decision-making [3,27,28].

2.2. Supply and Demand Optimization

Supply and demand optimization plays a key role in integrating green energy into data centers by dynamically balancing computing workloads with grid signals [29,30,31,32]. This is achieved through a combination of demand response strategies, energy storage solutions, and intelligent workload scheduling [33,34,35,36,37]. Data centers use demand response programs to adjust computing loads in response to electricity price fluctuations and grid conditions, enhancing grid stability while reducing peak demand [29,31,32]. Workload management strategies, such as multi-task scheduling and job migration across geo-distributed data centers, enable the alignment of computational demand with renewable energy availability [1,17,30,38]. Techniques like dynamic power capping, virtualization-based workload migration, and “follow the sun” strategies further enhance energy efficiency by directing workloads to regions with abundant renewable energy sources [3,21,30]. Energy storage devices, including battery systems and thermal energy storage, play a complementary role by absorbing excess renewable energy during periods of low demand and discharging it during peak hours to smooth consumption profiles [23,33,36]. Advanced scheduling systems, such as GreenSlot and GreenHadoop, dynamically align workloads with solar and wind energy patterns, optimizing supply–demand balance [11,13]. Geographical load balancing and reinforcement learning-based scheduling also facilitate workload redistribution, ensuring data centers operate efficiently despite renewable energy intermittency [2,39,40]. Additionally, two-stage optimization frameworks, combining day-ahead and real-time scheduling, manage uncertainties in energy supply and workload demand, enabling flexible IT operations and load shifting [32,34,35]. These strategies reduce reliance on non-renewable energy sources and enhance operational resilience, lower energy costs, and support broader sustainability objectives within the data center industry [37,41,42,43].

2.3. Renewable Energy Integration

Renewable energy integration is a CSF for green energy adoption in data centers, ensuring a continuous, reliable, and clean power supply through both on-site and off-site solutions [9,20,21,30]. Data centers increasingly employ a combination of renewable sources—solar photovoltaics, wind turbines, hydroelectric power, and hydrogen energy storage—to reduce their carbon footprint [1,20,30]. Major operators such as Google, Amazon, and Microsoft have committed to 100% renewable energy through on-site generation and off-site procurement strategies, including power purchase agreements and renewable energy certificates [2,6,9,44]. On-site installations, such as solar arrays and wind turbines, are often co-located with data centers to minimize transmission losses and ensure a steady power supply [11,45,46]. Technological advancements, combined with government incentives, have increased the economic viability of integrating renewables, while hybrid solutions—combining solar, wind, and bioenergy—further strengthen supply reliability [15,18,21]. The incorporation of hydrogen fuel cells and renewable natural gas into on-site power generation systems exemplifies innovative approaches to secure continuous clean energy [20,25]. These integration strategies ensure sustainable data center operations and enhance long-term energy resilience and operational performance [27,40,43].

2.4. Carbon Neutrality Strategies

Carbon neutrality strategies are essential for integrating green energy into data centers, focusing on achieving net-zero emissions through a combination of targeted energy efficiency improvements, renewable energy adoption, and complementary measures. Energy efficiency enhancements play a pivotal role in reducing greenhouse gas emissions, with strategies such as workload scheduling, hybrid cooling systems, and the use of energy-efficient hardware contributing significantly to lower carbon footprints [1,15,17,24,42]. The adoption of renewable energy is central to carbon neutrality, with data centers increasingly sourcing power from wind, solar, and carbon-free electricity grids to replace fossil fuels [2,9,13,30,47,48,49]. Leading cloud providers like Amazon, Google, and Microsoft have committed to operating on 100% renewable energy, underscoring the industry’s shift towards sustainable operations [2,13,20,50]. Complementary measures, such as carbon offset programs, carbon credits, and renewable energy certificates, further support emission reduction targets [3,9,41]. Policy frameworks, including carbon taxation, trading systems, and incentives for low-carbon technology adoption, are instrumental in driving these initiatives [10,37]. Additionally, innovative approaches like waste heat utilization in district heating networks [16], embodied carbon assessments in data center design [20], and the integration of flexible fuel systems [25] enhance sustainability efforts. Advanced energy storage devices and demand response mechanisms enable better use of renewable sources, minimizing reliance on carbon-intensive brown energy [33,38,43]. In addition to operational emissions, a comprehensive carbon neutrality strategy must also account for the embodied carbon associated with the manufacturing, procurement, and transportation of data center hardware. These upstream emissions, though less visible, contribute significantly to the overall carbon footprint and are influenced by material choices, supplier practices, and logistics operations [9,20,46]. Emerging technologies, including AI-based resource management and spatio-temporal energy scheduling, offer significant emission reduction potential, achieving up to a 70% decrease in carbon outputs [40,41,48].

2.5. Sustainable Data Center Design and Infrastructure

Sustainable data center design and infrastructure are CSFs for integrating green energy into data centers, emphasizing eco-friendly construction, optimized layouts, and strategic site selection [6,9,20,22]. Selecting data center sites near renewable energy sources and locations with natural cooling potential—such as colder climates, submerged regions, or areas with access to river and lake water cooling—significantly reduces energy consumption and transmission losses [9,10,13,21,45,51,52]. Modular and transportable data centers, including containerized and prefabricated designs, enhance scalability, facilitate rapid deployment, and enable placement in optimal geographic locations [9,20,21,30]. Utilizing eco-friendly construction materials, such as engineered timber and reflective roofing, further reduces the embodied carbon footprint [6,18,20,50]. Optimized architectural layouts, including hot aisle/cold aisle configurations, raised floors, and compact equipment arrangements, improve airflow, enhance cooling efficiency, and prevent energy wastage [15,50,52]. Infrastructure innovations, such as integrating liquid cooling systems, heat pipes, and waste heat recovery connected to district heating networks, improve energy efficiency while minimizing environmental impact [16,23,24,53,54]. Notable examples include Meta’s LEED-certified facilities, Huawei’s modular data centers, and the Facebook Arctic CDC, which leverage local natural cooling and renewable resources to minimize operational carbon footprints [9,21,22]. Advanced design approaches, like the construction of data centers within hollowed mountains or submerged environments, capitalize on stable temperatures for free cooling and reduced land use [20,55].

