Introduction

The United Nations (UN) estimates rapid growth of the world’s population, reaching 8.6 billion by 2030 and 9.8 billion by 2050 [1]. Today, we live in a consumption-oriented society where we produce and consume a large volume of products giving insufficient regard to what happens to the products before, during and after use or consumption. The FAO reported in 2011 that roughly one-third of food produced for human consumption, which is about 1.3 billion tons per year, is wasted globally [2]. The OECD [3] estimates an increase in the global middle class from 1.8 billion people in 2009 to 4.9 billion in 2030, consequently increasing the demand for resources [4]. The Ellen MacArthur Foundation [5] calls it “a potential consumption time bomb”. Many researchers thus call for discarding our traditional linear model of resource consumption, following a take, make and dispose pattern [6].

While the linear model of resource consumption is becoming less dominant, the concept of circular economy (CE) has increased in popularity and importance [7]. CE aims at an industrial economy that is restorative by intention and design and therewith also closing the material loops and replacing the end-of-life concept [4, 6]. Jawahir and Bradley [8] state that the CE is based on reducing wasteful resources by designing and implementing products and processes in a way to improve resource efficiency.

Currently, only 8.6% of the world economy is circular, as reported in the Circularity Gap Report of 2020 [9]. As a result, many initiatives concerning CE were introduced, such as circular business models, promotion of responsible consumption and innovative technologies that enable circularity. Such initiatives are best undertaken at the level of supply chains instead of individual businesses. The Ellen MacArthur Foundation [6] identifies supply chains as the key unit of action for driving change. In order to study and evaluate the processes in supply chains, reference models are used. The well-known and widely used reference model is the Supply Chain Operations Reference (SCOR) model [10,11,12]. SCOR categorizes SC supply chain processes into groups, namely, plan, source, make, deliver, return and enable, and provides detailed best practices and performance metrics [13, 14].

However, there is limited research on the application of SCOR to circular supply chain management. A few years back, Ntabe et al. [12] conducted a Systematic Literature Review (SLR) of papers published between 2000 and 2012 on the application of SCOR with special attention to environmental issues. They concluded that little attention has been given to environment-related and closed-loop supply chain approaches but observed that there is a positive trend in the number of papers about GreenSCOR following its release in 2008. Vegter et al. [15] did an SLR study on “circular economy” and “supply chains” through which they defined the concepts of supply chain processes and performance objectives of a supply chain in a circular business model. To our best knowledge, there has not been any comprehensive review of the literature that gives an overview of the application of SCOR to circular supply chain management. This research aims to fill this gap and give insights into how the SCOR model has been applied to circular supply chain management and which issues or challenges have occurred. To address this gap, the following research questions were defined:

  • RQ1)  To what extent has the SCOR model been applied in circular supply chain management?

  • RQ2) Which related approaches have been used for circular supply chain management and how do they differ from SCOR?

  • RQ3) What are the open issues and challenges in using SCOR for circular supply chain management?

The remainder of the paper is organized as follows. In the “Background” section, we present the background on circular economy and the SCOR model. The “Methodology” section describes the SLR methodology we adopted. The “Results” section presents the results of the SLR. In the “Discussion and Conclusion” section, we discuss the results and finally make concluding remarks.

Background

Circular Economy

There are many different definitions of CE which led Kirchherr et al. [16] to analyse the diverse definitions across 114 studies and conceptualize CE as an economic system which rejects the concept of product “end-of-life”. Various approaches are used to reduce the volume of products used or overcome that products are wasted. These approaches include the concepts of reduction, reuse, recycle and recover. These could occur during production, distribution and consumption processes at the micro (products, companies, consumers), meso (industrial parks) and macro (city, region, nation and global) levels in order to achieve sustainable and equitable development [16].

There are three popular conceptualizations of CE: biological and technical cycles, the “R frameworks” and sustainability. The Ellen MacArthur Foundation conceptualized the biological and technical material cycles graphically which is widely used in popular literature [4,5,6]. The foundation conceptualizes biological materials as those that are not hazardous and can therefore return to the soil through, for instance, composting or anaerobic digestion. Technical materials cannot be easily returned to the earth and should be designed to be recovered, refurbished and upgraded to close the loop.

