Abstract
Uncertain events such as earthquakes, epidemics, and wars have increased the risk of supply chain disruption. Due to the needs of carbon reduction policies and environmental protection, a large number of enterprises have started to produce both traditional and green products. Studying the issue of supply chain disruption for such enterprises has significant practical significance. We have developed a system dynamics model for a substitutable dual product supply chain with two levels of supply sources. Through simulation analysis, we found that (1) supply chain disruption can cause fluctuations in the manufacturer’s inventory, and disruptions from second tier suppliers have a higher impact on the manufacturer’s inventory than that from primary suppliers. In addition, the disruption of traditional products will cause consumers to flow to the green product market, resulting in a sudden increase in order for green products and components in a short period of time, causing a delayed impact on the inventory of suppliers and manufacturers of green products. (2) The disruption of upstream suppliers in traditional products causes the highest profit losses for all traditional product suppliers, while the disruption of downstream suppliers in green products causes the highest profit losses for the manufacturer and all green product suppliers. (3) From the perspective of the service level, compared to other components, the disruption of critical components in traditional products poses the highest risk of out of stock in the supply chain, while the risk of out-of-stock in the intermediate component of green product is the smallest. (4) Common sense may suggest that the more the suppliers disrupt, the higher the damage of the supply chain. However, due to the ripple effect, this article finds that from the perspectives of profit, inventory, and service level, multisupplier disruption is not necessarily inferior to single supplier disruption.
1. Introduction
In recent years, the global spread of the COVID-19 and the outbreak of the Russia-Ukraine conflict have accelerated the restructuring of the world’s development patterns and interest patterns. The worldwide political, economic, and social crisis has suddenly intensified, and the supply chain of many enterprises has been disrupted. Disruptions in supply chains influence trust by inducing overreactive behaviors across the network, thereby impacting the ability to consistently meet the resulting fluctuating demand [1]. The supply chain disruption has brought large losses to many enterprises. For example, due to the continuous disruption related to the COVID-19 war, natural disasters, and the continuous global chip shortage, the automobile manufacturer Toyota reduced its output [2]. Apart from Toyota, other companies are often affected by disruption. Hiroyasu et al. [3] simulate the economic loss resulting from supply chain disruptions triggered by the Great East Japan Earthquake (GEJE) in 2011, applying data from firm-level supply chains and establishment-level attributes to an agent-based model.
Due to limitations in production capacity or raw materials, automotive manufacturers often adopt a division of labor and collaboration approach to produce cars. For example, Toyota has long collaborated with the automotive semiconductor manufacturer Renesas and small companies with universal capabilities, such as the rubber manufacturer Fujikura, to produce complete products by purchasing their components [4]. So, once Toyota’s upstream suppliers have been disrupted, they will have to stop production. Most previous literature has focused on the disruption situation of their first tier suppliers. However, many first tier suppliers also purchase raw materials from their upstream suppliers. Therefore, many first tier suppliers’ disruptions are caused by second tier suppliers, which have forced enterprises to pay attention to the disruption problem of their second-tier suppliers. In addition, with the promotion of carbon reduction policies and the requirements of sustainable development, the current automobile manufacturers not only produce traditional cars but also new energy vehicles. Therefore, the supply chain of automobile enterprises has gradually formed a substitutable dual product supply chain framework. They need to produce both green and traditional products and need to consider the risk of disruption brought by their suppliers’ superior suppliers, and the problems they face are very complex. As a result, for a substitutable dual product supply chain with second tier supply sources, how to evaluate the risks of different disruption modes and improve the suppliers’ supply capacity is a very worthwhile issue to study. Our research endeavors to address the following key inquiries: RQ1: What disruption mode inflicts more significant damage to profitability? RQ2: Which disruption mode has a greater impact on service levels? RQ3: Does a multisupplier disruption result in greater losses for manufacturers compared to a single supplier disruption?
We have developed a system dynamics (SD) model in this article, which visualizes and compares the disruption risks of different supply sources and products through simulation methods. Drawing from the simulation results, our study reveals that the disruption of downstream suppliers leads to the most substantial profit losses for manufacturers. However, disruptions of upstream suppliers pose the highest risk of out-of-stock incidents in the supply chain. Interestingly, multisupplier disruption is not inherently inferior to single supplier disruption from an intuitive standpoint. In light of these findings, we offer managerial insights to help enterprises enhance their supply capacity and mitigate risks effectively.
