Circular Economy of Plastic: Revisiting Material Requirements Planning Practices for Managing Uncertain Supply

Abstract

Regulation and customer awareness pressurize manufacturers to use recycled plastic (RP) in the production system to reduce the negative environmental impact of plastic waste for sustainable production. Compared to virgin materials, the RP available in the market shows high variation in quality, composition, and properties, and often experiences higher variability in lead time. This renders the supply chain of RP and the production systems more vulnerable, making it difficult for material requirement planning (MRP) to decide the optimal quantity and reorder time. This paper first examines the RP supply chain and the sources of variations therein, identifies the associated uncertainties for operations management, reviews the current MRP design elements in managing supply uncertainty, and finally aligns strategies and design elements with the dimensions of the uncertainties. A set of valuable propositions is drawn for the plastic firms to manage variation from upstream suppliers and promote a high-value chain of plastic circularity. MRP practices at the operational level including safety stock, optimization techniques, and alternative bills of material are proposed to mitigate the variations in the supply chain. The work provides a conceptual foundation for the circular economy of plastic, which opens a new paradigm of future research in managing RP in the production system for sustainability.

1. Introduction

Circular economy (CE) strategies are imperative for sustainable development in the production sector [1], and they are significantly important for plastic processing industries. CE is defined as “an economic system that replaces the ‘end-of-life’ concept with reducing, alternatively reusing, recycling and recovering materials in production/distribution and consumption processes” [2]. Plastics are the most versatile and one of the most widely used materials in the modern economy, yet at the same time pose a significant environmental challenge related to their fossil origin and the pollution associated with their use and disposal [3]. Thus, there is a surge of calls for a new plastics economy at a socioeconomic and political level that reduces the negative environmental impacts of plastic through different principles of the circular economy—reduce, reuse, and recycle. For example, the European Union (EU) has recently proposed an amendment to the packaging regulation, calling for mandatory increases in recycling and reductions in plastic packaging [4]. Similarly, the United Nations Environment Programme’s (UNEP’s) “New Plastic Economy Global Commitment” has gathered signatories from both governmental institutions and the private sector, with the aim to reduce overall plastic use and increase the circularity of plastic [5]. Besides packaging, there is also a push towards a greater circularity of other plastic components, for example in the automotive industry, where the EU is working on an amended end-of-life directive to ensure the better circularity of all used materials, including plastics [6]. Plastic waste has potential for recycling in terms of sustainable development [7], and is considered one of the central pillars of a circular plastics economy that can contribute by extending the lifespan [8] through either mechanical or chemical recycling [9].

Yet, despite these pushes for a strengthened recycled plastics supply chain, the overall use of recycled materials remains low. Within the EU, which has one of the most effectively developed systems for the collection of recyclable materials, only 34% of plastic packaging was recycled into new plastic products in 2020 [10], with the rest being thermally recovered in incineration plants, or going to landfills. Research has identified several barriers associated with the low uptake of recycled plastics in manufacturing. Specifically, the use of recycled plastics creates new challenges across the supply chain [11], which complicate the material management of the production system [12] and operationalization at the level of the manufacturing firm [13], and act as a barrier to its adoption [14]. Accordingly, the operationalization of circular practices in manufacturing faces numerous new challenges and uncertainties that affect and challenge day-to-day decision-making [1].

One of the areas is material planning, which is concerned with decisions to ensure that sufficient stock levels are maintained to allow efficient production to meet demand. For these operations, manufacturers often rely on the practice of material requirements planning (MRP), a computer-based production planning method that facilitates the calculation of requirements in terms of materials and timing by converting three inputs—bill of material, inventory data, and master production schedule—into two main outputs, namely, planned order releases and rescheduled notices [15]. Since their inception, MRP systems have been further developed to deal with numerous operational challenges and decision uncertainties, such as uncertain demand [16], the increased complexity of flexible manufacturing systems [17], and supply volatility [12]. However, given the fundamental changes associated with recycled plastics, current MRP tools might place a limitation on effective and robust decision-making in a circular plastics economy. The purpose of this paper is therefore to systematically evaluate the challenges for MRP arising from circular plastics in manufacturing, and suggest new concepts related to how to design the MRP, its inputs, and its output, so as to support robust material management decisions for cleaner production in a circular supply chain.

Specifically, this paper aims to identify the limitations of current MRP designs related to dealing with uncertain supply and its impact on production operations. Therefore, we analyze the recycled plastics supply chain to identify sources of variations in the supply and associated uncertainties for the production system. This research conceptualizes the MRP practices with the supply-oriented uncertainties in the recycled plastic (RP) supply chain. The extensions are drawn by developing concrete propositions regarding how to evolve MRP to provide robust decision support and achieve practical implications in firms working with recycled plastics supply streams. The paper is structured as follows: Section 2 represents the practices of MRP in an uncertain supply environment. Section 3 depicts the research methodology, which is followed by Section 4 discussing the challenges ad current developments seeking to enable MRP in the context of supply chain variabilities. The developed design requirements of the MRP for the recycled plastic supply chain are represented in Section 5. Finally, Section 6 provides a concrete proposition based on the findings, regarding how these new challenges and requirements could be best integrated with the practices and strategies of MRP for a circular plastics economy, and how to unleash the potential for future research avenues.