The 6R framework is one of the “R frameworks” that started with the three Rs: reuse, reduce and recycle [7, 8]. However, Bradley et al. [17] noted that the traditional 3R framework is short-sighted, as it mostly follows a cradle-to-grave approach in which resources will eventually be disposed of in a so-called grave such as landfills or incinerators [18]. The 6R framework, which we used in this paper, aims at the cradle-to-cradle approach, where materials maintain their status as resources [4, 5]. The 6R framework adds the recover [8, 19, 20], remanufacture [8, 21] and redesign [17, 21] principles to the existing 3R principles.

Sustainability is a closely related concept, though CE is considered by its proponents as a new approach [22]. According to Geissdoerfer et al. [23], sustainability is known for its three pillars called people, profit and planet, referring to the balanced integration of social, economic and environmental performance. CE is known for the idea of closed loops and an industrial economy that is restorative by intention and design. CE has clear aims and directions, while the goals of sustainability are open and are generally applied in diverse contexts.

The SCOR Model

The SCOR model is used to manage supply chains effectively and ensure adequate collaboration between the different entities of supply chains. The SCOR model was created in 1996 and is now a product of APICS, after the Supply Chain Council merged with APICS in 2014 [13]. The model consists of standard process descriptions, key performance measures, descriptions of best practices and related software functionalities [10, 13, 24].

The SCOR model is a process reference model that aligns supply chain processes with key business functions and goals. The model contains 5 major elements, namely, performance, processes, practices, people and special applications. It is probably most known for its detailed descriptions of supply chain processes and performance indicators, and it is those aspects of SCOR that are considered in this study.

In the SCOR model, supply chain processes are described at three aggregation levels. At the highest aggregation level, level 1, six main processes are defined in SCOR version 12.0, namely, plan, source, make, deliver, return and enable [13]. Source, make and deliver concern the procurement, manufacturing and delivering of goods or services. The make process includes all types of material conversions or creation of content for services, meaning that activities like testing or simple processes like (re-) packaging are also included. The return process regards the reverse flow of goods, excluding remake activities. The plan process is about the planning of all the above processes. At last, the enable process refers to the overall management activities of supply chains.

The SCOR model includes performance measures, which are used to assess the outcomes of the execution of supply chain processes. APICS [13] describes three elements of performance measures, namely, performance attributes, metrics and process/practice maturity. The performance attributes are strategic characteristics that are used to prioritize and align supply chain’s performance with business strategy. SCOR performance attributes are reliability, responsiveness, agility, costs and asset management efficiency. SCOR includes SustainableSCOR, which includes multiple strategic environmental metrics to evaluate environmental performance. It is currently based on the Global Reporting Initiative (GRI) standard [25], ensuring a more complete set of metrics and alignment with widely accepted standards. In the older version of SCOR (prior to version 12), SustainableSCOR was referred to as GreenSCOR.

The Intersection Between the SCOR Model and CE Frameworks

The above background description clearly shows that CE prescribes processes such as reuse, recycle and recover that are either missing or insufficiently covered in the SCOR model. In CE, these circular processes are at least as significant as the processes included in the SCOR model, such as source, make and deliver. Though the return process in SCOR is directly related to some of the processes of CE, such as reuse, not all CE processes and the associated performance metrics are included in the SCOR model. The absence of essential circular processes and metrics in the SCOR model will undermine the model as a management tool for businesses that adopt circular supply chain practices, such as recycling and refurbishing. In addition, supply chain managers who wish to adopt CE practices within their businesses must now work with two different sets of frameworks that are not necessarily compatible with each other. Therefore, in a more direct manner, the intersection between the SCOR model and the CE frameworks across the various economic sectors needs to be identified and analysed, which we attempt to address in this study.

Methodology

A Systematic Literature Review (SLR) is a methodology that aims at a fair evaluation of a research topic or question, by evaluating and interpreting all relevant and available research [26]. This SLR study is based on the steps outlined by Kitchenham and Charters [26], which consists of the three main phases of planning the review, conducting the review and reporting the review. We describe below how the three phases are applied in our research.

In the planning phase, we defined the research questions, formulated the search strings, selected the literature databases and defined search and quality criteria. The main research questions are already presented in the “Introduction” section. Based on the scrutiny of the background, we further identified the following sub-questions to the first research question, RQ1:

  1. a.

    In which sectors is the SCOR model applied in circular supply chain management?

  2. b.