2.6. Regulatory Compliance, Certification, and Reporting

Regulatory compliance, certification, and reporting are CSFs for the integration of green energy in data centers, ensuring environmental responsibility, operational efficiency, and stakeholder accountability [6,9,18,20]. Adhering to environmental regulations is paramount, with data centers aligning their operations with international and regional standards such as ISO 14001 and ISO 50001 for environmental and energy management, respectively [14,20]. Certification systems like LEED and the EU Code of Conduct on Data Centres Energy Efficiency not only validate sustainability efforts but also promote continuous improvement [6,9,18]. Several national frameworks reinforce this compliance; for instance, China’s green data center policies, including the National Green Data Center Pilot Work Plan and Power Usage Effectiveness targets, set efficiency benchmarks and reporting obligations [10,56]. Similarly, Japan’s Act on Rationalizing Energy Use mandates regular reporting and energy efficiency improvements [47]. The European Union’s Circular Economy Action Plan further emphasizes eco-design compliance and extended equipment lifecycles [20]. Transparent Environmental, Social, and Governance (ESG) reporting is increasingly demanded, with evolving regulations like the potential SEC-mandated carbon disclosures underscoring the need for accountability [9,18]. Incentive-based regulations, such as Finnish tax benefits for waste heat utilization and U.S. investment tax credits, encourage compliance while advancing sustainable practices [16,25]. Leading data centers exemplify these principles through high-standard certifications—Apple’s Maiden and Switch’s Las Vegas facilities demonstrate environmental commitment and robust reporting mechanisms [50].

2.7. Sustainable Lifecycle and Waste Reduction

Sustainable lifecycle and waste reduction are CSFs for integrating green energy into data centers, aiming to minimize environmental impacts and reduce the embodied energy of equipment [9,18,20,22,46]. Emphasizing circular economy principles, data centers can extend hardware lifespans through refurbishment, repurposing, and responsible e-waste management [9,20]. Leading companies like Google and Meta exemplify this approach by refurbishing server components and implementing comprehensive waste diversion strategies [9,20]. Lifecycle assessments play a vital role by evaluating the environmental footprint across production, transportation, installation, operation, and recycling stages, guiding more sustainable decisions [46]. Extending hardware life through server consolidation, virtualization, and retiring underutilized equipment further reduces e-waste and embodied energy [15]. Innovative storage solutions, such as tape storage and the elimination of dark data, help minimize unnecessary data retention and associated energy use [44]. Metrics like Waste Recycle Rate and Water Reuse Rate offer quantitative means to assess waste reduction efforts and promote resource conservation [18]. Additionally, reusing infrastructure, as seen in repurposing coal power plants into thermal energy storage facilities, curbs the embodied energy otherwise required for new construction [36]. Utilizing waste heat through technologies like combined cooling, heating, and power systems further reduces lifecycle energy waste [53].

2.8. Governance and Organizational Culture

Governance and organizational culture play a crucial role in the successful integration of green energy into data centers, as they promote leadership commitment and encourage cross-functional collaboration [9,16,19]. Leadership from top management, as demonstrated by tech companies setting ambitious green energy targets, sets the tone for organizational priorities and sustainability investments [9]. However, these commitments often fall short without proper incentives for IT and facilities staff, highlighting the importance of aligning organizational goals across departments [19]. Siloed decision-making remains a significant barrier, as it impedes collaboration on energy efficiency initiatives and compromises operational efficiency [14]. Cross-functional collaboration is particularly crucial in initiatives like waste heat utilization, which require coordinated efforts and long-term agreements between data centers and external partners, such as district heating companies [16]. Moreover, building institutional trust—especially in programs like demand response—helps mitigate concerns over economic risks and quality-of-service issues, further reinforcing the need for robust governance structures [5]. Overcoming organizational barriers such as misaligned incentives, risk aversion, and lack of collaboration is essential to cultivate a culture that actively prioritizes and invests in sustainable, green initiatives [19].

2.9. Investment Strategy

A robust investment strategy is crucial for integrating green energy into data centers, requiring comprehensive financial planning and the development of strong business cases [9,13,51]. Key components of such strategies include Return on Investment and Total Cost of Ownership analyses, which guide decision-making by balancing capital expenditures, operational expenses, and potential long-term savings [20,46,51]. Economic assessments highlight that despite high initial costs for renewable infrastructure, operational savings from improved energy efficiency and optimized cooling systems can lead to favorable payback periods and increased profitability [9,49,54,57]. Various funding mechanisms, such as green loans, power purchase agreements, and incentive programs, further enhance financial viability [9,11]. Studies show that integrating lifecycle costs—including maintenance, carbon taxes, and electricity expenses—into investment considerations enables organizations to optimize system design and operational strategies [31,46,55]. While declining renewable technology costs and government incentives improve Return on Investment, investment decisions can still be hindered by financial constraints, risk aversion, and organizational split incentives [14,16,19]. To overcome these barriers, data centers are increasingly adopting economic models that assess cost–benefit trade-offs and profitability metrics, such as Internal Rate of Return, Net Present Value, and Levelized Cost of Energy, thus strengthening the business case for sustainability investments [16,29,56]. Moreover, utilizing hybrid energy systems and co-locating renewable energy generation with data centers can reduce transmission losses and infrastructure costs, further supporting informed investment decisions [30,45,47]. Ultimately, the combination of detailed financial analyses and innovative funding approaches is essential for securing investments in renewable infrastructure and achieving long-term energy efficiency and sustainability goals [13,23,53].

2.10. Skills and Workforce Readiness

Ensuring skills and workforce readiness is crucial for the successful integration of green energy in data centers [14,19,20]. A significant challenge in this context is the siloed expertise across technical domains, which can hinder the holistic management of renewable-driven operations [14]. Addressing this requires developing specialized knowledge in energy systems, smart grid technologies, AI analytics, and the Internet of Things. The absence of such expertise has been identified as a barrier to implementing energy-efficient operations, highlighting the importance of targeted training programs to prepare staff for adopting and managing advanced technologies [19]. Moreover, deploying digital solutions to develop circular data centers further underscores the necessity of a technically adept workforce [20]. Workforce succession planning is equally vital, as knowledge loss due to retirements can compromise operational efficiency [14].

