Advancing Circular Supplier Selection: Multi-Criteria Perspectives on Risk and Sustainability

Abstract

Supplier selection is a crucial factor for ensuring compliance with the circular economy’s principles. Existing approaches often overlook the integration of circularity and risk assessment in supplier evaluation, limiting their effectiveness in achieving sustainability goals. This paper addresses this gap by applying suitable criteria and proposing a structured decision-making model for circular supplier selection. The model innovatively integrates Multi-Criteria Decision Analysis (MCDA) techniques with risk evaluation, providing a comprehensive framework for assessing suppliers in circular supply chains. By advancing the theoretical understanding of circular supplier selection, this research contributes to both academia and practice, reinforcing the alignment between supply chain decision-making and the Sustainable Development Goal (SDG), particularly Target 12.5.

1. Introduction

A Circular Supply Chain (CSC) integrates the sustainable principles of the circular economy (CE) into supply processes, promoting a continuous flow of resources through reuse, reduction, and recycling. Unlike the traditional linear model, CSC is grounded in closed-loop processes, resource efficiency, and waste minimization [1]. A fundamental enabler of CSC effectiveness is the adaptability of supplier management [2]. However, conventional supplier selection frameworks often lack integration with CE principles, leading to challenges in aligning procurement decisions with sustainability goals. Existing models predominantly emphasize cost and operational efficiency, neglecting critical sustainability-driven criteria essential for a circular approach [3].

However, transitioning from linear to circular models presents significant challenges, including the absence of specialized knowledge, insufficient management support, and financial constraints. The process can also be hindered by the lack of standardized performance measurement systems and gaps in environmental regulations. Additionally, a lack of urgency or awareness can further complicate the transition [4]. Supplier selection plays a crucial role in overcoming these challenges, as organizations must ensure that suppliers align with circular principles to enhance resource efficiency and minimize environmental impact. The United Nations Sustainable Development Goals (SDGs), particularly Target 12.5 (Responsible Consumption and Production), reinforce the importance of circularity by aiming to reduce waste generation through prevention, reduction, recycling, and reuse [5]. Achieving this goal requires structured decision-making processes that integrate sustainability as a core criterion.

Although previous studies [6,7] have explored supplier selection in sustainable supply chains, most frameworks remain anchored in conventional Multi-Criteria Decision Analysis (MCDA) methodologies without adequately incorporating circularity principles. Furthermore, while risk assessment has been widely applied in supplier evaluation, its integration within CSC decision-making remains underexplored. The lack of a structured approach that simultaneously accounts for CE principles, sustainability-driven criteria, and risk factors represents a critical gap in the literature. Addressing this gap requires an advanced decision-making framework that moves beyond traditional MCDA applications by embedding CE priorities into supplier evaluation. These methods prioritize aspects such as resource efficiency, waste minimization, closed-loop logistics, environmental compliance, and risk factors.

A well-structured decision-making process significantly enhances the effectiveness of supplier selection [8]. By incorporating MCDA techniques, such as Analytic Hierarchy Process (AHP) and Benefits, Opportunities, Costs, and Risks (BOCR), organizations can systematically assess suppliers and support more informed decision-making [9,10].

This paper aims to bridge this gap by identifying suitable criteria and establishing a structured procedure for assessing sustainable suppliers within the chemical industry. Specifically, the research addresses the following questions: (RQ1) Which criteria suit circular supplier selection (CSS)? (RQ2) How can CSS be assessed? To answer these questions, the study proposes an integrated decision-making model that combines MCDA techniques with risk evaluation, ensuring a more comprehensive and systematic assessment of circular suppliers. Unlike previous studies, this research extends theoretical understanding by positioning supplier selection at the intersection of CE, risk assessment, and MCDA methodologies. Through this approach, it not only advances supply chain sustainability but also strengthens the alignment between CSC practices.

2. Literature Review

2.1. Circular Supply Chain

A CSC refers to a supply chain model rooted in the principles of the CE, aiming to reduce waste, optimize resource usage, and close the product lifecycle by encouraging reuse, recycling, and remanufacturing. The transition from linear to CSC plays a vital role in fostering sustainable production and consumption. A fundamental element of the CSC is minimizing waste by enhancing resource efficiency, implementing closed-loop logistics, and extending product life cycles. These principles strongly correspond with SDG’s Target 12.5, which seeks to significantly reduce waste generation through prevention, reduction, recycling, and reuse [11]. By embedding CE practices, this framework, CSC serves as a strategic model that enables businesses to transition from conventional supply chain structures to circular approaches that align with sustainability objectives.

The implementation of a CSC faces various barriers, including operational, cultural, technological, and structural challenges, which differ depending on the organizational context. Identifying, categorizing, and prioritizing these obstacles is crucial to overcoming difficulties and ensuring a successful transition to circular practices [12]. Among these, financial, economic, and technological barriers play a significant role, with financial pressures standing out as a key obstacle, particularly in the automotive sector, as noted by [13,14], and further explored by [15].

Moreover, CSC challenges, as highlighted by [16], include a lack of understanding, difficulties with data security in relationship management, insufficient expertise in data management, and unawareness of the benefits within CSC. Despite its sustainability potential, adopting CSC faces many challenges, including technological and implementation barriers. However, innovative technologies contribute to mitigating these obstacles [17]. Additionally, CSC approaches have been assessed, and barriers have been addressed by [18]. Equally important are the evaluation of solution strategies [19], the investigation and ranking of barriers, and the exploration of factors that facilitate the transition of business models.

Challenges encompass any factor that can impede the progress of a project or the achievement of its goals. These obstacles can be both internal and external, such as shifts in market conditions, technological advancements, and resource limitations. The impact of these challenges is evident in projects that experience delays, budget overruns, and reduced features and functions compared to their original plans. In this context, challenges include tracking the United Nations SDG [20], adopting CE principles [21], and implementing CSC [22,23].

Drivers are crucial factors to understand when executing a project. They can be related to the initial need, the desired final product, the imposed constraints, the operational environment, the key players involved, or the overall business framework. In this context, the resource-based view, resource-dependence theory, and stakeholder theory were used to identify and assess drivers [24]. Incorporating resilience practices into CSC empowers companies to meet their sustainability goals [25].