The novelty and contributions of this paper are as follows:(i)Existing research on the supply chain of substitutable products predominantly focuses on pricing and competitive decision-making [5–7], with minimal exploration of disruptions in the supply chain of alternative products. Our study aims to fill this gap and contribute to the enrichment of the research field related to substitutable product supply chains.(ii)Current literature on supply chain disruptions primarily examines single disruption modes [8–11], neglecting the complexity of mixed interruptions involving various levels and products. In contrast, our study delves into these intricacies and demonstrates that the impact of multisupplier disruption may be less severe than that of single supplier disruption.
The organization of this paper is as follows. In the second section, we introduce the relevant research. In the third section, we describe the research problems and model framework. In the fourth section, we constructed a formula and developed a stock flow diagram based on Vensim DSS software. In the fifth section, we tested the effectiveness of the SD model. In the sixth section, we analyzed the behavior of the supply chain system under different disruption modes, and in the seventh section, we summarized the conclusions and enlightenment.
2. Literature Review
This section introduces the current academic research on substitutable product supply chain decision-making and the application of SD in supply disruption.
2.1. Substitutable Product Supply Chain Decision-Making
Research related to the supply chain of alternative products mainly focuses on product pricing decisions, green level decisions, etc. Motlagh et al. [12] investigated the impact of substitutable green products on supply chains compared to nongreen products. Animesh et al. [13] developed and compared centralized and decentralized game theory models to analyze competition between green and nongreen products. Giri et al. [5] studied the dynamics of selling two substitutable products and one complementary product in a two-level supply chain, formulating various pricing strategies. Wang [6] optimized pricing and warranty decisions for fuzzy supply chains. Nazari and Seifbarghy [7] examined dual-channel sales under stochastic linear demand. Dong et al. [14] explored pricing decisions in bilateral monopolies. Cheng et al. [15] constructed an evolutionary game model for omni-channel strategies, and Zhang et al. [16] analyzed a dual-channel supply chain system.
In summary, research on the supply chain of substitutable products mostly revolves around pricing and competitive decision-making, and there is little involvement in the disruption of the supply chain of alternative products.
2.2. Application of SD in Supply Chain Disruption
Chopra and Sodhi [17] identified supply chain disruptions caused by natural disasters, human threats, and regulatory changes. Using SD to analyze supply disruption can provide insights into the dynamic effects of such disruptions.
Zhang et al. [18] considered a closed-loop supply chain with supply disruption and employed SD to model the dynamics of a Stackelberg game. Zhang et al. [8] used SD to study the financial system of a three-echelon supply chain during the epidemic. Lai et al. [9] used system dynamics simulation and game theory to study the supply interruption and international trade risks of offshore global production networks, established a system dynamics simulation model considering supply interruption risks, evaluated the impact of different supply interruption modes on OEM profits, and finally proposed dynamic and static penalty mechanisms. Azizsafaei et al. [10] mainly use system dynamics modeling methods to generate system risk management models, analyze the impact of dynamic risks on the behavior of the food supply chain system, and help improve management’s insight into the key role of system dynamics models in analyzing various types of risks and improving their efficiency in the food supply chain process. In order to study the impact of government subsidy strategies on supply chain recovery in the context of supply chain interruption, Ju et al. [11] focused on products with high demand during the epidemic period and used the selection of government subsidies under the influence of production capacity and transportation interruption as the entry point for recovery strategies. Using the cumulative total profit of chain members as the judgment indicator, modeling and simulation are conducted using system dynamics to construct a secondary supply chain for manufacturers and distribution centers, simulating eight scenarios of different production capacity levels and transportation interruptions, and clarifying the impact of government subsidies on supply and chain recovery. Abhijeet et al. [19] defined, applied, and demonstrated the ability of system dynamics modeling to identify and visualize supply, demand, and logistics interruptions, as well as the cascading effects of supply, demand, and logistics mergers and simultaneous interruptions. Wang et al. [20] studied how upstream and downstream enterprises embedded in supply chain networks participate in decision-making under interruption risk. The decision evolution problem of supply chain networks as complex adaptive systems (CAS) was studied using evolutionary game theory (EGT) and cellular automata (CA) methods from two aspects: temporal dynamics and spatial characteristics. Davoud et al. [21] examined resource sharing between suppliers during disruptions using an evolutionary game model and SD simulation.