    Which SCOR levels have been used in circular supply chain management?

  3. c.

    Which processes of the SCOR model are used in circular supply chain management?

  4. d.

    Which performance attributes and metrics are used in circular supply chain management?

  5. e.

    Are GreenSCOR and SustainableSCOR applied to circular supply chain management and if so, how?

Likewise, we identified the areas that need attention in relation to RQ2 and RQ3, namely, sustainability metrics, and the lacking or complexity of sustainability measures and processes. The sub-questions and the specific areas of attention are addressed in dedicated sub-headings in the “Results” section.

We selected Scopus, ScienceDirect, Web of Science, Wiley, Springer Link and IEEE Xplore literature databases. We defined the search keywords based on the themes identified in the “Background” section, namely, SCOR, GreenSCOR, circular economy and sustainability. After performing the analysis of the literature found in an initial test search, including literature found through snowballing, we added the terms closed-loop and recycling into the search string. The resulting search string that we used to search in the title, abstract and keywords fields of the literature is given below:

“(SCOR OR “supply chain operation* reference” OR GreenSCOR) AND (circular* OR green OR sustainab* OR “closed-loop” OR recycl*)”

The search strings were adjusted for each database in order to fulfil the requirements of the interface the database offers. A total of 525 articles were identified via the search string, and one additional article was found through snowballing (see Table 1).

Table 1 Search results per literature database
Full size table

To select the relevant papers, exclusion criteria (Table 2) and quality criteria (Table 3) were defined and applied.

Table 2 Study exclusion criteria
Full size table
Table 3 Study quality criteria
Full size table

Papers after and including 2008, the year that GreenSCOR was introduced, were selected. The application of the exclusion criteria excluded 495 papers, leaving 31 papers for quality assessment. The quality assessment was done using a three-level scoring scheme: 1 (high quality), 0.5 (medium quality) and 0 (poor quality). Two papers scoring below 4.5 (out of the possible maximum score of 9) were excluded leading to a total of 29 primary studies. Figure 1 summarizes the quality scores of the papers.

Fig. 1
figure 1

Quality assessment scores

Full size image

The 29 papers were analysed to answer the research questions. The analysis includes the steps of data extraction and data synthesis. In order to extract data, we read all the primary studies thoroughly. A data extraction spreadsheet form was prepared in which every research question and sub-question had its own data column where the data extracted from the paper were recorded. Additional columns for each research question and sub-question were added as necessary when papers were read, and new types of data were discovered.

During data synthesis, we used the ISIC (International Standard Industrial Classification) and SCOR process levels as a means of classification and categorization schemes. The ISIC standard is used around the world as a classification scheme of national economic activities for the purpose of economic analysis [27]. The 21 different activities are reproduced in Table 4. The activity manufacturing of food products falls under manufacturing according to the classification of ISIC, yet it is added separately in this study, as it is interesting for our analyses.

Table 4 The economic activities of ISIC with number four taken separately (based on UN (2008)
Full size table

In addition, performance metrics and attributes were used leading to 183 columns for each primary study. All metrics were checked to see if they are related to SCOR version 12.0, and those that are not directly related to SCOR were mapped to closely related categories. We used the 6R processes to evaluate to what extent the SCOR model supports CE processes. We also used performance attributes of SCOR and mapped them to the closest GreenSCOR or SustainableSCOR attribute. When the performance attributes used in the papers were not directly related to the SCOR model performance attributes, we presented the approach the papers used to derive new performance attributes.

Results

In this section, the results of the analysis of the research questions are presented. An overview of the 29 selected papers is shown in Table 5.

Table 5 Selected papers for the SLR
Full size table

The publication date varies between 2008 and 2020, of which a little over half of the papers were published between 2017 and 2020. More than half of the papers were found in Scopus; the rest were found in Web of Science, IEEE and Science Direct databases in comparable proportions. Surprisingly, none of the primary studies were found in Wiley and Springer Link databases.

RQ1(a): Sectors

Table 6 depicts an overview of the number of papers related to the 7 ISIC economic activities mentioned in the papers. The specific sectors or industries mentioned in these papers are provided in the last column. When the papers do not mention any specific economic activity (which is the case for review papers), the economic activity is labelled as N/A (not applicable). When papers mentioned the SCOR activities plan, source, make, deliver, return or enable, we did not immediately interpret them to be related to, for instance, manufacturing or transportation and storage. Only the specifically explicitly mentioned economic activities were included during data extraction.