Enablers support the essential activities needed to make a given technology viable, providing fundamental support for the respective process. These elements, commonly referred to as enablers, include items related to infrastructure, architecture, and compliance. They play a crucial role in creating the necessary conditions for the effective implementation and operation of the technology, ensuring its alignment with established requirements and standards. In this context, a framework for developing blockchain-enabled sustainable manufacturing practices [26].

In CSC, risks can arise from various sources, including economic, environmental, social, and technological factors. These risks encompass supply chain disruptions, waste management challenges, and potential resource scarcity. Effective risk management in this context safeguards business continuity, enhances resilience, and promotes long-term sustainability by addressing these vulnerabilities and ensuring smooth operation. The risk factors impacting the sustainability of the supply chain within the CE strategy were analyzed by [27]. Ref. [28] evaluated and analyzed the risks associated with implementing CSC practices with MCDA, while [29] proposed calculating a risk awareness indicator. Moreover, the transition from a linear economy to a CE was examined by [30], and CSC was analyzed by [31].

2.2. Circular Supplier Selection

Selecting suppliers within a CSC requires a shift from traditional procurement criteria toward a more comprehensive evaluation that aligns with CE principles. Beyond cost, quality, and delivery performance, CSS must prioritize resource efficiency, waste minimization, closed-loop logistics, and regulatory compliance. One of the main challenges is ensuring that suppliers have the necessary capabilities to operate in a circular system. This includes remanufacturing expertise, reverse logistics infrastructure, material recovery processes, and adherence to sustainability standards. Additionally, external factors such as policy incentives, financial constraints, and technological limitations can impact the feasibility of transitioning to a fully CSC.

Risk evaluation is vital in shifting to circular models, addressing supply, market, and operational uncertainties. Organizations should assess suppliers’ adaptability and sustainability commitment. A robust selection process ensures suppliers align with CE goals while addressing feasibility and mitigating risks. This selection process focuses on evaluating suppliers based on key criteria, such as resource efficiency, waste minimization, and environmental sustainability. By embedding the CE principles, businesses can align their supply chains with both economic and operational objectives, while simultaneously advancing broader environmental and social goals. This strategy not only fosters long-term sustainability but also enhances the resilience of CSC.

In this context, CSS and order allocation in Closed-loop Supply Chains address inventory-location-routing problems and sustainable evaluation methods [1,32,33]. Key criteria and applications in manufacturing were also studied [34,35,36]. Additionally, Supplier evaluation for CE, selection, and allocation, as well as green criteria development, have been explored [37,38,39,40,41]. Furthermore, an approach for reliable CSS and CLSC network design, emphasizing collaborative costs, shortages, and circular criteria, was developed by [42].

The transition from traditional linear supply chains to CSC is a key step in promoting sustainable production and consumption. Researchers have explored the challenges, opportunities, and potential risks involved in adopting CSCs, with a focus on financial, technological, and operational challenges [15,16]. Additionally, studies have examined how suppliers are selected within CSC, highlighting factors like reducing waste, improving the efficient use of resources, and complying with regulations [36,37]. For instance, suppliers can be assessed based on their ability to implement reverse logistics processes, such as the collection and recycling of end-of-life products [43].

Several types of research reinforce the need for better supplier selection procedures by employing strategies that manage uncertainty. Quantitative and qualitative approaches were utilized to improve the robustness of the assessment, with comparative case studies highlighting advantages over other procedures [44].

Furthermore, increasing research is arising involving frameworks that merge sustainability, CE, and supplier viability criteria in decision-making. For instance, Integrated framework that merges advances computational with improvement approaches have shown potential to strengthen supplier performance evaluation across multiple industries [45]. Additionally, integration supplier assessment with sustainability objectives, some approaches incorporate—environmental, economic, and social—through decision-making methods. These strategies contribute to advancing supplier selection management with sustainability objectives [46].

Despite these advancements, limited research has integrated a MCDA approach to evaluating CSS. While earlier studies have highlighted important sustainability criteria, a well-defined framework for decision-making remains largely unexplored. addresses this gap by proposing a comprehensive MCDA-based evaluation model that aligns supplier selection with CE principles. By integrating risk assessment, this research contributes to a more robust and strategic method for evaluating suppliers in CSC.

The MCDA approach to CSS enables a systematic evaluation based on multiple criteria, such as resource efficiency, waste management, and environmental sustainability. This process may involve methods such as AHP or Analytic Network Process (ANP), a decision-making framework to rank suppliers according to these factors [47,48].

Supplier Risk Assessment and SDGs Alignment

As sustainability objectives become a strategic priority, risk assessment has emerged as a critical element in supply chain decision-making, particularly those outlined by the SDGs. Among these, SDG’s Target 12.5 is strongly connected to supplier performance in environmental, economic, and operational dimensions. The current literature underscores the necessity of integrating risk-based thinking into sustainable supplier selection frameworks. For instance, ref. [49] proposes an SDG-informed multi-period decision model that incorporates supplier-related risks using fuzzy Failure Mode and Effects Analysis (FMEA), optimizing for both cost and risk while promoting SDG’s Targets 8 and 12.5. This highlights how supplier selection decisions can be aligned with global development priorities by recognizing and mitigating risks from social, environmental, and economic dimensions.

Within high-risk sectors such as the chemical industry, sustainability-linked supplier risks are particularly salient. Ref. [50] employs a hybrid MCDA approach, tailored to the “high-risk, high-pollution, and high-efficiency” nature of chemical operations, in order to identify risk-prone stages in supplier management. Similarly, ref. [51] combines spherical fuzzy (SF), AHP and combined compromise solution (CoCoSo) methods to evaluate supplier sustainability, emphasizing the importance of considering risk factors when selecting suppliers in hazardous and volatile industrial contexts.

Social sustainability risks are also increasingly incorporated into supplier assessment practices. Ref. [52] notes a shift in socially responsible purchasing strategies, wherein supplier development activities are initiated before supplier selection to proactively mitigate risks associated with non-compliance or unethical practices. This repositioning not only strengthens governance but also reduces the likelihood of supply chain disruptions stemming from unsustainable supplier behavior. Ref. [53] further explore the relationship between supplier risk and sustainability in the fashion industry, proposing a quality function deployment (QFD)-based framework that incorporates resilience capabilities—such as resource reallocation and shared risk responsibility—as mechanisms to counteract vulnerabilities like supplier delays, political instability, and poor material quality.