2.11. Supply Chain and Vendor Sustainability Alignment

Aligning the supply chain and vendor partnerships with sustainability goals is essential for the effective integration of green energy in data centers [14,20,24]. Long-term relationships with trusted vendors often shape procurement decisions, significantly influencing the adoption of energy-efficient technologies [14]. While such partnerships provide operational stability, they can also hinder the integration of newer, more sustainable solutions if there is an overreliance on established vendors resistant to change [14]. To overcome this challenge, data center operators should prioritize working with suppliers that demonstrate strong sustainability credentials. Initiatives like the EU’s Circular Electronics Initiative advocate for sourcing eco-friendly materials and collaborating with vendors committed to reducing environmental impacts, emphasizing the importance of ethical sourcing and responsible recycling practices [20]. Moreover, vendor-driven innovations in hardware, such as Intel’s power-saving technologies and IBM’s hot-water-cooled systems, underscore the potential of strategic vendor alignment to enhance energy efficiency [24].

2.12. Comprehensive Overview of CSFs for Green Energy Integration in Data Centers

This section provides a concise overview of the CSFs for green energy integration in data centers, as discussed in the previous sections. The first seven can be categorized as technical and operational factors, while the latter four are organizational and strategic factors. In

Table 1. Overview of CSFs for green energy integration in data centers.

	CSFs	References
Technical and Operational CSFs	CSF1: Operational energy efficiency and monitoring. Deploy high-efficiency servers and advanced cooling (including water-efficient systems and waste heat recovery) coupled with real-time energy and thermal monitoring to optimize overall operational performance.	[3,5,6,9,10,11,14,15,16,17,18,20,21,22,23,24,25,26,27,28]
	CSF2: Supply and demand optimization. Dynamically balance computing workloads with grid signals by leveraging demand response strategies, energy storage solutions, and smart grid integration to smooth out peak demand and enhance energy efficiency.	[1,2,3,11,13,17,21,23,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43]
	CSF3: Renewable energy integration. Secure and integrate on-site/off-site renewable energy sources—including strategic procurement—to ensure a continuous, reliable, and clean power supply.	[1,2,6,9,11,15,18,20,21,25,27,30,40,43,44,45,46]
	CSF4: Carbon neutrality strategies. Pursue net-zero emissions through targeted energy efficiency improvements, renewable energy adoption, and complementary measures such as carbon offset programs to meet sustainability goals.	[1,2,3,9,10,13,15,16,17,20,24,25,30,33,37,38,40,41,42,43,46,47,48,49,50]
	CSF5: Sustainable data center design and infrastructure. Design data centers with eco-friendly materials, optimized layouts, and strategic site selection near renewable resources, ensuring scalable energy and cooling infrastructure for long-term sustainability.	[6,9,10,13,15,16,18,20,21,22,23,24,30,45,50,51,52,53,54,55]
	CSF6: Regulatory compliance, certification, and reporting. Adhere to environmental regulations and recognized standards (e.g., LEED, ISO 50001) and maintain transparent ESG reporting to drive continuous improvement and accountability.	[6,9,10,14,16,18,20,25,47,50,56]
	CSF7: Sustainable lifecycle and waste reduction. Extend hardware lifespans through refurbishment, repurposing, and responsible e-waste management to reduce environmental impact and the embodied energy of equipment.	[9,15,18,20,22,36,44,46,53]
Organizational and Strategic CSFs	CSF8: Governance and organizational culture. Foster leadership commitment and cross-functional collaboration to build an organizational culture that prioritizes and invests in sustainable, green initiatives.	[5,9,14,16,19]
	CSF9: Investment strategy. Develop robust business cases—including Return on Investment and Total Cost of Ownership analyses—to secure funding for renewable infrastructure, energy efficiency upgrades, and other sustainability technologies.	[9,11,13,14,16,19,20,23,29,30,31,45,46,47,49,51,53,54,55,56,57]
	CSF10: Skills and workforce readiness. Cultivate specialized expertise in energy systems, smart grid technologies, AI analytics, and Internet of Things to empower the workforce in managing and optimizing a renewable-driven data center.	[14,19,20]
	CSF11: Supply chain and vendor sustainability alignment. Ensure that all partners and vendors uphold strong sustainability credentials, aligning the entire supply chain with the organization’s green energy and sustainability goals.	[14,20,24]

, each factor is presented with its core definition and corresponding references.

3. Methodology

This section presents the DEMATEL methodology, along with the sampling and data collection procedures used in this study.

3.1. DEMATEL Methodology

The primary objective of this study is to analyze the CSFs for green energy integration in data centers, identified through a systematic literature review (see Section 2). DEMATEL, originally developed at the Battelle Memorial Institute’s Geneva Research Center in the early 1970s, is particularly effective in clarifying complex causal relationships among various factors, distinguishing itself from other multi-criteria decision-making methods such as the Analytic Hierarchy Process (AHP), Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), Analytical Network Process (ANP), and Interpretive Structural Modeling (ISM) [58,59,60,61]. Its widespread application across sectors—including business, healthcare, technology, environmental management, and energy systems—demonstrates its versatility in addressing complex issues through a structured analytical approach [62,63,64,65,66].

DEMATEL’s unique capability to visualize and quantify interrelationships allows it to categorize factors into cause-and-effect groups, enhancing the understanding of both direct and indirect interactions [58,59,60,65,67]. This structured visualization is vital for decision-makers aiming to prioritize CSFs and assess the impact of strategic decisions in dynamic environments [60], particularly in the context of transitioning data centers to green energy solutions. Matrices and diagrams generated by DEMATEL provide a hierarchical view of interdependencies, enabling stakeholders to comprehend complex causal networks and develop effective solutions [65,67]. The methodology employed in this research follows a structured five-step process [60,64,65,68], with mathematical notations and formulations adapted from Moktadir et al. [60] to ensure precise calculation of interrelationships.

3.1.1. Step 1: Constructing the Initial Average Direct Relation Matrix

The first step of the DEMATEL methodology involves constructing the Average Direct Relation Matrix to quantify the direct influence among the identified CSFs. Experts assess the extent to which each CSF affects the others using a standardized scale from 0 (no influence) to 4 (very high influence) [58,59,64,66]. If there are $n$ identified CSFs and $H$ experts involved, each expert $k$ provides an $(n\times n)$ matrix, denoted as $X^k=\left[x_{ij}^k\right]$, where $x_{ij}^k$ represents the influence of CSF $i$ on CSF $j$ as evaluated by expert $k$. The set of all initial matrices from the $H$ experts is represented as $X^1=\left[x_{ij}^1\right],X^2=\left[x_{ij}^2\right],\dots ,X^H=\left[x_{ij}^H\right]$. The Average Direct Relation Matrix $M=\left[{\tilde{x}}_{ij}\right]$ is then calculated by averaging these expert-provided matrices, using the formula

$${\tilde{x}}_{ij}={\displaystyle \frac{1}{H}}{\sum }_{k=1}^H{x}_{ij}^k$$

(1)

This matrix consolidates the collective evaluations of all experts and provides the foundational input for subsequent analyses within the DEMATEL framework.