Table 6 Number of related papers per economic activity
Full size table

Most papers are related to either one economic activity or did not specify any economic activity; two papers specifically mentioned two economic activities, which are Effendi et al. [38], concerning sugar production, and Marimin and Safriyana [40], regarding the palm oil supply chain.

Manufacturing

Though the papers focussed on diverse industries, when classified according to ISIC, manufacturing (including the manufacturing of food products) is the most dominant economic activity that was the focus of most of the papers. The papers focussed mainly on the SCOR green purchasing performance evaluations [28, 29, 32, 41,42,43, 47, 50]. Four papers related to the economic activity of agriculture, which were interesting because they focussed both on existing and new models for performance analysis of agriculture or sustainable supply chain [30, 34, 40, 45]. Additionally, three papers were about manufacturing of food products which included performance evaluations in relation to new food supply chains [37, 38, 40].

There were relatively fewer papers related to economic activities other than manufacturing and agri-food. Two papers were about water supply, sewerage, waste management and remediation activities [36, 46]; one of them applied SCOR and Activity Based Costing (ABC) for Integrated Biomass Logistics Centres (IBLCs) of a waste recycling supply chain. Two other papers used SCOR to make an evaluation system for green construction, and one paper used SCOR in relation to supply chain performance [44, 52]. Two papers were concerned about transportation and storage, including performance evaluation and closed-loop logistics [35, 38]. At last, there was one paper related to activities of extraterritorial organizations, as it was concerned with the mapping and evaluation of eco-industrial parks (EIP) through the SCOR model [54].

Nine papers did not relate to any economic activity at all. This was because some papers did a review [12, 15, 31], built a new framework or model that is not specific for an economic activity [39, 48, 49, 51, 53], or did a comparison between GreenSCOR and ISO 14001 [33].

RQ1(b): SCOR Levels

In Fig. 2, the complete overview of the number of papers per SCOR level is shown. Level 3 processes were only rarely and marginally mentioned that we did not consider them for further analysis.

Fig. 2
figure 2

Number of papers related to the separate SCOR levels

Full size image

The three SCOR levels have been explicitly identified when performing the analysis. However, if a paper mentioned a SCOR process from a lower level (for instance, level 2) then that paper is automatically considered being related to the corresponding higher level (for instance, level 1) process. This is because the SCOR model processes are hierarchical, and in the high-quality primary studies we selected, no paper includes a lower level SCOR process without referring to the corresponding higher-level process. It can be seen obviously that there are far more studies focussing on higher level processes than lower levels. Five out of the 29 papers were not related to SCOR levels at all. Most of these papers used metrics from SCOR, yet they did not mention any SCOR process in the paper [32, 38, 44, 52]. One paper Ntabe et al. [12] reviewed previous papers but did not use SCOR levels and processes. It should be pointed out that the papers used the SCOR processes in different ways, as some papers apply it to a certain economic activity, whereas others use them to develop a new model or approach.

RQ1(c): SCOR Processes

In Fig. 3, an overview can be seen of the total number of papers that were related to the level 1 processes plan, source, make, deliver, return and enable.

Fig. 3
figure 3

Total number of papers related to the different level 1 processes

Full size image

The SCOR sourcing process has been investigated more often than the other SCOR level 1 processes, and there were studies which were only, or mostly, about the sourcing process [39, 42, 43]. Though critical for CE, the return process has been applied less than the other processes except, probably, the enable process. However, since enable level 1 processes was not defined in SCOR 10 or earlier versions but were instead embedded as level 2 processes in the other five level 1 processes, we cannot be certain if the return process has been applied more often that the enable process.

It is interesting to see the cross relations of the processes used in the papers. Figure 4 provides an overview of the number of papers that used the different level 1 process combinations. In this figure, the processes are referred to as P (plan), S (source), M (make), D (deliver), R (return) and E (enable).