Systemic and relational risks are also addressed in broader strategic frameworks. In ref. [54], through a systematic review, it is highlighted that supplier development strategies must deliver Triple Bottom Line (TBL) outcomes while acknowledging the knowledge risks that arise when TBL principles are not fully understood or integrated. In parallel, ref. [55] maps organizational barriers—such as resistance to change and implementation failures—that represent internal risks affecting the transition to sustainable production systems.

Moreover, governance-related risks stemming from market dynamics are gaining attention. Ref. [56] identifies how customer concentration influences supplier risk-taking behaviors, revealing governance-based mechanisms that alter a firm’s sustainability posture. Ref. [57] contributes to this by calculating country-specific sustainability risk scores based on SDG alignment and legal compliance, emphasizing how the geographic and institutional context of suppliers can significantly impact procurement risk exposure.

Finally, methodological critiques have also emerged. Ref. [58] analyzes the limitations of commonly used tools like risk matrices in supply chain decision-making. They argue that inadequate information quality can impair risk perception and judgment, reinforcing the need for more robust, SDG-aligned assessment tools in supplier management. In summary, these studies illustrate a multidimensional understanding of supplier-related risks that span compliance, operational, and strategic concerns. When integrated into supplier evaluation frameworks, such risk assessments not only support SDG’s Target 12.5 but also promote more resilient, responsible, and forward-looking supply networks.

2.3. Circular Supplier Selection Criteria

Selecting suppliers that adhere to the CE principles is a crucial step in implementing a CSC. The selection criteria should encompass environmental, economic, and operational factors to ensure resource efficiency and minimize negative environmental impacts. One of the primary goals in this process is to align supplier selection with SDG’s Target 12.5, emphasizing the need for suppliers to adopt practices that contribute to waste reduction through material recovery, closed-loop systems, and sustainable production methods. By incorporating MCDA methods, it is possible to systematically evaluate suppliers based on their ability to meet CSC requirements, ensuring that procurement decisions align with sustainability objectives.

This section provides an overview of the selection criteria, focusing on the most crucial factors influencing the sustainability and circularity of the supply chain. It is essential to balance subjective and objective criteria, integrating both analytical and intuitive thinking to provide a comprehensive evaluation [8]. Furthermore, only the necessary amount of information and analysis should be employed to resolve dilemmas, preventing decision fatigue or delays. Moreover, this approach aims to substantially reduce waste generation by 2030 through prevention, reduction, recycling, and reuse. The following criteria can be considered when evaluating suppliers in a CSC (Figure 1).

2.3.1. Waste Prevention and Reduction Criteria

This subsection discusses key strategies for waste prevention and reduction, focusing on eco-design practices, material efficiency, process optimization, and waste reduction: Eco-design practices focus on how suppliers integrate sustainable product design to minimize waste at the source. Waste management is the main challenge in CSC. Ref. [59] adds value to identifying and ranking barriers in the copper industry by applying systematic decision-making methods. Moreover, it contributes to waste management policies and strategies for collaboration between suppliers, decision-makers, and corporations, to minimize waste and improve material efficiency. Similarly, ref. [60] manages the complexities of waste-to-energy solutions, introducing a strategic framework seeking lower emissions and promoting cleaner energy cycles that employ circular practices.

This includes adopting design strategies that facilitate material reuse, recyclability, and extended product life cycles. A collaborative eco-design approach can enhance sustainability by engaging stakeholders and evaluating secondary raw materials for improved recycling integration [61].

Material efficiency focuses on the optimization of raw material use, and suppliers should prioritize strategies that reduce material consumption while maintaining functionality and performance. This includes efficient packaging solutions that minimize waste and improve life cycle assessments. Material selection plays a crucial role, as differences in weight and composition can significantly impact resource efficiency and sustainability metrics [62]. Suppliers adopting lean manufacturing principles optimize processes, eliminate inefficiencies, and minimize waste. Continuous improvement strategies enhance efficiency and align with CE goals [63].

2.3.2. Recycling and Reuse Capabilities

This subsection discusses key strategies for the use of recycled or recovered materials, end-of-life product management, and reverse logistics capabilities: The use of recycled or recovered materials should be encouraged as suppliers incorporate these into their production processes to reduce dependency on virgin resources. Considering industrial synergy and product lifecycle extension, [64] examines how circular principles and collaborative business models can enhance e-waste recovery. Their research demonstrated that CSC—when supported by finance incentives—lead to better returns, lower emissions, and more effective reuse of electronic products. Integrating life cycle assessments enables cost-effective product development while reducing CO₂ emissions. Recycling end-of-life products is essential for emissions reduction and resource efficiency in a CE [65].

End-of-life product management encourages suppliers to implement product take-back programs, remanufacturing, and refurbishment initiatives to extend product life cycles and minimize waste. Incorporating recyclability into product design enhances material recovery and promotes sustainable resource use. Structural design strategies, particularly in industries like automotive manufacturing, facilitate efficient dismantling and material reuse [66].

Reverse logistics capabilities require suppliers to possess infrastructure and processes to facilitate material recovery, remanufacturing, and reintegration into the supply chain. A well-structured CLSC is fundamental to achieving circularity, requiring suppliers to demonstrate operational efficiency, system management, and circular waste management. Additionally, effective coordination enhances the implementation of reverse logistics processes [67].

2.3.3. Circular Business Model Alignment

This subsection discusses key strategies for CLSC integration, Industrial symbiosis participation and Product-as-a-service (PaaS) adoption: CLSC integration ensures that suppliers demonstrate the ability to retain materials in circulation through reprocessing, refurbishment, and reuse. A multi-period closed-loop supply chain approach enables strategic and operational decisions regarding network design, inventory control, production planning, supplier selection, and service level requirements, ensuring the efficient reintegration of recovered materials [68].