2. Material Requirements Planning

Material Requirements Planning (MRP) is a tool used to plan material ordering and develop production schedules. MRP is a common computational technique for the resourcing and procurement of raw materials, which allows manufacturers to plan and optimize their procurement by supporting decision-making in, amongst others, reordering and capacity planning. MRP has been an important productivity tool used in creating a competitive advantage in the global economy, and plays a vital role in inventory management in the manufacturing of complex industrial products [18]. Today, MRP is an integral and main part of most enterprise resource planning (ERP) systems. It is thus embedded in the ERP database (supply, demand, capacity, programmed reception, etc.), and is executed periodically (daily or weekly) depending on the nature of the problem.

2.1. Uncertain Supply for MRP

Ho [19] has categorized the uncertainties affecting MRP as environmental and system uncertainty. Environmental uncertainty relates to the uncertainty of supply and demand, whereas system uncertainty relates to production process uncertainty, i.e., production quantity, production lead time, quality issues, changes to production, etc. Both types of uncertainty pose a challenge to decisions supported by MRP. They can, if improperly considered, lead to a high cost of production due to more frequent rescheduling, changing orders, service-level issues, unplanned machine setups, lost sales and goodwill, and high inventory costs [20]. Moreover, uncertain demand and uncertain supply lead time may create “nervousness” in the MRP [12], whereas small input fluctuations have large—and likely costly—effects on production. MRP uses various approaches to deal with these uncertainties and reduce nervousness, such as safety stock, safety lead time, and improved forecasting [21]. While much attention has been paid to work on the strategies of MRP and operations strategies to mitigate uncertainties in demand in the extant literature, very limited research offers solutions to supply-oriented uncertainties [22]. In recent years, there has been literature available related to the demand-driven MRP [16,23] and nervousness [24], and it has been shown that the effect of the uncertain supply causes a ripple effect impacting the production system [25,26]. However, none of the extant literature has focused on operationalizing how MRP approaches in the production system cope with the uncertainties from upstream supply to strengthen the circular economy of plastic.

2.2. MRP Strategies to Deal with Uncertain Supply

Supply chain management systems and practices have, with time, become much better at dealing with the challenges originating from uncertainty and variability through a range of different approaches. Wijngaard and Wortmann [21] compared three MRP approaches, i.e., safety stock, safety time, and hedging, by considering stochastic uncertainties in multi-stage production systems with converging and diverging nodes. Murthy and Ma [20] suggested an MRP model, where output quality is considered a random variable as a function of input quantity and uncertainty in operations. They used overplanning approaches with lot sizing in the master production schedule to cope with the quality uncertainty. Dellaert and Jeunet [27] studied the impact of the positive lead time on the multi-lot-sizing rule, and found that it can be used instead of safety stock. Angkiriwang et al. [28] worked on flexible supply chain management using buffering and redesigning coordination strategies to manage uncertainty. Thuannadee et al. [29] utilized various lot-sizing rules, i.e., a lot for lot (LOL), period order quantity (POQ), part balancing, silver meal, least unit cost, and the Wagner Whitin method of MRP to optimize the inventory of the plastic bag processing factor to achieve a minimum cost of production. Gozali and Ali [30] considered plastic jar companies, seeking to optimize the cost by formulating safety stocks, MRP, and lot size for the forward-loop supply chain.

Companies are more oriented towards flexibility from the demand side, and thus they do not have enough ability to absorb complexity from the supply side (mass customization). Flexible settings in the production system are also required to deal with this uncertain supply in the context of MRP. Bertrand and Rutten [31] proposed flexibility in the recipe of the product to deal with the uncertainty in the material by using an experimental design with the application of a computer. Lyon et al. [32] used material resources planning to develop an alternative bill of material (BOM) for the semiconductor industry, where they used alternative production processes and devices for manufacturing the final product. Lin et al. [33] proposed a critical MRP for the liquid crystal display (LCD) industry using nonlinear programming by use of a network graph, with a statement that “the traditional MRP is no longer enough to deal with the planning problem”. The research also recommended an alternative BOM for customized products according to the requirements of the specific customers. Wang et al. [34] developed a framework and design with the flexibility of an optimized operation for a supply chain network in the face of uncertainty.

2.3. MRP Optimization Techniques for Uncertain Supply

To convert the MRP into a decision-making tool in the production system, researchers used various mathematical approaches.

Table 1. Research work using various mathematical approaches in the MRP environment to deal with different uncertainties (supply-oriented).