3.1.2. Step 2: Calculating the Normalized Direct Relation Matrix

The second step of the DEMATEL methodology involves normalizing the previously calculated Average Direct Relation Matrix $M$ to derive the Normalized Direct Relation Matrix $P$, which facilitates the comparison of influence scores among the CSFs. Normalization ensures that all influence values are proportionally adjusted within a standardized range between 0 and 1. To perform the normalization, a scaling factor $S$ is calculated as the minimum of the reciprocals of the maximum row and column sums of the absolute values in matrix $M$, expressed as

$$S=\min \left[{\displaystyle \frac{1}{{\sum }_{j=1}^n\left|{\tilde{x}}_{ij}\right|}},{\displaystyle \frac{1}{{\sum }_{i=1}^n\left|{\tilde{x}}_{ij}\right|}}\right]$$

(2)

Once $S$ is determined, the normalized matrix $P=\left[p_{ij}\right]$ is obtained by multiplying each element of $M$ by $S$ using the formula $P=M\times S$. This process ensures that the maximum element in $P$ does not exceed 1, preserving the proportional relationships among the CSFs and maintaining the integrity of the influence data for the subsequent stages of the DEMATEL analysis.

3.1.3. Step 3: Calculating the Total Relation Matrix

The third step of the DEMATEL methodology involves constructing the Total Relation Matrix $T$, which captures both direct and indirect interactions among the CSFs for green energy integration in data centers. Building upon the Normalized Direct Relation Matrix $P$ derived in Step 2, this step utilizes matrix algebra to fully represent the complex interrelationships between CSFs. The Total Relation Matrix $T$ is calculated using the formula $T=P{\left[I-P\right]}^{-1}$. In this equation, $I$ denotes the identity matrix with the same dimensions as $P$, and ${\left[I-P\right]}^{-1}$ represents the inverse of the matrix obtained by subtracting $P$ from $I$. This operation accounts for the cumulative effects of each CSF, capturing both their direct influences and the indirect impacts transmitted through the network of interrelated factors. By multiplying $P$ by ${\left[I-P\right]}^{-1}$, the resulting matrix $T$ provides a comprehensive view of how each CSF affects and is affected by others within the system. The Total Relation Matrix $T$ is vital in the DEMATEL framework, as it offers decision-makers an in-depth understanding of the full spectrum of interactions, thereby supporting informed strategies for managing the examined CSFs.

3.1.4. Step 4: Calculating and Visualizing the Prominence and Net Effects of CSFs

The fourth step of the DEMATEL methodology involves determining and illustrating the prominence and net effects of the CSFs. This analysis uses the Total Relation Matrix $T={\left[t_{ij}\right]}_{n\times n}$, derived in the previous step, to quantify both the influence each CSF exerts on others and the influence it receives. To calculate these effects, two indices are computed for each CSF. The total influence exerted by a CSF is denoted as $r_i$ and is calculated using the formula

$$r_i={\sum }_{j=1}^n{t}_{ij},\forall i$$

(3)

The total influence received by a CSF is represented as $c_i$, calculated as

$$c_j={\sum }_{i=1}^n{t}_{ij},\forall j$$

(4)

The prominence of each CSF is determined by the sum $(r_i+c_j)$, which reflects the overall significance of the factor within the system. The net effect is calculated as $(r_i-c_j)$, indicating whether the CSF functions as a cause or an effect. A positive value of $(r_i-c_j)$ suggests that the CSF predominantly influences others, while a negative value indicates that it is more influenced by other factors.

The prominence and net effect values are then plotted on a diagram to visually represent the interrelationships among the CSFs. The horizontal axis displays the prominence, while the vertical axis represents the net effect. This visual depiction enables decision-makers to identify the most influential factors and understand the dynamic interactions within the network of CSFs, thereby facilitating informed decision-making.

3.1.5. Step 5: Visualizing the Most Significant Causal Relationships Among CSFs

The fifth and final step of the DEMATEL methodology involves the graphical representation of the most significant causal relationships among the CSFs. This visualization aims to highlight the most impactful interactions derived from the Total Relation Matrix $T$, enabling decision-makers to focus on the key factors influencing the system. To ensure that the diagram emphasizes only the most relevant relationships, a threshold value $\theta$ is established. This threshold is calculated using the formula $\theta =\mu +\sigma$, where μ is the mean and $\sigma$ is the standard deviation of the entries in the Total Relation Matrix. Relationships with influence values exceeding $\theta$ are considered significant and are depicted using directed arrows in the causal relationship diagram. This approach, supported by studies such as those by Moktadir et al. [60], Huang et al. [69], and Kouhizadeh et al. [59], ensures that the visualization remains clear and focused by excluding less impactful interactions that could obscure critical insights. By concentrating on the most substantial relationships, this step provides a comprehensive understanding of the causal structure among the CSFs, aiding strategic decision-making.

3.2. Sample Information and Data Collection Method

Participants for the DEMATEL analysis were selected using convenience and snowball sampling, common methods in this type of research [59,60,68]. The selection focused on engineers and managers with extensive knowledge of data center operations; their role in this study was to evaluate the interconnections between CSFs for green energy integration.

The initial identification of these CSFs was based on a review of the existing literature (see Section 2). To ensure their relevance and precision, a panel of three industry experts, each with approximately 20 years of experience, was consulted. This panel included a Head of Strategy Division from an energy conglomerate, a Technical Coordinator in data center construction from a project management consultancy, and a Principal Senior Land Development Manager from a data center real estate firm. During the validation process, the experts assessed the CSFs based on their professional experience and current industry practices. To facilitate their evaluation, participants completed questionnaires that detailed each CSF and used an 11 × 11 pairwise comparison matrix. This matrix applied a five-point scale (0 = no influence, 4 = very high influence) to quantify the interdependencies between the CSFs.