A quantitative approach introduced by [69] considered CE metrics integrating resiliency and market perspectives to optimize both supplier selection and order allocation. Utilizing key metrics and multicriteria models, their research contributes to the design of circular business models that enhance resource allocation. Ref. [70] aligns with this view by assessing how players and emerging technologies shape the circular transformation of supply chains.

Industrial symbiosis participation involves suppliers engaging in collaborative networks to repurpose waste streams as raw materials, enhancing resource efficiency and sustainability. Industrial Symbiosis fosters CE principles by facilitating exchanges of materials, expertise, and services, reinforcing the commitment to environmental sustainability and CLSC practices [71].

PaaS adoption enables suppliers to implement leasing, pay-per-use, and subscription-based models to extend product lifecycles and reduce resource consumption. The adoption of PaaS is crucial for the CE, particularly in emerging economies, where consumer acceptance of remanufactured products and service-based models presents both challenges and opportunities. Understanding consumer behavior in this context can guide businesses and policymakers in fostering the CE practices [72].

2.3.4. Environmental Compliance and Certifications

This subsection discusses key strategies for regulatory adherence, third-party sustainability certifications, and transparent reporting on circularity metrics: Regulatory adherence requires suppliers to comply with waste management regulations and the CE standards as part of their sustainability commitment. The sustainable supply chain management approach evaluates suppliers based on criteria such as green initiatives, waste reduction practices, and regulatory compliance, enhancing both environmental performance and operational efficiency. The adoption of SSCM not only benefits business operations but also fosters supply chain circularity, encouraging the integration of emerging technologies to strengthen sustainable practices [73].

Third-party sustainability certifications require suppliers to hold recognized environmental certifications, such as ISO 14001 (Environmental Management), cradle to cradle, or zero waste to landfill, demonstrating their commitment to sustainability. Additionally, managerial commitment to sustainability, supplier compliance with ISO 14000 standards, and practices such as waste heat recovery are key factors in supplier evaluation [74].

Transparent reporting on circularity metrics requires suppliers to disclose key CE performance indicators, including waste reduction, resource recovery rates, and material reintegration. Transparent reporting enhances stakeholder trust and regulatory compliance, facilitating informed decision-making in sustainable supply chain management. The use of standardized sustainability metrics allows for comparative assessments and continuous improvements in circular practices [75].

2.3.5. Innovation and Technological Adoption

This subsection discusses key strategies for advanced recycling technologies, digitalization for circularity, and sustainable material substitutions: Advanced recycling technologies assess suppliers based on their investment in cutting-edge recycling and remanufacturing methods to enhance material recovery and waste reduction. The adoption of innovative recycling solutions contributes to CE goals, while emerging technologies, such as blockchain, support traceability and efficiency in closed-loop systems [45,46,76]

Digitalization for circularity applies artificial intelligence, internet of things, and blockchain to monitor, track, and optimize resource flows within CSC. Digital technologies and address barriers to product circularity, enhance traceability to CE practices, benefiting both environmental sustainability and business performance [77].

Adoption of biodegradable, recyclable, or low-impact materials to enhance circularity and reduce environmental impact. Future material selection and design should integrate sustainability metrics, engage stakeholders across the supply chain, and incorporate green chemistry principles. This approach supports materials circularity, alternative assessments, and alignment with global sustainability policies [78].

2.4. Literature Gaps

In the supplier selection process, uncertainty is intrinsic, which allows for wide opportunities to solve problems with different approaches, concepts, methods and even the application of algorithms in decision-making environments [69,79]. Furthermore, a controlled process is a requirement when it involves partners both in supplier selection and in monitoring circularity activities through performance measurement [45,70].

There is a need to combine diverse applications such as artificial intelligence, and game theory to address complex realities related to environmental sciences, and renewable energy [44,79]. Equally important, studies on how sustainability-driven supplier selection through integrated decision-making tools can be explored mainly by considering regulations [46,59]. It is possible to highlight the need for other approaches that involve the combination of lean production tools in a CSC, mainly in eliminating and avoiding waste [59]. Ref. [64] highlighted waste-to-energy, as well as the need for data reliability, in addition to models that consider uncertainty in dynamic demand scenarios.

Given the inherent uncertainty and complexity in supplier-related decisions, academic contributions point to several unresolved issues in the integration of risk management into supplier selection and allocation decisions. Ref. [49] highlights the need to enhance existing frameworks by incorporating stochastic demand modeling, scalability for large datasets, and real-time risk profiling, especially in volatile and global supply chain environments. However, risk dimensions such as environmental and social impacts, geopolitical exposure, and uncertainty from emerging technologies like IoT and blockchain remain under-addressed in practical decision models.

Ref. [56] identifies gaps in understanding how customer concentration shapes corporate risk behavior, particularly regarding credit and investment risk in supplier relationships. Ref. [51] also signals the need for adaptable decision methods that respond to high-risk, crisis-prone contexts—yet few studies systematically connect MCDA approaches with dynamic risk scenarios. Similarly, ref. [54] reveals that supplier development strategies face operational and strategic risks, especially when monitoring and feedback systems are not effectively integrated to sustain TBL outcomes.

In sector-specific studies, ref. [50,53] proposes risk-informed models for industries with critical vulnerabilities such as fashion and chemicals, though broader frameworks connecting disruption risk, sustainability goals, and strategic supplier assessment are still limited. Ref. [7] adds that scenario-based approaches and resilience strategies under uncertainty are promising but not yet consolidated across diverse industrial settings. Therefore, there remains a clear opportunity to advance the literature by incorporating comprehensive risk management mechanisms into supplier-related decision-making—especially in alignment with sustainability imperatives.

3. Methodology

Aligning supplier selection with SDG’s Target 12.5 requires a structured approach that prioritizes circular suppliers and waste reduction. To develop the proposed model, this section first presents the AHP, which is used to prioritize supplier selection criteria identified in the literature review. Additionally, risk assessment and adjustment formulas are incorporated into the AHP framework to enhance its robustness, ensuring a more structured and reliable decision-making process.

The AHP is one of the most successfully applied MCDA methods for supplier selection [80]. According to [81], the AHP is comparable the human mind functioning when facing a complex situation composed by various elements, these are reunited in groups according to their common characteristics. Such groups can then be grouped at a superior level, with another set of common characteristics, until a maximum level is reached, which constitutes the final objective of the decision-making process.