Author	Uncertain Environment	Approach
[35]	Uncertainty index: safety lead time and safety stock calculations	Fuzzy logic controllers and artificial neural networks
[36]	Uncertain demand, capacity, and costs	Fuzzy approach
[37]	Epistemic uncertainty and integrity conditions of the objective functions: cost, time, and backorder minimization	Fuzzy goal programming
[38]	Procurement lead time: total cost minimization	Stochastic lead time: chance-constrained programming
[39]	Uncertain lead time: fuzzy costs	Robust fuzzy suppliers’ lead time
[40]	Uncertain lead time	Fuzzy multi-objective
[41]	Uncertain order size and lead time	Stochastic variables
[42]	High lead time, high computational time	Mixed Integer Linear Programming-based approach with solver
[43]	Uncertain capacity, lead time, and inventory	Fuzzy systems
[44]	Uncertain production system: optimal lot sizing and production plans	Fuzzy credibility-based double-sided chance-constrained programming
[45]	Supply and demand uncertainty: safety stock and safety lead time	Stochastic lead time and dynamic demand

shows modeling and optimization approaches in the MRP environment considering different supply and production parameters with uncertain supply. These approaches are used for modeling depending on the nature of the problem and the required data. The fuzzy-based approach is the most commonly available in the literature, which converts uncertain parameters into fuzzy numbers with a membership function. The stochastic approach considers uncertain parameters as random variables with known probability. The stochastic approach is used when the uncertain parameters are converted into random variables following distribution. The final approach is a robust optimization, which considers several scenarios wherein a specific value is given to the uncertain variable, to represent the expected occurrences of the outcome. Table 1 shows modeling and optimization approaches in the MRP environment considering different supply and production parameters with uncertain supply.

The MRP system is used to manage the right material at the right time, in the right quantity, and its adaptation is crucial for recycled plastic so as to integrate the recycled materials within the production process, improve the efficiency, and optimize the inventory. MRP can also integrate the data from suppliers (lead time, batch quantity, etc.) and production processes (capacity, process design, design, etc.) to provide insights into smooth flow, as well as reducing waste and enhancing customer services. It is the MRP that helps businesses to align production schedules with the availability of recycled material, minimizing disruption and efficiently increasing the utilization of RP [46]. Modern MRPs have become better equipped to deal with decision contexts of high uncertainty and volatility, wherein optimization tasks are non-trivial. The approach of enterprise resource planning (ERP) originated from the development of standard inventory packages, which later evolved into MRP II (manufacturing resource planning) and MRP. By the year 2017, the ERP had transformed into a cloud-based system for data. The final phase of the evolution includes the integration of robot process automation (RPA) and artificial intelligence (AI). Sustainability has been incorporated in the ERP system, deriving the S-ERP systems that can monitor sustainable operations, and companies have already incorporated modules into the material management system [47]. Numerous researchers have applied traditional and modern MRP tools in different case industries, e.g., Chandraju et al. [48] executed SAP (ERP-Module) in the sugar industry for material management; therefore, the product can be indented and received within a safety period. Santin et al. [49] applied MRP planning in the furniture industry, and improvements were seen in inventory cost reduction, greater effectiveness in the production and manufacturing processes, and better understanding in terms of information accuracy. Xiong, et al. [50] considered a global chemical company (producing oil additives such as engine oil additives, marine lubricants, chemical and components, and viscosity modifiers) as a case study to analyze the smooth flow of material integrating ERP (SAP) and a manufacturing execution system (MES) using latest ICTs and intelligent technologies i.e., Barcode, RFID, IoT, GPS, Cloud Computing, Big data, and Parallel Control Management. Transitioning to RP requires an improved MRP with new strategies to cope with supply-oriented uncertainties that account for these volatilities, and to manage supply streams of both recycled and virgin materials.

3. Research Methodology

To develop a conceptual foundation of the circular economy by understanding MRP’s role in the plastic circular supply chain, we followed a process of theoretical extension. This allowed us to establish if and how current MRP practices can accommodate the new challenges facing effective and robust material planning associated with the use of recycled plastics. Specifically, we organized the research into three distinct steps.

First, we identified the potential effects of the use of recycled plastics on material planning by reviewing changes resulting from the switch from virgin to recycled materials in the areas of sourcing, inventory management, production planning and control, as well as interfacing with manufacturing technology. For this analysis, we relied on a review of the literature, supplemented by insights from both structured and informal collaborations in which the authors have been engaged with manufacturers using plastics. This review of changes introduced when using recycled plastics has allowed for a comprehensive mapping of factors that can create additional complexities or uncertainties in material planning decision-making.

Second, we reviewed which of the MRP elements are potentially affected by these changes, and how. For this analysis, we mapped the identified changes onto the different MRP elements [51]. Specifically, we attended how changes derived from working with recycled plastics might affect the structure, reliability, and availability of information that serves as an input to MRP, the requirements for mathematical modeling within the MRP process, and the use of MRP outputs within operations decision-making. In addition, we ask, what are the best MRP practices to manage these supply-oriented uncertainties?

Third, we reviewed the MRP literature regarding which current methods and techniques are suitable to accommodate the challenges to material planning when working with recycled plastics. This allowed us to identify current gaps or issues that would need addressing in order to design MRP tools to support material planning decisions in a recycled plastics supply chain. We furthermore formulated propositions and outlined future research avenues based on these identified gaps.