In addition, when supply chains such as healthcare, agricultural products, and cheese are disrupted, SD methods can still be used for coordination and research. In response to the disruption of the medical supply chain caused by the coronavirus, Singala et al. [22] developed a conceptual SD model for medical supply chains during the coronavirus pandemic. Ana et al. [23] used SD to assess the robustness of a fresh agri-food supply chain. Zhu et al. [24] applied SD to study a cheese supply chain. Lee et al. [25] modeled power supply and demand in the context of earthquake disasters using SD methods and tested the model for the earthquake disaster that occurred in South Korea in 2016. Zhang et al. [26] simulated changes in inventory and orders during disruptions. Lu et al. [27] designed a case of supply disruption, and combined with SD, they built a simulation model of double chain one-way coordination strategy pattern and double chain multinode two-way coordination strategy pattern. Based on the two indicators of “inventory level” and “order accumulation rate,” the dynamic operation mechanism of the supply chain system under different mitigation strategies was derived. Raghuram et al. [28] mainly study the impact of inability to obtain parts from suppliers due to unpredictable events and the cost of recovering from such interruptions and construct system dynamics models to explore the interrelationships between different variables. Gökçe et al. [29] used SD to analyze payment terms and supply risk management practices.
In summary, it can be seen that system dynamics, due to its unique advantages in handling complex and dynamic problems, has been applied by many scholars to the study of supply chain disruption. However, the existing literature on system dynamics has little attention to the disruption problem of substitutable dual product supply chain, and there is also little mention of the disruption problem of multilevel supply sources related to assembly supply chain. For automobile manufacturers, they need to produce both green and traditional products and need to consider the risk of disruption brought by their suppliers’ superior suppliers, so the problems they face are very complex.
3. Problem Description
We analyzed an assembled supply chain consisting of three levels (as shown in Figure 1): the first level, second level, and the third level. The first level includes a single company, known as the “manufacturer,” which assembles manufactured products and sells them to the consumer market. This “manufacturer” represents car manufacturers, such as Audi, Tesla, and Toyota. Manufacturers produce and sell both traditional and green products at the same time, which have competitiveness and substitutability. For example, many car manufacturers produce both traditional fuel powered vehicles and new energy vehicles. Some consumers prefer fuel powered vehicles (because new energy vehicles have troubles such as cumbersome charging and insufficient range), and some consumers prefer new energy vehicles (because traditional cars have lower environmental friendliness and higher fuel consumption costs). Manufacturers assemble cars by purchasing critical components required for production from upstream suppliers, such as the air conditioning system or engine of the car. At the second level, or rather at the first level of the supplier, there are two first tier suppliers (supplier 1 and supplier 2), which produce the critical components required for traditional cars (supplier 1 is responsible for production) and new energy vehicles (supplier 2 is responsible for production) and supply them to the manufacturer. For example, the critical components for Toyota’s car air conditioning are purchased from the supplier. At the third level, or rather at the second level of the supplier, there are two second tier suppliers (supplier 3 and supplier 4), which produce intermediate components of the critical components required for traditional vehicles (supplier 3 is responsible for production) and new energy vehicles (supplier 4 is responsible for production) and supply them to the first level supplier. The second tier supplier may be a semiconductor manufacturer that provides electronic equipment used in air conditioning units.
Overall, we considered a system with two levels of supply sources, and existing research mostly focuses on the disruption of the primary supply source, while the disruption problem of the second tier supply source is rarely mentioned. This system of second tier supply sources is widespread in reality. For example, Audi’s air conditioning suppliers also need to purchase and manufacture air conditioning components from their superior suppliers, and Audi often only signs contracts with its first tier suppliers; Audi does not know who its second tier suppliers (i.e., their superior suppliers) are. However, in reality, many disruptions from first tier suppliers are often caused by second tier suppliers, not the first tier suppliers themselves. Therefore, this requires manufacturers to consider the disruption issue of their first tier suppliers’ superior suppliers.