AHP’s foundations include the Fundamental Scale of Absolute Numbers [81]. The process involves creating a pairwise comparison matrix A, followed by using Linear Algebra concepts to determine the eigenvector w and eigenvalue ${\lambda }_{max}$. This enables the derivation of relative priorities, providing a robust framework for decision-making. In AHP, priorities are determined by applying the Perron-Frobenius theorem, as outlined by [81] and expressed in Equation (1).

$$Aw={\lambda }_{max}w$$

(1)

Consistency is an essential property of the matrix $A$. If $A$ exhibits consistency in its comparisons, then $a_{ij}=w_i/w_j$ for $i,j=\mathrm{1,2},\dots n$ where $n$ is the order of $A$. In this manner, $a_{ij}=a_{ik}{a}_{kj}$. If $A$ is not a perfectly consistent matrix, then ${\lambda }_{max}>n$. The Consistency Index $CI$, calculated by Equation (2), serves as a measure of the deviation between ${\lambda }_{max}$ and $n$:

$$CI=({\lambda }_{max}-n)/(n-1)$$

(2)

The Consistency Ratio $CR$, calculated using Equation (3), also considers a Random Index $RI$ associated with $n$. If the $CR$ exceeds 0.10, a review of the comparisons may be necessary [71].

$$CR=CI/RI$$

(3)

AHP application occurs in many areas and subjects, among which supply chain management can be highlighted. MCDA methods have been developed to aid in the essential managerial task of decision-making. According to the contingency approach, these methods must consider situational factors that can affect the results of the decision-making process. As an MCDA method, AHP was structured to support the formulation and analysis of decisions. With this method, it is possible to break complex problems into components according to a given hierarchy, which facilitates the analysis and solution of the problem [10]. The application of AHP can be summarized in a few steps, including hierarchy construction, pairwise comparison, consistency verification, and results interpretation [82]. AHP applications are presented by [83].

One of the justifications for choosing AHP is that it has been a leading MCDA method for decades [80,84,85]. Another justification is that it is one of the most widely used MCDA methods in the context of the CE [86]. When contrasted with other approaches for ranking and choice problems [87], AHP brings a logically structured based on model that enables pairwise comparisons and consistency checks, which is particularly significant when using expert judgments. Although techniques like TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) and ELECTRE (ELimination and Choice Expressing Reality) are useful in ranking alternatives, they often require precise quantitative input or equalization techniques that may suggest bias. AHP, alternatively, ensures qualitative evaluation and enhances transparency in weighting and ranking, suitable for decision problems involving sustainability and circularity [88,89].

Despite this, the integration of AHP with a risk matrix is not a novel approach and has been applied in several prior studies [49,90]. However, our rationale for selecting standard AHP over more advanced methods such as fuzzy AHP or hybrid MCDA models lies in the clarity, transparency, and applicability of the traditional AHP framework in the context of the present study. The use of standard AHP allows for a consistent, reproducible, and practitioner-friendly structure that is especially relevant when communicating results to stakeholders with limited technical background.

While fuzzy AHP provides enhanced capability to handle uncertainty in subjective judgments [91,92], it introduces additional complexity using fuzzy numbers and membership functions, which may compromise interpretability. In contrast, the standard AHP—when supported by a rigorous consistency analysis—offers sufficient robustness for modeling expert preferences in structured decision contexts, such as the prioritization of circularity criteria under defined risk dimensions.

Moreover, this study integrates AHP with a risk-adjusted performance matrix, which enhances the analytical richness by incorporating contextual risk factors into the prioritization process. This approach, while methodologically established, remains effective and aligns with the applied objectives of this research, focusing on practical implementation rather than methodological novelty. Finally, it is recognized that the use of fuzzy AHP or other hybrid models could offer added value in contexts involving greater ambiguity or heterogeneous expert opinions, and this is suggested as a promising direction for future research.

Following the AHP ranking, suppliers undergo a risk assessment based on a risk matrix, which evaluates probability versus impact for key risks. These include financial risks (e.g., bankruptcy and economic instability), environmental risks (e.g., regulatory non-compliance and low material traceability), and operational risks (e.g., delays and insufficient production capacity).

Next, AHP results are integrated with the risk matrix, incorporating an adjustment based on risk Equation (4). Additionally, a threshold for acceptable risk is established, defining the maximum allowable risk score. Suppliers exceeding this limit are disqualified.

$$Adjusted Score=Overall\times (1-\alpha \times Normalized Risk Score)$$

(4)

where

• Overall: result from the AHP.
• Normalized risk score: risk value normalized to a 0–1 range to avoid distortions.
• α (penalty factor): adjustment parameter to calibrate risk impact (e.g., 0.05 or 0.1). This adjustment is applied to the final AHP score to incorporate the risk profile of each alternative. The detailed formulas, rationale, and illustrative examples of this adjustment are presented in Section 4.

The proposed model integrates

• AHP for selecting circular suppliers based on multiple criteria.
• Risk matrix for evaluating threats and uncertainties.
• AHP-risk integration to support decision-making and mitigate uncertainty.

By combining AHP with a well-designed risk adjustment mechanism, this methodology advances traditional approaches to supplier selection by embedding considerations of uncertainty and resilience into the decision-making process. Although earlier studies have utilized AHP for evaluating suppliers, the proposed method enhances this process by systematically revising rankings based on exposure to risks. This ensures that the final selection of suppliers adheres to both CE principles and effective risk management practices. This integrated approach strengthens decision-making, providing a robust framework that surpasses conventional applications of MCDA in fostering supply chain sustainability and waste reduction.

To avoid redundancy in the evaluation process, although the same group of experts participated in both the performance assessment and the risk evaluation, the process was carefully structured in sequential phases. Experts were instructed to focus solely on performance-related criteria during the AHP-based prioritization and to consider risk dimensions (operational, financial, and environmental) only in the final stage. This approach followed the company’s internal policy to standardize evaluations and reduce subjectivity in the integration of risk into decision-making.