A total of 24 valid responses were obtained from a diverse group of professionals. The sample includes individuals aged between 30 and 59 years, with a gender distribution of four females and twenty males. Participants hold senior positions in organizations that either host or manage data centers, representing a wide range of industries, including energy, software, insurance, real estate, construction, pharmaceuticals, manufacturing, financial services, logistics, food processing, shipping, and healthcare. Their roles span various functions, such as IT Managers, Environmental/Energy and Sustainability Managers, and Facilities/Infrastructure Managers. In terms of academic qualifications, most respondents hold master’s degrees in relevant fields, including Energy and Environmental Management, Engineering, Business Administration, and Information Systems. Their professional experience ranges from 10 to 33 years, with an average of approximately 18.9 years, reflecting a high level of expertise and long-term engagement in their respective fields. Detailed professional profiles of all participants can be found in

Table A1. Participant profiles.

ID	Age Range	Gender	Highest Degree	Company Classification	Current Position	Years of Experience
1	40–49	Male	Master’s in Supply Chain Management	Oil and Gas; Energy	Head of Strategy Division	20
2	40–49	Male	Master’s in Business Administration	Insurance Company	IT Risk and Compliance Manager	16
3	40–49	Male	Master’s in Energy Production and Management	Project Management; Consultant	Technical Coordinator in Data Center Construction	22
4	40–49	Male	Master’s in Mechanical Engineering	Real Estate Investment Company	Asset Development Senior Project Manager	19
5	40–49	Male	BEng in Civil Engineering	Construction Company	Environmental and Health and Safety Manager	20
6	30–39	Male	BEng in Mechanical Engineering	Pharmaceutical Company	Facilities Engineer	10
7	40–49	Male	BEng in Electrical Engineer	IT Infrastructure Consulting	Senior Consultant—Green Technologies	18
8	30–39	Male	Master’s in Business Administration	Manufacturing Company	Corporate Sustainability Coordinator	12
9	50–59	Male	BBA in Business Administration	Financial Services Firm	Director of IT Infrastructure	25
10	40–49	Male	Master’s in Operations Management	Logistics and Supply Chain Company	Regional Facilities Manager	22
11	40–49	Male	PhD in Environmental Management	Academic Institution	Lecturer in Environmental Management and Sustainability	20
12	40–49	Female	Master’s in Business Administration—Total Quality Management	Beverage Company	Sustainability Program Manager	17
13	50–59	Male	BSc in Industrial Engineering	Retail Corporation	Senior Operations Consultant	27
14	40–49	Male	Master’s in Information Systems	Banking Institution	IT Manager	28
15	40–49	Male	BEng in Mechanical Engineering	Food Processing Company	Director of Corporate Sustainability	19
16	30–39	Female	Master’s in Business Administration	Retail Corporation	Environmental Compliance Manager	15
17	50–59	Male	Master’s in Economics	Shipping Company	Global Infrastructure Advisor	24
18	30–39	Female	Master’s in Business Administration	Telecommunications Provider	Corporate Sustainability Specialist	12
19	50–59	Male	BSc in Physics	Consumer Electronics Company	Facilities Energy Supervisor	22
20	30–39	Male	Master’s in Computer Engineering	Healthcare Organization	IT Infrastructure Advisor	11
21	30–39	Male	Master’s in Energy	Energy Trading	Head of Long-Term Trading Department	11
22	50–59	Male	Master’s in Computer Science	IT Services	Consulting Practice Manager	33
23	40–49	Female	Master’s in Civil Engineering and Construction	Data Center Real Estate	Principal Senior Land Development Manager	15
24	40–49	Male	Master’s in Energy: Strategy, Law, and Economics	Oil and Gas; Electricity; Renewables	Head of Business Analysis, IT	15

in Appendix A.

To evaluate how consistently the experts assessed the interrelationships among factors in the DEMATEL framework, the Intraclass Correlation Coefficient (ICC) was applied. This statistical measure was calculated using a two-way mixed-effects model with average measures (i.e., ICC(3, k)). It is particularly appropriate in this setting, as it measures the degree of agreement among evaluators while accommodating the structured, ordinal nature of the scoring system used in DEMATEL studies. The analysis yielded an ICC value of 0.927 with a significance level of p < 0.001, indicating high consistency across the expert assessments. Although individual judgments may exhibit some variation, the collective pattern of responses demonstrates a strong level of alignment. This consistency reinforces the reliability of the data and supports the robustness of the resulting causal interpretations drawn from the DEMATEL analysis.

4. Results

This section presents the key findings from the application of the DEMATEL methodology, organized into two subsections. Section 4.1 examines the prominence and net effects of each CSF, as determined in stage 4 of the DEMATEL process. Section 4.2 analyzes the most significant causal relationships among the CSFs, corresponding to stage 5 of the methodology. Additional details on the core matrices used in these analyses—the Average Direct Relation Matrix, the Normalized Direct Relation Matrix, and the Total Relation Matrix (stages 1–3 of DEMATEL)—are available in Appendix A in

Table A2. DEMATEL matrices.