4. Challenges for Material Planning in a Recycled Plastics Supply Chain

There are two approaches to managing and dealing with the uncertain supply of recycled plastic when adopting a circular economy. One is to improve and strengthen the supply chain infrastructure, recycling process, material properties, and developing supply market, and the second one is to improve the production system and make it flexible to manage the supply variations in RP. The scope of this project is to deal with the uncertain supply in the production system and operations, by improving and redesigning MRP strategies, applying alternative bills of material, and employing optimization techniques. Plastic waste is the most heterogeneous material stream due to various factors. First, it exists in multiple forms with distinct properties, which must be separated from each other before the recycling process. The major and most common types are polyethylene terephthalate (PET), polypropylene (PP), polyethylene (PE), and polystyrene (PS), together accounting for more than 63% of the plastic demand in Europe [52]. In addition, plastic products often require different additives during processing to enhance their properties for specific applications. These additives can improve the plastic’s strength, durability, flexibility, heat resistance, or other characteristics. Common types of additives used in plastics include plasticizers, stabilizers, colorants, fillers, flame retardants, heat stabilizers, slip agents, residuals, antioxidants, antistatic agents, UV absorbers, modifiers, and lubricants [53]. The flame retardants are the most common additives used in electronics products; similarly plasticizers, antioxidants, and anti-UV are common additives used in plastic packaging [54]. Moreover, plastic types and additives used in food packaging require strict legislation regulations due to their chemical composition and the potential migration of problematic substances [55]. The proposed research considers all types of recycled polymers and additives used in the processing of plastic products.

The use of recycled plastics instead of virgin materials from fossil or renewable sources is a central element of a new plastic economy. There are two main types of recycling, i.e., chemical recycling and mechanical recycling, as illustrated in Figure 1. Today, the most common form is mechanical recycling, which entails the collection, sorting, cleaning, and compounding of used plastic products into granules that can be reused in the production of new plastic products through molding or extrusion. Besides mechanical recycling, substantial advances in chemical recycling have been made, i.e., the depolymerization of plastics into their chemical components (monomers), which can then be refined and re-polymerized into new plastics [56]. While chemical recycling promises purer recycled polymers with properties closer to virgin materials, the technical processes are currently not available at an industrial scale for all but a few specific types of polymers, such as PMMA. Moreover, chemical recycling requires high-temperature processes, yielding a higher environmental impact of the chemically recycled material [57]. These issues produce differences between the quality and properties of the virgin and recycled plastics at the receiving end, which produce uncertainties for plastic processing firms. For example, PET contains more Sb than PE, PP, and PS types, and a higher recycling rate may produce higher metal concentrations in RP [58]. In addition, recycled PET often exhibits increased plasticity and processability due to changes in molecular weight and the ratio of crystallinity [59]. Although able to be recycled multiple times, the yield and quality of the material in RP inevitably decline, which is significantly affected by impurities in the input feedstock [60]. RP pellets tend to have lower durability, and products made from them are prone to surface scratches when exposed to strong impacts during use. The mechanical properties of recycled polymers were generally lower than those of virgin polymers, and decreased crystallinity led to more uniform surfaces but potentially lower mechanical strength [61].

The availability and use of recycled plastics in manufacturing are further complicated through different configurations of supply chains. In summary, the transition from a linear economy to a circular economy affects the process of MRP due to uncertainties in upstream suppliers, which create problems in the entire production system. A generic view of the uncertainties with sources and their impacts on the plastic value chain in a circular environment is given in Figure 2. The figure emphasizes how material is supposed to flow in a circular loop, and how it could negatively impact the supply, production, quality, and customer satisfaction. It is clear that there are two main supply chain configurations—open loop and closed loop, as illustrated in green blocks. Open loop refers to the sourcing of plastic waste from numerous, often unknown sources, typically post-consumer waste from municipal waste collection schemes. The sources are typically characterized by a mix of different polymer types, high degrees of contamination with non-plastic content (metal, food waste, etc.), and limited insight into how the material has been used. As such, recycled material from open loop systems—also referred to as post-consumer resin (PCR)—often shows high variation in quality and availability, besides other challenges such as smells or (metal) pollutants that clog up the processing machines. Conversely, closed-loop systems build on known sources of plastic waste, typically post-industrial waste like scraps or quality rejects, but also post-use material recuperated through take-back systems for specific products. Thus, post-industrial resin (PIR) materials behave more predictably, but are in their supply naturally limited, as they are tied to concrete processes and products [63].

Overall, the main challenges related to the supply side of the recycled plastics cycle the material quality, availability, and price fluctuations [13,14]. To map the effects of these challenges on decision-making within materials planning, we reviewed the potential effect on the following domains of operations management: supply chain, production, and inventory management.

4.1. Supply Chain Management

The market for recycled plastics is characterized by a substantially different supplier landscape than the market for virgin plastics. While the latter is dominated by a few large, globally active polymer producers, such as BASF, SABIC, or Borealis, the market for recycled plastics, in particular mechanically recycled plastics, is much more fragmented and localized, with new types of suppliers entering the market as recyclers or compounders specialized in recycled materials. Additionally, in closed-loop systems, new forms of suppliers have emerged following the re-use of industrial waste across organizational boundaries, sometimes forming relationships such as industrial symbioses [64,65]. Yet, while the number of suppliers increases, their typically smaller size and more limited capacities increase the risk of supply defaults or the unavailability of supply from a particular source. This is exacerbated by an overall rising concern regarding the future availability of recycled material or feedstock for recycling processes, which creates a risk of a transition towards a seller’s market, where manufacturers need to buy what they can get when they can get it, rather than optimizing for the right restocking point. Consequentially, and seemingly paradoxically, prices for recycled plastics are—not only because of the more complicated processes—often more expensive than their virgin counterpart, and they follow more complex price drivers than virgin plastics, which largely follow the oil price [66]. Thus, manufacturers working with recycled plastics are already diversifying their supplier portfolio or the insource steps of the recycling process, such as compounding, to maintain better control over the supply.