4. A Decision Model for Circular Supplier Selection

Several case studies have demonstrated the value of applying MCDA methods to supplier selection problems in real-world CE contexts. For instance, ref. [93,94] evaluates green suppliers, while [15,95,96] incorporated sustainability into supplier selection. These applications highlight the importance of adapting decision models to sustainability goals.

This research examines a Fortune 500 company recognized as one of the world’s leading chemical producers and a major supplier of crop protection products in Latin America. The company operates across five key global regions: Europe, North America, South America, Asia, and the Pacific, focusing on research, development, manufacturing, and supply of chemical solutions for various markets. Its Latin American headquarters are located in Brazil, a region considered strategic due to its large agricultural market. The company operates more than ten production sites and several logistics hubs in the region, supporting efficient distribution. It also emphasizes sustainable supply chain practices and environmental risk mitigation across its operations.

In this context, sustainable supply chain management has become a strategic priority for the company, with an emphasis on integrated and self-optimizing systems that enhance resource efficiency and waste reduction. The company prioritizes minimizing waste generation through initiatives that promote material reuse, recycling, and circularity.

The research assessed experts and supply chain managers’ alignment with CE principles. Participants were consultants with over 15 years of experience in supply chain and sustainability, working in industrial sectors like chemical manufacturing across the Americas. All experts hold, at least, a bachelor’s degree in Business Administration (BA), Chemistry (C), Environmental Science (EI), or Industrial Engineering (IE), which is detailed in

Table 1. Experts profile overview.

Position	Number of Experts	Work Experience	Higher Education Field
Consultant	5	Up to 10 years	BA
Consultant	2	More than 15 years	C
Consultant	2	Up to 5 years	ES
Consultant	3	More than 15 years	IE

Source: our own elaboration.

.

The objective of the proposed model is to select suppliers. This process was carried out in a single, focused, in-person session, enabling direct interaction and collaborative engagement. At the beginning of the session, the AHP stages were explained, and key concepts were translated for the experts. This interaction helped refine the selection process, ensuring a well-defined set of criteria and sub-criteria. Based on their discussion, the experts proposed the criteria and sub-criteria by integrating insights from relevant literature (Section 2.3).

This collaborative effort ensured that theoretical foundations and practical insights were effectively integrated, making the criteria both comprehensive and contextually relevant. The identified supplier selection criteria reflect key aspects of circularity and sustainability, aligning with SDG’s Target 2.5 by emphasizing waste prevention, recycling, and resource optimization. The criteria were selected from the literature review in Section 2.3, including waste prevention and reduction, recycling and reuse capabilities, circular business model alignment, environmental compliance and certifications, innovation, and technological adoption. The sub-criteria were structured as follows (

Table 2. Main criteria and sub-criteria.

Main Criterion	Sub-Criterion Code	Sub-Criterion	Focus and Objectives	Reference
2.3.1. Waste Prevention and Reduction	W1	Eco-design Practices	To assess suppliers’ integration of sustainable product design to minimize waste at the source and facilitate product lifecycle extension.	[61]
	W2	Material Efficiency	To evaluate suppliers’ optimization of raw material use, reducing consumption while maintaining functionality and performance.	[62]
	W3	Process Optimization and Waste Reduction	To measure suppliers’ implementation of lean manufacturing principles to streamline processes, eliminate inefficiencies, and minimize waste generation.	[59,60,63]
2.3.2. Recycling and Reuse Capabilities	R1	Use of Recycled or Recovered Materials	To determine the extent to which suppliers incorporate recycled or recovered materials into their production processes, reducing reliance on virgin resources.	[64,65]
	R2	End-of-Life Product Management	To assess suppliers’ initiatives in product take-back programs, remanufacturing, and refurbishment to extend product lifecycles and minimize waste.	[66]
	R3	Reverse Logistics Capabilities	To evaluate suppliers’ infrastructure and processes for material recovery, remanufacturing, and reintegration into the supply chain.	[67]
2.3.3. Circular Business Model Alignment	C1	CLSC Integration	To verify suppliers’ ability to retain materials in circulation through reprocessing, refurbishment, and reuse within a closed-loop system.	[68,69,70]
	C2	Industrial Symbiosis Participation	To assess suppliers’ engagement in collaborative networks that repurpose waste streams as raw materials for other industries.	[71]
	C3	Product-as-a-Service (PaaS)	To examine suppliers’ adoption of leasing, pay-per-use, or subscription-based models to extend product lifecycles and reduce resource consumption.	[72]
2.3.4. Environmental Compliance and Certifications	E1	Regulatory Adherence	To ensure suppliers’ compliance with waste management regulations and CE standards as part of their sustainability commitment.	[73]
	E2	Third-Party Sustainability Certifications	To evaluate suppliers’ possession of recognized environmental certifications	[74,97]
	E3	Transparent Reporting on Circularity Metrics	To assess suppliers’ disclosure of key CE performance indicators, such as waste reduction and resource recovery rates.	[75]
2.3.5. Innovation and Technological Adoption	T1	Advanced Recycling Technologies	To evaluate suppliers’ investment in cutting-edge recycling and remanufacturing methods for enhanced material recovery and waste reduction.	[45,46,76]
	T2	Digitalization for Circularity	To assess suppliers’ application of digital technologies (AI, IoT, blockchain) to monitor, track, and optimize resource flows within CSC.	[77]
	T3	Sustainable Material Substitutions	To examine suppliers’ adoption of biodegradable, recyclable, or low-impact materials to enhance circularity and reduce environmental impact.	[78]

Source: our own elaboration.

):

In this model, suppliers A, B, C, D, and E represent the alternatives to be evaluated. The decision-making process aims to identify the supplier that best aligns with the established circularity criteria and contributes to waste reduction and resource efficiency (Figure 2).

In this stage of the research, the pairwise comparisons among the criteria were assigned through consensus among the experts. Concerning waste prevention and reduction, recycling and reuse capabilities, circular business model alignment, environmental compliance and certifications, innovation, and technological adoption criteria, the experts agreed that a balance among them is necessary for the supplier selection to be successful. Therefore, they assigned a weight of 0.200 to each criterion, ensuring that the total sum of weights amounts to 100%. The pairwise comparisons among the sub-criteria were assigned through consensus among the experts. Local priorities for all attributes can be determined by normalizing the right eigenvector of the combined comparison matrices. Once judgments for criteria and sub-criteria were made, the overall priorities were computed by multiplying each criterion’s priority values by the sub-criteria weights. The local and overall priorities of the criteria are presented in

Table 3. Local and overall priority.