Fig. 4
figure 4

The level 1 process combinations used in the papers

Full size image

The combination of plan, source, make, deliver and return has by far been mentioned most. Only three papers mentioned all processes, including enable [15, 28, 34]. One paper mentions the processes plan, source and enable [42], which is interesting because enable is generally used in combination with all other processes. Waaly et al. [42] focused on procurement, and so all level 2 processes of plan, source and enable were also focused on sourcing in their paper, namely source planning, source make-to-order and enable source. There were also two papers that exclusively focused on the sourcing process [39, 43]. The combination of plan, source, make and deliver, the basic processes that have been there before 2008, were mentioned in only three papers [37, 45, 50] and the combination of source, make and deliver by two papers [46, 54]. At last, one paper, Bin et al. [53], used the return process in combination with source, make and deliver. It can be seen that the return process has always been used in combination with most other processes and always at least with source, make and deliver.

We also kept track of the SCOR level 2 processes, which were found in 10 of the 29 papers. An overview of the total number of papers related to each level 2 process is shown in Fig. 5. For each level 1 process given at the bottom of the chart, the corresponding level 2 processes are shown above it as bubbles. The sizes of the bubbles are proportional to the number of papers (given next to the bubbles) related to the corresponding level 2 process. The numbers at the bottom represent the total number of papers for each level 1 process. For instance, 9 out of the 29 papers were related to the plan level 1 process, and each one of them is related to the sP2 level 2 process (plan source), but only 6 of them are related to the sP1 level 2 process (plan supply chain).

Fig. 5
figure 5

Number of level 2 process-related papers categorized by the corresponding level 1 processes

Full size image

There are more level 2 processes for the first four level 1 processes compared to those for return and enable processes. Moreover, there is only one paper related to the enable and four to the return level 2 processes, meaning about every seventh paper investigated the level 2 return processes, whereas about every third paper investigated the level 2 plan and source processes. Also note that the bubbles of sS3, sM3 and sD3 show that the engineer-to-order processes were less frequently mentioned than make-to-order and make-to-stock processes.

RQ1(d): Performance Attributes and Metrics

There were 14 papers that considered performance metrics (or attributes) in circular supply chain management, which is a bit less than half of the papers. Ten papers did not use performance metrics; in four papers, it was unclear whether the papers applied one or not; and in one paper, it was unclear which metrics were used. Besides the metrics from the SCOR model, multiple non-SCOR performance metrics were mentioned as shown in Fig. 6.

Fig. 6
figure 6

Number of related papers for each attribute mentioned in the papers

Full size image

The four most frequently mentioned performance attributes of the top five attributes in Fig. 6 are SCOR performance attributes. The attribute agility, which is part of SCOR, is mentioned less frequently than the attribute flexibility, which is not part of SCOR. Nevertheless, the terms agility and flexibility are similar, and the metrics of agility included flexibility in earlier versions of SCOR, which can explain the lower numbers related to agility. Three papers were related to sustainability, and most of the attributes they used are similar to the SCOR attributes.

The metrics found in the papers were checked whether they were related to the latest version, version 12, of SCOR. Out of the 165 metrics found in the different papers, 48 of them were unrelated to SCOR (see Table 7). The table shows that all metrics were mentioned only once except for metric “% of supplier with an EMS or ISO 14000 certification”, which was mentioned twice. Most of the additional metrics are about the environment, even when they were given as finance-related. These metrics are concerning the costs or investment for environment and were mostly found in Bai et al. [39], Liu et al. [44] and Kiriş et al. [32]. There were also multiple societal-related metrics, which are lacking in SCOR according to Büyüksaatçi Kiriş et al. [32], Kyllönen and Helo [37] and Stohler et al. [48].

Table 7 Overview of metrics not related to SCOR
Full size table

The metrics that could be related to SCOR were categorized under the five attributes, and a separate category was used for GreenSCOR. The categories were also used for metrics that were not directly related to SCOR. In which case, the closest category was used as shown in Table 8. Compared to the other categories, agility and asset management had fewer papers and metrics related to them. Moreover, it can be seen that source cycle time is a category that has been mentioned the most. This can be related to the large number of source process-related papers. Also, the cost to source category has a large number of associated metrics, while the make cycle time and deliver cycle time have fewer related metrics. The cost to deliver category has no associated metrics, and the cost to make category has almost the same number of metrics as the cost to source category. There are only limited metrics related to the return process, as there are no metrics related to return cycle time and the metric cost to return has only been mentioned once. However, there are many metrics related to GreenSCOR, which also has the second-highest number of associated papers.