In consequence, these variabilities in the supply chain, related to market, price, availability, and material quality, create new challenges to the production system in the management of material flow and planning to ensure efficient and undisturbed production.

4.2. Production

Virgin and recycled materials, and even different batches of recycled materials, differ in important properties, such as melting behavior, brittleness, or appearance [67]. These differences have implications for the production processes, and need to be controlled. In current production, recycled plastics are therefore often compounded together with virgin materials, whereas the percentage depends on certain criteria related to the output materials or product. This share is typically determined by both technical considerations regarding what proportion would still allow one to maintain the desired product properties, and by the availability of the recycled material on the market [68]. The latter factor means that many manufacturers working with recycled plastics will also flexibly adjust the proportion of recycled material in their products, depending on its availability and price.

Moreover, the typical variations in material properties and qualities between different batches of recycled material have implications for the production process, with two potential outcomes. Moreover, the typical variations in material properties and qualities between different batches of recycled material have implications for the production process; for example, process settings during molding or extruding, or adjustments to the additives required to reach the desired color, flexibility, or melting behavior [53]. Additionally, varying levels of regulated contaminants such as flame retardants may cap the maximum recycled content in the final product. Thus, manufacturers need to flexibly adapt the recipes, i.e., the proportions of plastics and additives, as well as the process parameters (temperature, flow, etc.).

In consequence, two products of the same kind may have different material compositions, which has inspired pragmatic approaches to traceability like a mass balancing approach [69]. Moreover, differences in the production parameters, as well as the commonly reported higher maintenance requirements when working with recycled plastics [13], can affect the overall equipment efficiency (OEE), lowering the throughput and thereby the volume of material being processed within a defined period.

4.3. Inventory Management

Inventory management strikes a trade-off between holding cost and service level provided to the customer. Given the different properties between batches of recycled material, some manufacturers have started to treat these separately, both physically through separate storage systems and in their ERP systems as materials with different inventory numbers. This increased number of stock items means that production planning—and consequentially material planning—needs to accommodate more complexity, given both the higher number of items and differences in how they can be used in production to meet demands [70]. Moreover, following the discussed change in the supplier landscape and the changes to the form of supplier relations, the novel risks to assuring necessary supply can also impact decisions regarding what constitutes an optimal buffer stock. Exacerbated by higher price volatility, manufacturers may decide to build up stock when prices are low—or when material is available—with consequential higher inventory fluctuations and possibly increased costs of holding.

The overall effect of the transition from the linear economy to the circular economy of plastic used for RP on supply chain, production, and inventory management, as discussed, is highlighted below in

Table 2. Major changes to operations management domains following a transition from a linear to a circular plastics supply chain.

Operations	Area of Change	Virgin Plastics	Recycled Plastic
Supply Chain Management	Quality (material properties)	Variations within a narrow range between batches	Can show high variations in material properties between and within batches
	Price	Following oil price (simple) Large-scale production processes with high-cost efficiency	Following several market drivers (complex) Often novel and complex production technologies and small-scale production
	Number of suppliers	Globally consolidated—few big suppliers	Many small, local suppliers and a few large globally acting suppliers
	Type of supplier partnerships	Transactional or strategic partnerships	Anything from transactional partnerships to industrial symbiosis
Production	Recipe—material composition	Stable composition of polymers and additives	Flexible composition of recycled and virgin polymers and additives to reach desired properties
	Recipe—process parameters	Stable, optimized process parameters (temperature, flow…)	Flexible adjustment of parameters to accommodate material properties
	OEE	Relatively predictable	More variable and higher risk of disruptions due to contaminations
	Production quality	Relatively predictable	More variability in quality, risk of higher quality reject rate
Inventory Management	Physical storage	Different batches can be stored in the same storage	Different batches might need separation
	Information systems	Different batches can be treated as the same stock items	Different batches might require different identifiers in the system
	Buffer stock levels	Predictable supply streams allow for optimized buffer stock levels	Variable prices and material availability can influence buffer stock levels

.

5. Design Requirements for an MRP Suitable for a Recycled Plastics Supply Chain

Building on our analysis of the changed decision context for material planning in the circular plastics supply chain, we will now evaluate these changes and challenges in terms of their effect on MRP components, specifically MRP inputs, processing, and outputs. Thereafter, we will develop suggestions on which currently available MRP designs are suitable to accommodate this changed decision context, and highlight the need for further MRP development to support the operationalization of a circular plastics economy.