Criteria and Sub-Criteria	Local	Overall
Waste prevention and reduction	0.200	0.200
W1 Eco-design practices	0.163	0.033
W2 Material efficiency	0.297	0.059
W3 Process optimization and waste reduction	0.540	0.108
Recycling and reuse capabilities	0.200	0.200
R1 Use of recycled or recovered materials	0.179	0.036
R2 End-of-life product management	0.709	0.142
R3 Reverse logistics capabilities	0.113	0.023
Circular business model alignment	0.200	0.200
C1 CLSC integration	0.701	0.140
C2 Industrial symbiosis participation	0.202	0.040
C3 Product-as-a-service	0.097	0.019
Environmental compliance and certifications	0.200	0.200
E1 Regulatory adherence	0.135	0.027
E2 Third-party sustainability certifications	0.281	0.056
E3 Transparent reporting on circularity metrics	0.584	0.117
Innovation, and technological adoption	0.200	0.200
T1 Advanced recycling technologies	0.243	0.049
T2 Digitalization for circularity	0.088	0.018
T3 Sustainable material substitutions	0.669	0.134

Source: our own elaboration. Note: In this framework, the term “product” includes both tangible goods and intangible services, acknowledging that service-oriented models such as product-as-a-service are also forms of product delivery with circular potential.

.

The overall priorities for sub-criteria R2 (End-of-life product management) are highly prominent in the category, reaching 14%, associated with the recycling and reuse capabilities. C1 (CLSC integration) demonstrates a significantly greater contribution, reaching 14%, in circular business model alignment. T3 (Sustainable material substitutions) presents the greatest contributions, reaching 13%, in Innovation, and technological adoption. E3 (Transparent reporting on circularity metrics) presents the greatest contribution, reaching 12%, in environmental compliance and certifications. W3 (Process optimization and waste reduction) presents the greatest contribution, reaching 11%, in waste prevention and reduction. However, its contribution is low in W1 (Eco-design practices) and W2 (Material efficiency) in waste prevention and reduction, R1 (Use of recycled or recovered materials) and R3 Reverse logistics capabilities in Recycling and reuse capabilities, C2 (Industrial symbiosis participation) and C3 (Product-as-a-service) in Circular business model alignment, E1 (Regulatory adherence) and E2 (Third-party sustainability certifications) in Environmental compliance and certifications, T1 (Advanced recycling technologies) and T2 (Digitalization for circularity) in Innovation, and technological adoption. The comparison of the sub-criteria can be observed in Figure 3.

An advantage of using ratings is the opportunity to avoid biases. When comparing alternatives pairwise (relative measurement), certain historical trends need to be considered. Comparing alternatives against a standard (absolute measurement) appears to provide a less partial or more unbiased measurement. The level of performance corresponding to the attributes in linguistic scales varies from ‘weak’ to ‘excellent’. The adoption of absolute measurement is due to its potential to significantly reduce conflicts in decision-making processes [81]. In this research, performance levels were used, and alternatives were compared individually to a scale, such as excellent (1.000), very good (0.830), between good and very good (0.670), good (0.500), between weak and good (0.250), and weak [98]. The overall priorities of the alternatives are presented in

Table 4. Overall priorities of the alternatives.

Alternative	Overall
A	0.710
B	0.700
C	0.675
D	0.724
E	0.688

Source: our own elaboration.

.

Following the development of the decision model for CSS using the risk matrix (

Table 5. Risk matrix.

Probability			Impact
Very high = 5	5	10	15	20	25
High = 4	4	8	12	16	20
Average = 3	3	6	9	12	15
Low = 2	2	4	6	8	10
Very Low = 1	1	2	3	4	5
	Very Low = 1	Low = 2	Average = 3	High = 4	Very high = 5

Source: our own elaboration. Note: Colors indicate risk levels — green for low, yellow for moderate, and red for high risk.

). The risk matrix (Table 5) was established based on expert insights, considering probability and impact. The risk matrix is the result of the product of probability and impact. The following range was considered for probability and impact: very low = 1 to very high = 5.

The risk categories—financial, environmental, and operational—were defined to capture key challenges in CSS. Environmental risks encompass issues such as low material traceability and regulatory non-compliance, which directly impact circularity. Operational risks include barriers such as delays and insufficient production capacity, as well as challenges in reverse logistics and closed-loop supply chain integration.

The numerical parameters reflect a team consensus approach, ensuring a balanced interpretation of risk severity in the decision model. For each alternative and risk category, a raw risk score was calculated by multiplying the assigned probability and impact, calculated by Equation (5).

$${Raw risk score}_{ij}={Probability}_{ij}\times {Impact}_{ij}$$

(5)

where i represents the alternative and j the risk categories financial, environmental, and operational. These raw scores were then normalized by dividing by the maximum possible score (25) to ensure comparability calculated by Equation (6).

$${Normalized risk score}_{ij}={\displaystyle \frac{{Raw risk score}_{ij}}{25}}$$

(6)

For instance, replacing values in Equation (5), for Alternative A (financial risks), we obtain a raw risk score of 3 × 2 = 6. Similarly, by applying the values in Equation (6), the normalized risk score yields 6/25 = 0.24. This structured risk assessment supports a comprehensive evaluation, aligning with the principles of CSC (

Table 6. Risk scoring of alternatives.

Alternative	Risks	Normalized Score
A	Financial risks	0.24
	Environmental risks	0.32
	Operational risks	0.24
B	Financial risks	0.64
	Environmental risks	0.24
	Operational risks	0.16
C	Financial risks	0.24
	Environmental risks	0.16
	Operational risks	0.64
D	Financial risks	0.08
	Environmental risks	0.48
	Operational risks	0.24
E	Financial risks	0.12
	Environmental risks	0.20
	Operational risks	0.16

Source: our own elaboration.

).