Table 8 Categorized metrics found in the papers related to SCOR
Full size table

RQ1(e): GreenSCOR/SustainableSCOR

The selected papers mentioned only GreenSCOR and not SustainableSCOR, even though SustainableSCOR was introduced in 2017 and includes more specific metrics. Out of the 29 papers, there were only 8 papers related to GreenSCOR, and 18 papers did not mention the application of GreenSCOR. For two papers, this research question was not applicable, and for the paper of Kiriş and Börekçi [32], it was unclear whether GreenSCOR was applied or not.

GreenSCOR was used in different ways, but mostly for evaluation purposes. Pulansari and Putri [29] used KPIs partially similar to GreenSCOR, Susanty and Hidayatika [28] used best practices as indicators, and Effendi et al. [38] formulated green objectives. Three papers used so-called green processes, based on GSCM and GreenSCOR [29, 41, 47]. Furthermore, metrics of GreenSCOR were checked for alignment with the Triple Bottom Line [48], compared with ISO 14001 [33], and used to extend a proposed performance measurement framework [30]. For the last purpose, the specific metrics used were unclear. This shows that though some papers used metrics from GreenSCOR, the specific metrics used were not always clear or completely similar.

RQ2—Approaches Based on SCOR

The papers discussed many different approaches, which either meant proposing a new model based on SCOR or using SCOR in combination with another approaches. In total 27 different types of approaches were found in 20 papers, and some papers combined multiple approaches as shown in Fig. 7.

Fig. 7
figure 7

Number of different approaches found in the papers based on SCOR

Full size image

Many papers used normalization, weighting or other methods to determine relationships among indicators. Within this category, Analytical Hierarchy Process (AHP) and Snorm de Boer (SdB) were used in five and four different papers, respectively (counted only once in Fig. 7). The other methods were used in a single paper. Four papers discussed a model including extra processes, besides the ones from the SCOR model itself [35, 45, 51, 53]. To give a better insight into how SCOR is applied within circular supply chain management, the different types of approaches are explained below in more detail.

Approaches with Extra Sustainability Metrics

Four papers introduced extra sustainability metrics, besides the SCOR metrics. Two of these papers proposed a model that included extra sustainability metrics [32, 37]. One paper, Bai et al. [39], used business and environmental measures for supply chain sourcing, of which some were already existing and some were new. At last, Stohler and Rebs [48] tried to match sustainability metrics to SCOR in order to propose more integration of social metrics.

Normalization, Weighting and Determining Relationships Among Indicators

Eleven papers used approaches of normalization and weighting of indicators or determining relationships among them, for instance, to prioritize them or determine their values. These approaches were all used or proposed in combination with the SCOR model. AHP and SdB have been used a lot and most of the time within a single paper. In these papers, the AHP method was used to weigh the importance of the KPIs from SCOR that affect performance, and SdB was used to normalize the performance values to ensure the same scale for each KPI [28, 29, 38, 42, 50]. Other approaches used to determine the weights of indicators are the G1 together with the entropy method by Liu and Xu [44] and the Analytical Network Process (ANP) by Marimin and Safriyana [40]. Also, the Shannon entropy fuzzy TOPSIS method was used to set priorities and rank the KPIs of the supply chain [30].

Marimin and Safriyana [40] used the Hayami method, besides ANP, to measure the added value of the indicators, and Liu et al. [44] used the grey relational method, besides determining weights of indicators, to analyse the correlation among them. Effendi et al. [38] used the DEcision MAking Trial and Evaluation Laboratory (DEMATEL) approach, besides AHP and SdB, to see the interdependent relationships among metrics. Additionally, cause and effect relations among metrics were found by the fuzzy DEMANTEL method to prioritize metrics by Kiriş et al. [32]. At last, Tan et al. [45] used fuzzy TOPSIS as an approach together with their CLASCM model to rank alternatives in decision-making for project performance.