5.1. MRP Input

MRP applications require input in the form of four types of data: bills of material (BOM), inventory data (on-hand and scheduled receipts), supplier lead times, and master production schedule. Of these, the bill of material is most fundamentally affected by the challenges and changed manufacturing practices associated with recycled plastics, alongside the limited solutions to be found in the current literature.

Table 3. MRP input data analysis for virgin and recycled plastic.

Input Data	Description/Information	Virgin Plastic	Recycled Plastic
Bill of Material (Recipe)	Relationship between product, intermediate components, and raw material.	Fixed relationship between materials and end-product	Complex relationships because of alternative possible recipes/BOM to produce the end product
Inventory Data	On-hand Inventory Type of item Stock level (quantity) Location	Lower number of different items, stable allocation to different storage locations possible	A higher number of different items with changing storage locations (separate treatment of different batches). Higher risk of data inaccuracies
	Scheduled receipts How much is ordered? When will it be received and how much?	High predictability/low variance for scheduled receipts due to the type of suppliers	Higher risk/higher variance due to the type of suppliers
Lead times	Expected time between purchase order and receipt of order	Relatively stable and short.	Risk of longer and more volatile lead times between different orders.
Master Production Schedule	Customer demand due dates. Capacity input BOM information	A few given factors: due dates are fixed; capacity constraints are satisfied	More uncertain factors: changing due dates; incapable of processing varieties

provides an overview of the type of information covered by the different input data types, and the changes to these inputs following the contextual changes of a circular plastics supply chain, which is further explicated in the remainder of this section.

5.1.1. Bill of Material

The bill of material (BOM) is an input to the MRP systems, which includes information on raw materials, or semi-finished products, as well as other engineering design and manufacturing data. In process technologies, such as plastic extruding or molding, the term “recipe” is typically used similarly, referring to the material composition and the process parameters for production. As such, it provides the information necessary to calculate the raw material needed to produce a specified number of products of a certain type. The alternative BOM gives all information to the MRP to produce a reliable plan for the RP supply chain. Typically, a single end product will have a single BOM following an optimized product and production design. However, when using recycled plastics, a single product might have several different recipes—built on different compositions of recycled and virgin polymers, as well as additives. Each variety is mixed with different additives and under different conditions to produce the same plastic product (appearance, durability, quality, etc.). Many different additives can be introduced during the production phase to control the properties of the plastic and enable it to fulfil the requirements for use in specific applications. These include additives such as colorants, fillers, plasticizers, lubricants, and antioxidants [53]. Specifically, the BOM used as input must represent the actual flexibility of using different material variants, a challenge not thoroughly covered in the literature.

To fully capture the complex production reality of working with alternative recipes, we thus require a novel way to structure the BOM input data. A simple case is illustrated in Figure 3, which presents a comparison between BOMs for products made from virgin plastics and products made from a variable content of recycled material. The illustration is important in helping to understand the difference between traditional linear product configuration and how complex the product configuration of the recycled plastic is. The applicable BOMs might even be further complicated in practice, through the possibility of adjusting the recycled content gradually, instead of stepwise. Specifically, to optimize material planning when working with alternative recipes, the MRP needs to solve a complex optimization problem considering demand, the available recycled material (both on stock and possible to receive in due time), and the impact on production capacities when following alternative recipes with different process parameters.

5.1.2. Inventory Data

Inventory data include information on the on-hand inventory of raw materials and work-in-progress, as well as data on the expected receipt of material and semi-finished products already ordered. Given the often separately handled batches of recycled material, the inventory data structure is expected to be more complex [70], with more item numbers and storage locations. While this increased complexity does not fundamentally change the structure of the input MRP, it raises the risk of inaccuracies in the data due to the increased demands for controlling stock levels. Moreover, the discussed changes to the supplier landscape can increase the risk of inaccuracies related to scheduled receipts, as defaults and stock shortages can become more likely in a fragmented supply network. Thus, the data need to capture the uncertainty inherent to inventory data, and the MRP needs to be designed so that it can consider and process these uncertainties adequately when solving the various optimization problems.

5.1.3. Lead Time

Following similar environmental uncertainties, there is also an increased risk that the lead time deviates from the expected value, and future orders face longer lead times [71]. However, here, it is essential to understand the level of uncertainty the supply of recycled material is experiencing because of the unavailability of material and suppliers in the market. The lead time of recycled plastic shows a great deal of uncertainty. It requires a robust solution of optimal safety stock and optimal lead time for the undisrupted flow of materials with a minimum inventory cost.

5.1.4. Master Production Schedule

The master production schedule (MPS) balances expected demand with available capacity. Due to the inherent uncertainty of demand, studies in the literature have paid increased attention to Demand-Driven MRP (DDMRP), which has led to a fundamental shift in MRP logic from static safety stocks to dynamic buffers. While the overall demand is not expected to be fundamentally more uncertain for products based on recycled materials compared to virgin materials, the other pillar of the MPS, capacity, is more likely to be affected. Specifically, the use of recycled materials can create more uncertainty in production set-up time [72], create changes to process parameters affecting throughput, and increase the risk of machine breakdowns due to contaminants like metal scraps [13]. All of these can reduce the overall equipment efficiency (OEE) or render it more variable. Moreover, production planning methods themselves are becoming more dynamic to cope with the challenges of uncertainty [73], leading to potentially more complex input data files, or even the need for full and dynamic integration between MPS and the MRP application.