Average Direct Relation Matrix
CSFs	CSF1	CSF2	CSF3	CSF4	CSF5	CSF6	CSF7	CSF8	CSF9	CSF10	CSF11
CSF1		2.750	2.292	2.875	2.833	2.417	1.792	1.708	2.208	1.875	1.583
CSF2	2.708		2.667	2.542	2.333	2.250	1.375	1.875	2.583	2.208	1.583
CSF3	2.625	2.667		3.917	3.208	2.750	1.708	2.042	3.042	2.000	2.208
CSF4	3.125	2.500	3.667		2.958	3.250	2.792	2.958	3.250	2.208	2.375
CSF5	3.333	2.625	3.250	2.833		3.083	2.833	1.917	2.917	1.958	2.083
CSF6	2.667	2.167	2.833	3.375	3.042		2.708	2.542	2.500	1.917	2.708
CSF7	1.750	1.500	1.542	2.250	2.750	2.125		1.958	1.917	1.667	2.083
CSF8	2.917	2.625	2.958	3.333	2.958	3.333	2.792		3.542	3.167	3.083
CSF9	3.208	3.083	3.375	3.458	3.208	2.833	2.583	3.042		2.625	2.708
CSF10	2.333	2.792	2.125	2.250	2.125	2.458	2.042	2.667	2.125		1.708
CSF11	1.625	1.542	2.208	2.375	2.500	2.625	2.958	2.333	2.333	1.500
Normalized Direct Relation Matrix
CSFs	CSF1	CSF2	CSF3	CSF4	CSF5	CSF6	CSF7	CSF8	CSF9	CSF10	CSF11
CSF1	0.000	0.090	0.075	0.094	0.092	0.079	0.058	0.056	0.072	0.061	0.052
CSF2	0.088	0.000	0.087	0.083	0.076	0.073	0.045	0.061	0.084	0.072	0.052
CSF3	0.085	0.087	0.000	0.128	0.104	0.090	0.056	0.066	0.099	0.065	0.072
CSF4	0.102	0.081	0.119	0.000	0.096	0.106	0.091	0.096	0.106	0.072	0.077
CSF5	0.109	0.085	0.106	0.092	0.000	0.100	0.092	0.062	0.095	0.064	0.068
CSF6	0.087	0.071	0.092	0.110	0.099	0.000	0.088	0.083	0.081	0.062	0.088
CSF7	0.057	0.049	0.050	0.073	0.090	0.069	0.000	0.064	0.062	0.054	0.068
CSF8	0.095	0.085	0.096	0.109	0.096	0.109	0.091	0.000	0.115	0.103	0.100
CSF9	0.104	0.100	0.110	0.113	0.104	0.092	0.084	0.099	0.000	0.085	0.088
CSF10	0.076	0.091	0.069	0.073	0.069	0.080	0.066	0.087	0.069	0.000	0.056
CSF11	0.053	0.050	0.072	0.077	0.081	0.085	0.096	0.076	0.076	0.049	0.000
Total Relation Matrix
CSFs	CSF1	CSF2	CSF3	CSF4	CSF5	CSF6	CSF7	CSF8	CSF9	CSF10	CSF11
CSF1	0.342	0.398	0.420	0.460	0.443	0.422	0.360	0.352	0.408	0.333	0.337
CSF2	0.422	0.314	0.428	0.450	0.428	0.415	0.347	0.355	0.417	0.341	0.336
CSF3	0.479	0.448	0.410	0.552	0.514	0.490	0.410	0.412	0.489	0.383	0.404
CSF4	0.529	0.478	0.554	0.480	0.547	0.542	0.474	0.470	0.532	0.419	0.441
CSF5	0.500	0.450	0.507	0.527	0.423	0.502	0.444	0.411	0.488	0.384	0.403
CSF6	0.479	0.434	0.493	0.538	0.510	0.408	0.439	0.426	0.474	0.380	0.418
CSF7	0.355	0.324	0.357	0.398	0.399	0.373	0.271	0.324	0.360	0.294	0.318
CSF8	0.544	0.501	0.556	0.600	0.569	0.566	0.494	0.402	0.561	0.463	0.478
CSF9	0.545	0.506	0.560	0.595	0.567	0.544	0.480	0.484	0.450	0.441	0.460
CSF10	0.415	0.402	0.417	0.446	0.427	0.426	0.370	0.382	0.409	0.278	0.344
CSF11	0.386	0.357	0.411	0.441	0.429	0.423	0.391	0.365	0.407	0.318	0.285

Notes: Threshold value = mean + standard deviation = 0.435 + 0.074 = 0.509. Values exceeding this threshold are highlighted in red italics.

.

4.1. Prominence and Net Effects of CSFs

This analysis evaluates the impact of different CSFs on green energy integration in data centers.

Table 2. Prominence and net effect values.

Critical Success Factors	${\mathit{r}}_{\mathit{i}}$	${\mathit{c}}_{\mathit{j}}$	${\mathit{r}}_{\mathit{i}}+{\mathit{c}}_{\mathit{j}}$	${\mathit{r}}_{\mathit{i}}-{\mathit{c}}_{\mathit{j}}$	Impact
CSF 1: Operational energy efficiency and monitoring.	4.276	4.997	9.272	−0.721	Effect
CSF 2: Supply and demand optimization.	4.253	4.612	8.865	−0.359	Effect
CSF 3: Renewable energy integration.	4.991	5.112	10.104	−0.121	Effect
CSF 4: Carbon neutrality strategies.	5.467	5.488	10.955	−0.022	Effect
CSF 5: Sustainable data center design and infrastructure.	5.038	5.258	10.296	−0.220	Effect
CSF 6: Regulatory compliance, certification, and reporting.	5.000	5.112	10.112	−0.111	Effect
CSF 7: Sustainable lifecycle and waste reduction.	3.773	4.479	8.252	−0.706	Effect
CSF 8: Governance and organizational culture.	5.733	4.382	10.114	1.351	Cause
CSF 9: Investment strategy.	5.633	4.995	10.628	0.637	Cause
CSF 10: Skills and workforce readiness.	4.316	4.035	8.351	0.281	Cause
CSF 11: Supply chain and vendor sustainability alignment.	4.213	4.224	8.438	−0.011	Effect

presents an overview of the significance and net effect values for each CSF. Additionally, Figure 1 provides a visual representation of these scores, with prominence plotted on the x-axis and net effect on the y-axis. This visualization helps illustrate the influence dynamics among the CSFs, offering a clear understanding of how each factor contributes to the broader strategic objectives.

Six CSFs received high prominence scores (i.e., above 10), highlighting their crucial role in the green energy integration framework for data centers. Among them, “carbon neutrality strategies” stands out as the most prominent CSF. Its near-zero net effect value suggests that it functions both as a cause and an effect, depending on its specific interactions and interdependencies with other factors. “Governance and organizational culture” and “investment strategy” exhibit positive net effects, identifying them as causal factors. These factors act as key drivers of change within the data center ecosystem, exerting a stronger influence on other CSFs than they receive. Their role as primary catalysts highlights their capacity to initiate and shape transformations across the system. On the other hand, “renewable energy integration”, “regulatory compliance, certification, and reporting”, and “sustainable data center design and infrastructure” display negative net effects, classifying them as effect-based factors. This suggests that they are more influenced by other CSFs rather than acting as primary drivers. Changes in these areas are typically triggered by developments in the causal CSFs, indicating a dependency where these factors adapt and respond to shifts initiated by governance structures, investment strategies, and other key drivers of green energy adoption. Modifications in all the aforementioned domains could substantially influence energy integration practices in data centers. In contrast, CSFs such as “sustainable lifecycle and waste reduction”, “skills and workforce readiness”, and “supply chain and vendor sustainability alignment” received lower prominence scores (i.e., below 8.5), indicating their relatively lesser influence within the broader green energy integration framework in the data center sector.