Other Approaches

He et al. [34] created a Product Carbon Footprint (PCF) model for each SCOR process to calculate PCF across the sustainable supply chain. Yadav et al. [30] introduced a Supply Chain Performance Measurement (SCPM) system based on IoT data, to focus on the importance of data collection and communication. Activity Based Costing (ABC) was used together with SCOR to propose an integrated framework to comprehensively evaluate the costs of supply chains of Integrated Biomass Logistics Centres (IBLCs) [46]. Tramarico [50] introduced a training assessment model to assess the effectiveness of Green SCM training, and Le Tellier et al. [54] adapted SCOR to map processes inside an Eco-Industrial Park (EIP). Besides introducing performance metrics, Bai et al. [39] used a different approach, namely, the grey-based neighbourhood rough set methodology. This method was used to monitor and control the number of performance measures used to manage supply chains [39]. Liu et al. [44] used the Balanced Scorecard (BSC) with SCOR to create a performance evaluation indicator system of green construction supply chain, in addition to using the G1 method, entropy method and grey relational method.

RQ3—Open Issues and Challenges

In total, eleven out of the 29 papers mentioned issues or challenges. The five distinct issues and challenges that were found are summed up below:

  1. 1.

    The SCOR model lacks sustainability measures [32, 37, 39, 48].

  2. 2.

    There is an overload of measures which can be costly to manage and are inefficient [39].

  3. 3.

    SCOR cannot capture the complex interactions between different processes and the payoffs between them [49].

  4. 4.

    Certain existing processes of SCOR should shift their focus as they do not support a circular business model [15].

  5. 5.

    Certain processes are missing in the SCOR model [15, 35, 45, 51, 53].

Lacking Sustainability Measures

Four papers mentioned that sustainability measures were lacking in SCOR. Kiriş et al. [32] state that GreenSCOR is driven by economic and environmental metrics and that it is, therefore, insufficient to achieve real sustainable supply chain management approach. They note that environmental metrics are not fully developed, and social metrics are neglected. Stohler et al. [48] argued that social metrics are not easily compatible with an economically oriented supply chain process model, which focuses more on products than people. The older papers from 2012 mentioned that environmental and social metrics were lacking [37, 39]. The environmental metrics added by Kyllönen and Helo [37] could be related to SustainableSCOR, but not those from Bai et al. [39] (see the “RQ1(d): Performance Attributes and Metrics” section).

Overload of Measures

Bai et al. [39] introduced a methodology to help managers in determining which performance measures to utilize to achieve effective management of overall business and ecological outcomes. The authors argued that there is an overload of measures which can be costly and inefficient to manage, especially when just a few measures provide information on the overall outcome.

Complex Interactions Between Processes Not Captured

Badurdeen et al. [49] states that the SCOR model version 9.0 does not take into account issues of sustainability over the life of the product. Moreover, the authors stated that the model has a focus on specific processes and not able to capture the complex interactions between different processes and the payoffs among the actors, both within and between supply chains. They note that because of its focus on specific processes, it is unable to track the return on assets due to recycling throughout the life cycle of a product.

Existing Processes Should Shift Focus

Vegter et al. [15] discuss that some of the already existing processes and new processes that were identified should shift their focus to support a circular business model. According to them, the plan process should, for instance, consider the environmental boundaries, the source process should focus more on alternative materials to be sourced with low environmental impact, and the suggested to exclude maintenance and recovery from the make process and put them in a separate process as these have a more prominent place in circular business models.

Missing Processes

There were five papers that mentioned processes that were missing in the SCOR model. The missing processes identified by the papers are shown in Table 9; the numbers between brackets are the number of papers that identified the given missing process.

Table 9 Overview of the missing processes of SCOR found in papers
Full size table

First of all, there are four papers, published before 2013, that proposed an extra process named either make return or reverse make [35, 45, 51, 53]. The new process was introduced as an addition to the SCOR processes of source, make, deliver, source return and deliver return. This process was introduced by the papers because the authors missed a process for repairing, separating, waste recycling and other comparable activities [51, 53], or disassembly, repair and recovery [45]. Besides, Tan et al. [45] renamed all return processes into reverse processes in their model as the authors noted that reverse concerns also the reclaiming and re-manufacturing processes instead of only returns of unqualified and surplus goods.

Additionally, Xiao et al. [35] proposed an additional process called logistics information to handle the forward and reverse flow of logistics operation information in a closed-loop logistics operation. The authors suggested the corresponding new level 2 processes that are missing in SCOR. These processes are given in Table 9 as the last seven processes under the “level Level 2 processes” column. All the level 2 make return processes added in the table are about some aspects of recycle product processes.