5.2. MRP Processing

MRP processing concerns the mathematical operations carried out within the MRP application to determine material requirements and optimize material order scheduling, through the steps of BOM explosion, gross/net requirements calculation, planned order release, and order scheduling [74]. The RP supply chain is a complex problem for MRP because of the high level of uncertainty of the input parameters and the dynamic interdependencies of these parameters, such as safety stock, lot sizing, lead time, and alternative BOM information. The expected changes to the input data—particularly BOMs with alternative compositions, but also more adaptive production scheduling—complicate the requirement calculations. Probabilistic input data for lead time, which is used to determine the optimal re-ordering points, further increases the complexity. The threefold challenges—more input data, probabilistic input data, and dynamic input data—mean that the optimization problems to be solved within MRP are non-trivial, and require suitable approaches.

5.3. MRP Output

The MRP program generates output data as a part of production planning and managing plant operations. The outputs include (1) planned order releases, which provide the authority to place orders that have been planned by the MRP; (2) reports of planned order releases in future periods; (3) exception notices. MRP provides the information for crucial decisions for ordering required material at the right time to be available for the operation department. It thereby helps to avoid stock-out and consequential ad-hoc changes to the production schedule, which can destabilize the production system and necessitate rapid changes to setups and resource allocation. In the case of the RP supply chain, the conditions are uncertain, and there are more changes expected in the material requirement planning due to variations in the quantity, quality, time, and composition. Too frequently, material planning processes with uncertain lead time, availability, and quantity of the recycled content will result in high costs of production, unplanned machine setups, high inventory costs, shortages, and customer loss.

6. Discussion and Propositions

The study aims to provide a research agenda for companies to adopt circular operational practices for sustainable production and circular economy development. The discussion is based on the findings of how the RP supply chain is different from other linear supply chains, along with intended variations, their effects on operation decisions, and MRP approaches and techniques, to derive several propositions. Through a systematic process of theoretical extension—identifying the sources of variation in the circularity of RP for exploring associated uncertainties in the production system—the paper maps the MRP and operations practices against the uncertainties of a recycled plastics supply chain. There are numerous routes and levers to control the issue of variance, i.e., developing better systems to support planning under new conditions and working with the requirements, making sure to utilize the space of opportunity that may have over specification/safety margins, developing interactions between material and production processes, whereby both can be made more apt to variance, and finally collaboration with the upstream RP suppliers to provide the right material and/or appropriate engagement with the customers to understand the accepted level of variation. The following propositions provide an opportunity for researchers, academicians, and experts to find new avenues to reach effective and efficient solutions to support the circular economy for plastics:

• The quality variation in the supply of the RP affects the production process and results in waste, maintenance, and the reduced quality of the output product. This variation cannot be handled at the material management level, and therefore production settings must be sufficiently flexible to deal with different compositions of plastic material with different recipes, also called alternative BOM, as explained in Section 5.1. Here, it is important to understand the properties of the material and the process parameters. The alternative BOM is designed to represent all possible alternative recipes of the end product from the combination of the raw materials. Many different additives can be introduced during the production phase to control the properties of the plastic and enable it to fulfil the requirements for use in specific applications. These include additives such as colorants, fillers, plasticizers, lubricants, and antioxidants [53]. The alternative BOM gives all information to the MRP that enables it to produce a reliable plan for the RP supply chain. This flexibility can be part of automation or a technological improvement, and will also provide agility to the production system to improve its resilience. It is also important to discuss the level of flexibility required to deal with the uncertain quality variation of the lots received from the upstream suppliers. Process flexibility and optimal process parameters are the limitations imposed on an alternative BOM, and the potential avenues of research related to managing the uncertain variety in quality are huge.
  Proposition01. An alternative BOM for process customization provides a viable solution to managing uncertain quality in the recycled plastic supply chain. However, it requires extensive knowledge of the process parameters and additives based on the composition of the RP. In addition, alternative BOMs also require flexibility in machine settings to control and manage the process of producing quality output products;
• Lead time is the main input into MRP, which is directly under the control of the management. The managers must set this lead time for each component, and need to understand how the performance has been affected by changing the lead time [75]. It becomes uncertain due to the immature supply market of RP, and as a result, the RP becomes short and will not be available on time. Lead time variability is therefore a key risk driver for the smooth flow of material through production, and MRP aims to avoid this issue and ensure on-time deliveries. A mitigation strategy is required to reduce its negative impact on the production system, and in the MRP context, there are various solutions and approaches given in Section 5.2. However, safety stock and safety lead time are the two major managerial approaches followed by MRP to reduce unwanted fluctuations in the lead time. The supply market of RP is underdeveloped for various reasons, e.g., competition against incineration, many small and new suppliers, lack of investment, and lack of technology. Therefore, big firms and the government should collaborate to develop legislation and subsidize RP processing to enable the development of the supply market. The other reason for the uncertainty in the lead time is the unavailability of the data of the RP at the required time. The uncontrolled nature of the lead time data in the case of RP requires the aid of emerging technologies, i.e., RFID, IOTs [76], and a digital product passport [77], to directly illuminate the uncertain changes in the supply chain and to estimate the lead time. Once one has collected data on time, then the manager can make effective decisions to mitigate the uncontrolled negative impact on the production system due to RP supply.
  Proposition 02. An uncertain lead time in the RP supply chain is the key input parameter of MRP, and mitigation strategies are required that use safety stock and safety lead time in the production, as well as supplying market maturity and using emerging technologies;
• The supply-oriented variation due to uncertain quantity is the biggest issue in MRP, and it is essential to select the best modeling technique to decide the right quantity order considering uncertain parameters. Researchers used several modeling techniques, i.e., fuzzy, stochastic, and robust approaches, to model uncertainties depending on the complexity of the problem and the company’s objectives. However, these improvements provide optimal plans in controllable uncertain environments. The nature of the RP, covering many input variables, constraints, and parameters under an uncertain environment, makes the MRP problem more complex. Therefore, it is required to look deeply into the applications of advanced techniques and analyze the impacts on managing RP supply uncertainties by performing an empirical study. This will help the organization find the best ordering schedule and production plan under uncertain supply variations for RP.
  Proposition 03. Recycled plastics supply chains operate at a higher level of uncertainty as compared to supply for virgin materials. As a consequence, MRP problems with uncertain input parameters become more complex, enabling them to formulate and calculate optimal re-ordering scheduling. The development of advanced modeling techniques based on stochastic, fuzzy, and robust approaches to meta-heuristic algorithms for optimization will enable MRP to deal with the uncertain supply of RP. However, the evaluation requires empirical studies to find the best-suited modeling approach for the RP problem;
• The firm must be externally aligned with stakeholders in the context of a circular economy [78] to manage the variations. Suppliers must be aware of the process capability and flexibility of the production firms providing the required recycled material. The processing firms are required to align the quality of raw material with the capability of the production system for the resilience of plastic supply chain management [79]. Prosman and Wæhrens [65] draw a relation of significance of integration between suppliers and manufacturers to manage variations in the quality of the received raw material in cases of industrial symbiosis, but they considered cement as a product for which different processing conditions are required. However, they derived better results in managing supply variation when integrating suppliers with the requirements and capability of the production. The production firms must integrate suppliers in their production planning process to ensure the quality of the supply with the capability and flexibility of the process to deal with the quality variations of RP. In addition, this also necessitates customer engagement to understand specific demand specifications at a granular level, and therefore to understand the important constraints and points where leeway for variation can be found and accepted. The organizations must help the customers understand the changes in the product due to the transition from a linear to a circular supply chain of RP.
  Proposition04. Collaboration with suppliers and engaging customers are significant measures in the process of reducing the supply variation in the recycled plastic supply chain.

7. Conclusions

Despite strong calls for a circular plastics economy, the operationalization of circular plastic supply chains still comes with many challenges regarding quality, available volumes, or the predictability of availability. These challenges have a negative impact on material management, production, and output quality in production. This research therefore explored how operations management decisions for a recycled plastics supply chain can be supported. We did so by first discussing how the circularity of recycled plastic creates changes in supply chain, production, and inventory management. We then outlined the resulting demands for the material requirement planning process to deal with the new challenges effectively. Specifically, this study shows how current MRP practices are often rigid and incapable of managing supply-oriented variations, as faced in a recycled plastics value chain. This paper therefore aimed to outline alternative approaches for MRP to increase the resilience against the supply uncertainty of recycled plastic, and points towards potential research avenues, formulated as concrete propositions.

As we have discussed, many drivers of recycled plastic supply variations can create vulnerabilities for production and supply systems designed following the optimization logic of production systems for virgin materials. This may lead to MRP approaches that are too rigid, and are incapable of managing the more heterogeneous and variable supply of recycled plastics. Alternative MRP approaches need to integrate the higher uncertainty of the supply side through more robust modeling approaches, and consider the increased flexibility of production technologies using alternative BOMs and dynamic integration with production planning systems. Thus, the study also contributes to identifying production and supply chain decisions connected to the MRP in managing inventory, maintenance, waste, and cost for a resilient supply chain, without stockouts, late deliveries, poor quality, and high prices.

Overall, MRP decisions are an integral part of production planning and play a crucial role in creating resilience when facing uncertainty in recycled plastics supply. The suggested alternative approaches for MRP in a recycled plastics supply chain also benefit other important production operations, such as supply chain, production, and inventory management. The suggested alternative approaches are geared towards creating a robust strategy to deal with uncertain lead time or material unavailability. Moreover, they consider flexibility in process (machine settings) and inventory management to provide agility in the production system. These solutions are not empirically verified and are based on concepts, and therefore invite future empirical work, representing a step towards establishing a resilient production system by identifying uncertainties and shortcomings in current systems, and suggesting new practices to manage variations resulting from the use of recycled plastics. This study has therefore extended the current academic discussion of MRP by attending to the specific challenges of working with the circular supply chain of recycled plastics for the circular economy and sustainable development.