4.2. Key Causal Relationships Among CSFs

The Total Relation Matrix provides a comprehensive assessment of the causal relationships among the identified CSFs. In this study, the threshold was determined by adding the mean and standard deviation of all values within the matrix, as described in Section 3. This threshold serves to highlight the most significant relationships, with values exceeding this benchmark distinctly marked in red italics within the matrix (refer to Table A2 in Appendix A for details). Based on these values, the key causal relationships diagram was developed and is visually represented in Figure 2. In this figure, solid arrows denote one-way effects, whereas dashed arrows represent reciprocal effects.

This diagram shows the interconnections among the main CSFs influencing green energy integration in data centers. Understanding these structured interactions provides valuable insights into how targeted strategic actions can influence the network and improve the effectiveness of green energy integration efforts.

In this analysis, “governance and organizational culture” and “investment strategy” emerge as the most influential causal CSFs, exerting a strong direct impact on all other key CSFs within the model. Additionally, “governance and organizational culture” directly affects “investment strategy”, positioning the latter as a mediating factor that connects governance-related aspects with all other CSFs.

“Carbon neutrality strategies” holds a central role within the network, acting as a crucial mediator. Specifically, it serves as a link between the two primary causal factors mentioned above and four effect factors: “renewable energy integration”, “sustainable data center design and infrastructure”, “regulatory compliance, certification, and reporting”, and “operational energy efficiency and monitoring”. While “carbon neutrality strategies” primarily drives these effect factors, it also maintains reciprocal relationships with the first three. These reciprocal links show how achieving carbon neutrality involves continuous adaptation—while it drives advancements in these areas, it is simultaneously reshaped by the evolving technological and regulatory contexts they represent.

Furthermore, “renewable energy integration” and “regulatory compliance, certification, and reporting” function as additional bridging elements, linking “governance and organizational culture”, “investment strategy”, and “carbon neutrality strategies” with “sustainable data center design and infrastructure”.

5. Discussion and Conclusions

This study examined the CSFs influencing green energy integration in data centers. By applying the DEMATEL methodology, it mapped the cause-and-effect relationships among these CSFs. A combination of insights from a systematic literature review and industry experts highlighted the significant influence and roles of these factors. This analysis identified key areas where strategic interventions could substantially enhance green energy integration practices. Additionally, it provided a deeper understanding of the interactions among various CSFs, forming a strong basis for this study’s findings and implications.

5.1. Discussion of Key Findings

This section presents a discussion of this study’s main findings, analyzing how different CSFs interact to shape green energy integration in data centers.

5.1.1. Governance and Investment as Key Causal Factors

The DEMATEL analysis identifies “governance and organizational culture” and “investment strategy” as primary causal factors with substantial influence on key areas of green energy integration within data centers. “Governance and organizational culture”, through leadership commitment and cross-functional collaboration, establishes the strategic direction, promoting a sustainability-focused mindset that shapes decision-making and operational priorities across the data center ecosystem. By embedding sustainability as a core value, governance structures guide regulatory compliance, infrastructure planning, and operational standards, directly influencing key effect factors such as “renewable energy integration”, “carbon neutrality strategies”, “sustainable data center design and infrastructure”, “regulatory compliance, certification, and reporting”, and “operational energy efficiency and monitoring”. It is important to note that the “governance and organizational culture” factor has received limited attention in research on green energy integration in data centers [9,19]. However, this study shows that without a strong governance framework, sustainability initiatives often lack coherence and long-term viability, leading to fragmented adoption of green energy solutions.

Equally central is “investment strategy”, which exerts direct influence on the same five effect factors, ensuring that financial resources are effectively allocated to support energy efficiency initiatives, infrastructure modernization, and regulatory compliance. By prioritizing expenditures on renewable energy procurement, energy storage solutions, and advanced monitoring systems, investment decisions directly impact the effectiveness of sustainability measures. As a mediating factor, “investment strategy” bridges the strategic intent set by governance structures with the operationalization of sustainability initiatives. This intermediary role ensures that the ambitions of governance are translated into tangible financial commitments that drive the transformation of data center operations toward greater energy efficiency and sustainability. Investment thus operates both as a catalyst and as a reinforcing mechanism, where early financial commitments can generate operational benefits that justify and accelerate further investment. Such recursive dynamics highlight how financial strategy can evolve through learning, performance feedback, and institutional adaptation. While previous studies highlight financial constraints as barriers to green energy adoption, they often treat investment as a static limitation rather than a dynamic enabler [14,16,19].

5.1.2. Carbon Neutrality Strategies as a Central Mediator

The role of “carbon neutrality strategies” as a key mediating factor highlights its importance in structuring the interconnections between governance, investment, and operational sustainability outcomes. As a bridge linking “governance and organizational culture” with the primary effect factors, “carbon neutrality strategies” translates broad sustainability commitments into actionable pathways toward decarbonization. This involves integrating energy efficiency measures, offsetting residual emissions, and deploying innovative renewable solutions to reduce overall carbon footprints. The influence of “carbon neutrality strategies” on “renewable energy integration”, “sustainable data center design and infrastructure”, “regulatory compliance, certification, and reporting”, and “operational energy efficiency and monitoring” demonstrates its role as a channel through which governance-driven sustainability objectives materialize in practice. Prior studies emphasize carbon neutrality as an end goal [2,9,20,50], but its function as a structural intermediary suggests a more complex role in balancing governance priorities with operational realities. From a systems perspective, carbon neutrality is both an output of prior strategic decisions and an input that reshapes future priorities, forming a feedback loop that reinforces and evolves the sustainability agenda over time.