Vegter et al. [15] argued that there is no framework available that sufficiently addresses processes of a supply chain in a circular business model. The authors introduced additional processes at level 1 called use and recover and a number of processes at level 2 (see Table 9). The process use was suggested for addition to the SCOR model, because the authors determined that consumption by the end-users should be part of the supply chain in a circular business model, and because recover is an important component of CE [15].

Discussion and Conclusion

In this paper, we reviewed the literature systematically and reported the extent to which the SCOR model has been applied to circular supply chain management. In this section, we first discuss the implications of the results and then present the limitations of our study.

Interpretations and Implications of the Results

We studied application of the SCOR model in circular supply chain management for major economic sectors. We found that only seven of the 29 ISIC categories were the focus of the literature, and most of those were related to manufacturing. We found only six papers that are related to the agri-food sector. This number is low considering the fact that multiple papers note the importance of circularity to agricultural production and food systems due to the challenges of feeding the growing world population [55,56,57]. It is also interesting to see that the SCOR model was not applied as much in the waste sector, which by its very nature strives to realize a circular economy.

Most papers dealt with SCOR level 1 processes; level 2 got less attention; and level 3 processes were rarely mentioned. The level 1 process return was used sixteen times, which is slightly less than the other processes, and was used relatively less at level 2 in comparison to other level 2 processes. The paper Ntabe et al. [12] found eleven papers published between 2000 and 2012 that used the return process and concluded a positive trend in papers related to environmental issues. We, therefore, expected a higher number of level 2 return processes among the research papers published between 2008 and 2020. However, among the papers selected in our review, which focussed on CE, we found fewer publications on return processes. This suggests that the return processes of SCOR may not have been viewed in the same way as the return principle of CE.

It was interesting that there were multiple attributes and metrics found in the papers that were not related to the SCOR model. Most attributes were similar to SCOR attributes, except those related to sustainability, and which were studied by multiple papers. Out of the 165 metrics, 46 were not related to SCOR. Even though some of the new metrics were vague concepts, such as “satisfactory”, they might be interesting additions after making them clear and purposeful. Whereas many papers focussed on metrics in relation to environmental costs, it seems that SustainableSCOR does not account for those. Nevertheless, as most of these additional metrics were found in only one paper, their value should be considered with care when reflecting on their addition to SCOR. Though return processes were important in CE, there were only a few metrics related to them, which was unexpected.

We identified the application of GreenSCOR and SustainableSCOR to circular supply chain management and found that they were not mentioned as much as expected; namely, only eight papers were related to these aspects. It turns out though that if metrics are mentioned, they were generally about GreenSCOR. However, the papers did not always use the specific metrics from GreenSCOR. It is also interesting that SustainableSCOR has not been mentioned at all, considering that it was introduced in 2017 and includes a number of extra and more specific metrics, useful for evaluating circularity.

We found extra approaches based on SCOR for normalization and weighting of indicators or determining relationships among the indicators, for instance, to determine and prioritize their value. Some papers concluded that SCOR lacks sustainability metrics; others suggested additional social and environmental metrics. Some others missed processes in the SCOR model for representing reference circular production processes. However, APICS [13] claims that the make process in SCOR model caters for common material conversion processes including recycling and remanufacturing.

The results show that there is little synergy between research done on the SCOR model and CE frameworks. The SCOR model clearly lacks processes that are common in CE, such as those found in the recycling industry. The results also indicate that the existing SCOR model needs a fundamental rethinking and aligning with CE frameworks.

Limitations and Future Research

First, it should be taken into account that this study included only 29 papers; thus, the results were also based on the limited number of papers available. We have made our best efforts that relevant papers will not be missed due to the specific selection of the search strings and our interpretation of exclusion criteria. Furthermore, the interpretation of the results and categorization is also done carefully. Yet, some relevant papers and information may have been missed. We have, however, followed the guidelines of Kitchenham and Charters [26] and ensured that our results are reproducible and can form a basis for future research.

There were less papers than expected for agri-food and waste, and also the return process was used less than expected which indicates that there is still more focus on a linear economy in the literature. We suggest future research to focus on these areas giving special attention to the challenges and open issues we summed up. In addition, SustainableSCOR does not seem to account for environmental costs and could be considered in future research on how to incorporate them in the SCOR model. In general, only few papers applied SCOR as-is, which indicates that the application of SCOR in its current form to circular supply chains is challenging, and it may need careful further investigation that guide the future releases of SCOR.