After obtaining the normalized score, the penalty factor α = 0.1 is adopted to adjust the risk impact. For this study, the value α = 0.1 was selected based on expert judgment during the assessment process. Replacing values in Equation (4), for Alternative A we have the adjusted score = 0.710 × (1 − 0.1 × 0.32) = 0.688. The same procedure was performed for the other alternatives. The risks integration with AHP and the adjusted score for alternatives is presented in

Table 7. Risks integration with AHP and adjusted scores for alternatives.

Alternative	Overall	Adjusted Score	Rank
A	0.710	0.688	2
B	0.700	0.656	4
C	0.675	0.633	5
D	0.724	0.690	1
E	0.688	0.677	3

Source: our own elaboration.

.

Initially, the AHP results (overall) were used to rank the alternatives. However, after incorporating the adjusted score, noticeable shifts in rankings occurred. While alternative D retained its leading position, other alternatives—A, E, B, and C—experienced significant changes. These adjustments stem from the risk factors associated with each alternative, emphasizing their influence on the final ranking. By integrating risk considerations, the evaluation process provides a more nuanced prioritization, ensuring that decision-making accounts for potential uncertainties. This underscores the importance of applying a penalty factor to refine the assessment of risk impact, leading to more informed and resilient supplier selection decisions. Moreover, this approach supports a sustainable supply chain by encouraging the selection of suppliers that minimize waste and enhance resource efficiency, reinforcing the CE principles within the CSC.

A sensitivity analysis was conducted to evaluate the robustness of the decision model. For this purpose, the criterion waste prevention and reduction was selected due to its recognized relevance in CSC. While the choice of this criterion is illustrative, it allows us to observe how variations in its weight—from the original 20% up to 50%—affect the final scores of the alternatives. The corresponding results are shown in Figure 4.

Figure 4 illustrates the impact of increasing the weight of the criterion waste prevention and reduction from 20% to 50%. The black vertical line represents the original weight, while the red line indicates the adjusted weight. Notably, the ranking of alternative D remained unchanged, demonstrating its robustness under the weight variation. Meanwhile, alternative C shifted positions with alternative B, indicating a sensitivity to this criterion. Alternatives A and E maintained their original positions, suggesting moderate resilience to this change. These results should not be broadly generalized, as they are based on a specific case context. Future research could extend the findings by applying the model to different organizations or sectors, thereby enhancing its external validity.

5. Discussion

The proposed model for supplier selection requires a structured approach that prioritizes circularity and waste reduction. It emphasizes the need for suppliers to adopt practices that minimize waste through material recovery, closed-loop systems, and sustainable production methods.

The criteria used in this study include waste prevention and reduction, recycling and reuse capabilities, circular business model alignment, environmental compliance and certifications, innovation, and technological adoption. Additionally, 15 relevant sub-criteria were identified and integrated into the AHP calculation, enabling the evaluation of the five alternatives defined in the model.

Moreover, an important advancement of the proposed model was the integration of risk assessment into the decision-making process. The model considers financial, environmental, and operational risk categories, capturing key challenges in CSS. This addition enhances the robustness of the model, making it applicable to diverse industry contexts and particularly valuable for CSC practitioners and researchers.

From a theoretical perspective, this research aligns with recent studies such as [99,100], which addressed CE and supplier selection. This work makes a significant contribution to the study of CSS by addressing the gap in MCDA methods that explicitly incorporate risk-adjusted prioritization. While earlier studies have employed AHP for supplier evaluation, they often overlook the dynamic risks that can influence decisions in CSC. By introducing a structured mechanism for risk adjustment, this study enhances the applicability of AHP in uncertain environments, offering a more comprehensive decision-making model that integrates sustainability objectives with operational resilience. Additionally, by aligning supplier selection criteria with the SDG’s Target 12.5, this research underscores the importance of structured decision frameworks in advancing circularity. It presents an innovative approach that moves beyond traditional economic and environmental considerations, systematically incorporating risk as a central factor in CSC management.

From a managerial perspective, the proposed model offers a methodical and data-centric approach to supplier selection, enabling decision-makers to integrate CE principles into CSC strategies. This integration not only enhances supply chain sustainability but also mitigates risks associated with regulatory compliance, environmental impact, and resource scarcity. Furthermore, incorporating risk assessment fortifies the decision-making process, providing managers with a comprehensive framework to balance sustainability goals with operational and financial considerations. The model’s versatility allows its application across various industries, aiding organizations in transitioning towards more responsible and resilient supply chains.

While MCDA methods such as AHP, TOPSIS, and Multi-Criteria Optimization and Compromise Solution (VIKOR) have been widely used for supplier evaluation, their application in circular contexts remains limited in addressing core dimensions such as systemic circularity, risk integration, and real-world applicability [101,102,103]. Unlike conventional approaches that often include sustainability criteria through indirect or auxiliary means, the present model embeds circularity-driven priorities—such as waste prevention capabilities, material reuse strategies, and environmental compliance—into the core decision structure. Moreover, while prior studies frequently assume deterministic settings or rely solely on expert judgment, this study integrates a structured risk-adjusted matrix that systematically accounts for exposure to financial, environmental, and operational risks, thereby increasing analytical robustness and contextual relevance [79]. This combination of theoretically grounded circularity dimensions with a transparent and practitioner-oriented decision framework enhances both academic insight and decision-making effectiveness in supplier evaluation for circular supply networks.

6. Conclusions

The research answered the questions by identifying the relevant criteria for CSS in its development, as well as developing a model based on MCDA, the risk matrix. Additionally, future research could extend the analysis to multi-echelon CSS, and incorporating the ANP method would provide deeper insights into the decision-making capabilities of the proposed model.

The research offers certain limitations. The models’ outcomes are based on evaluation provided by experts in a specific CSC scenario, potentially restricting the broader consideration of the outputs. Although the model can be applied to different industries, the results will naturally differ due to sector-specific characteristics. Nevertheless, future research could explore the application of fuzzy AHP or other hybrid MCDA models to better capture uncertainty and vagueness in expert judgments, especially in more complex or dynamic CSC environments. Furthermore, the use of statistical validation tools could be considered to empirically test the robustness and generalizability of the proposed framework across different datasets and contexts. This integration would enhance the robustness of the model, which is committed to waste reduction and sustainable production practices, supporting the development of more sustainable and efficient CSC across various industries.