Similarly, “carbon neutrality strategies” mediates the relationship between “investment strategy” and the same effect factors by defining how financial resources are allocated to achieve net-zero goals. Investment in carbon reduction initiatives, such as energy-efficient cooling technologies, direct renewable power procurement, and waste heat recovery systems, strengthens the sustainability profile of data centers while reinforcing compliance with environmental regulations. This mediating role reflects a dynamic business mechanism in which financial decisions and decarbonization strategies co-adapt, adjusting in response to regulatory developments, technological breakthroughs, and competitive positioning. Notably, the reciprocal relationships observed between “carbon neutrality strategies” and “renewable energy integration”, “sustainable data center design and infrastructure”, and “regulatory compliance, certification, and reporting” suggest an iterative process in which advancements in carbon neutrality drive further refinements in these domains, while regulatory shifts and technological innovations reciprocally influence carbon reduction strategies. This reflects a learning-oriented model in which data centers can iteratively optimize their sustainability performance as part of an adaptive strategic system.

5.1.3. Renewable Energy Integration and Compliance as Structural Mediators

Beyond “carbon neutrality strategies”, the roles of “renewable energy integration” and “regulatory compliance, certification, and reporting” as additional mediators highlight the complexity of green energy transition within data centers. These CSFs do not operate in isolation but form part of an interconnected system where influences are transmitted and reinforced across the network. Specifically, they serve as critical linkages connecting “governance and organizational culture”, “investment strategy”, and “carbon neutrality strategies” with “sustainable data center design and infrastructure”.

The mediation of “renewable energy integration” reflects the necessity of aligning organizational directives and financial commitments with the physical and technical feasibility of renewable energy adoption. Effective governance frameworks promote renewable procurement policies, while investment strategies allocate capital for infrastructure capable of integrating on-site and off-site renewable sources. Consequently, “renewable energy integration” ensures that sustainability ambitions translate into practical energy solutions that enhance the long-term resilience of data centers. This interaction forms a feedback structure in which strategic decisions influence renewable implementation, which in turn informs future governance priorities and investment adjustments.

Similarly, “regulatory compliance, certification, and reporting” function as a structural mediator by institutionalizing governance-driven sustainability commitments and investment-backed initiatives into formalized standards and transparent reporting mechanisms. This mediation advances accountability and clarity in sustainability practices, thereby guiding and reinforcing sustainable design decisions within data center infrastructures. Moreover, these compliance mechanisms often trigger a reinforcing loop: as organizations demonstrate progress through certifications and disclosures, they are incentivized to pursue further sustainability improvements to maintain stakeholder trust and competitive advantage. Consistent with the existing literature, adherence to frameworks such as ISO 14001 and ISO 50001, as well as certifications like LEED, promotes continuous improvements in operational energy efficiency and enhances stakeholder trust in sustainable data center operations [6,9,14,18,20].

5.2. Managerial Implications

The findings of this study emphasize significant managerial implications for decision-makers in data centers working towards the effective integration of green energy. Foremost is the establishment of a robust governance framework that incorporates sustainability as a core strategic value. Leadership must promote cross-functional collaboration and embed sustainability principles into all operational and investment decisions, ensuring alignment across departments—particularly procurement, facilities management, and IT operations.

Another key factor of successful green energy integration is developing strategic investments that support the transition to energy-efficient technologies and the adoption of renewable energy sources. Businesses must formulate clear business cases demonstrating the Return on Investment and Total Cost of Ownership over time. Securing adequate capital for investments in renewable energy infrastructure, energy storage, and advanced cooling technologies is crucial for maintaining operational resilience and sustainable growth.

The carbon neutrality strategy serves as another critical link between strategic leadership and operational improvements. Managers should develop clear and actionable carbon reduction pathways that encompass renewable energy integration, enhanced energy efficiency, and the use of carbon offset mechanisms. Continuous monitoring and adaptation of these strategies through feedback mechanisms and regulatory compliance will enable businesses to dynamically respond to technological and regulatory developments.

Furthermore, regulatory compliance and the integration of renewable energy sources are not just operational requirements but structural factors that can enhance the resilience of data centers. Managers must ensure adherence to internationally recognized environmental and energy standards and certifications, thereby strengthening transparency, credibility, and investor confidence. Participation in ESG reporting and alignment with sustainable procurement practices will further contribute to reinforcing corporate responsibility and competitiveness.

Taking this study’s findings into account holistically, the adoption of an integrated approach that acknowledges the interdependencies between governance, investment, carbon neutrality, and regulatory compliance is clearly recommended. This will enable data centers to achieve sustainable energy strategies and long-term competitiveness within the industry.

5.3. Societal Implications

The results of this study have significant societal implications for governmental institutions and regulatory authorities forming the sustainable development of digital infrastructure. As “governance and organizational culture” emerged as one of the most important drivers of green energy integration, policymakers are encouraged to create clear, enforceable sustainability mandates for data centers, including national frameworks for carbon accounting, mandatory disclosure of embodied emissions, and incentives for procurement of low-carbon hardware. Public–private green financing tools could also help support investment methods in line with long-term decarbonization targets, strengthening early-stage infrastructure modernization. Such policy guidelines should provide a supporting environment encouraging transparency, innovation, and alignment among corporate governance, energy regulation, and sustainability reporting.

5.4. Limitations and Future Research Directions

This study has certain limitations that require further investigation. First, the use of the DEMATEL method relies on subjective expert judgments, which may affect the accuracy of the causal relationships among the CSFs. Future research could enhance the validity of the findings through empirical validation by incorporating actual energy consumption data from data centers. Second, the use of convenience and snowball sampling, while common in exploratory studies, limits the generalizability of the results. Future research could incorporate a broader sample, including policymakers, renewable energy providers, and researchers, to strengthen the external validity of the findings. Additionally, DEMATEL provides a static representation of the causal relationships among factors, without capturing the dynamic changes driven by technological advancements and regulatory shifts. The application of longitudinal studies or repeated DEMATEL analyses could reveal how the interactions of CSFs evolve over time, particularly with the introduction of new technologies such as AI-driven energy optimization and next-generation storage solutions. Future research should also explore the reciprocal relationship between these optimization technologies and the very infrastructure they depend on. For instance, a focused investigation could examine how using AI to manage energy more efficiently might actually increase the need for computing power and the number of servers needed in large-scale data centers. Investigating these unintended consequences through systems-based modeling or dynamic impact assessments could offer valuable insights into long-term sustainability trade-offs. Finally, this study primarily focuses on internal organizational and technical factors without considering external influences such as shifting markets, consumer preferences, and regulatory fluctuations. Future research could integrate multi-criteria decision analysis or System Dynamics approaches, focusing on a deeper understanding of the interaction between operational and strategic factors.