An assembly supply chain system is with two levels of supply sources, in which the disruption mode is more complex than a single-level supply source system. At the same time, traditional products and green products often have competitiveness and substitutability. The different disruption modes of these two products can also make enterprise decision-making more complex. Overall, in this paper, we attempt to answer the following questions:(1)What are the differences in the impact of disruption from first tier and second tier supply sources on supply chain behavior?(2)What impact will the multiple disruptions of first tier and second tier supply sources have on supply chain behavior? Is multisupplier disruption necessarily more damaging to the supply chain than single supplier disruption?(3)For traditional and green products, what impact will the disruption of one product have on the operation of the other?
4. Model Setting
4.1. Method Introduction
SD is a continuous, time-varying, and visual simulation method, and the SD model is essentially composed of complex differential equations. In practice, due to the scale and nonlinearity of the system, it is difficult to obtain the analytical expression of the complex system, and only its numerical solution can be obtained. Therefore, SD differentiates the differential equation and then uses the computer for simulation analysis. Any decision in actual management cannot be made every moment but rather over a certain period. Therefore, treating differential equations differently is in line with the operational rules of actual management. In this paper, we use Vensim DSS software to visualize the SD model.
The variables in the SD model are divided into state variables, flow rate variables, and auxiliary variables. State variable represents the general solution of the differential equation, while the initial value of the state variable in the simulation represents the particular solution of the differential equation. State variable is the accumulated quantity that changes over time and is the storage link of matter, energy, and information. State variables in a causal chain can change the overall dynamic properties of the system. Flow rate variable represents the differential in the differential equation. The function of the flow rate variable is to transform various factors that affect the system state, information, plans, and decisions from inside and outside the system, into actions that change the system state. The flow rate variable determines the size of the state variable. Auxiliary variables represent the structure of differential equations, and their expressions are similar to those of variables. They are algebraic operations and have no standard form. In SD models, auxiliary variables represent information within the system and can be any combination of constant terms, state variables, rates, or other auxiliary variables.
All variables in this article are shown in Table 1. The constants used in SD simulation are all included in Table 2.
4.2. Mathematical Formulation
4.2.1. Structure of Demand for Final Product
The total demand of consumers for traditional products and green products in the market is random. We model it as a random variable subject to normal distribution (equation (1)). At the manufacturer level, the manufacturer sells the traditional products and green products to consumers at the same time. When the manufacturer’s product inventory is greater than the market demand, the manufacturer can fulfill all market demand. When the manufacturer’s product inventory is less than the market demand, there will be a shortage of products, and the consumer’s demand cannot be fulfilled. Therefore, the sales of the two products are determined by the product inventory and market demand (equation (2)). The total market demand is divided between traditional and green products. For example, many car manufacturers now produce both traditional fuel powered and new energy vehicles. Some consumers prefer fuel powered vehicles (due to the hassle of cumbersome charging and insufficient range), while others prefer new energy vehicles (due to their low environmental friendliness and high fuel consumption cost), the market demand for green products is often related to the level of green technology of the product, and consumers are more willing to purchase low emission cars and pay higher price for them. The demand for traditional products is the difference between the total market demand and the demand for green product (equation (3)).
4.2.2. Structure of Disruption
There are many types of supply chain disruption, such as transportation disruption, capacity disruption, and demand disruption. Different disruption modes have different impacts on the supply chain. In our model, we mainly focus on transportation disruptions, which are disruption modes where products cannot be sent to designated locations due to natural disasters, road damage, road closures, and transportation vehicle failures. Transportation disruption is widespread. For example, due to volcanic eruptions and frequent earthquakes, Toyota’s suppliers in Japan often have a transportation disruption. In reality, many suppliers may face disruption not due to their own reasons but rather due to their superior suppliers. Therefore, in our model, we consider the disruption problem with two levels of supply sources, which expands previous research on assembly supply chains. Traditional products can have two superior suppliers, while green products also have two superior suppliers; therefore, there are four disruptions in the entire supply chain system: (1) disruption of the intermediate components produced by the second tier suppliers of traditional products, (2) disruption of the critical components produced by first tier suppliers of traditional product, (3) disruption of intermediate component produced by second tier supplier of green product, and (4) disruption of critical component produced by first tier supplier of green product. The disruption mode is jointly determined by the start time and duration of the disruption. We define the disruption mode as a piecewise function (equation (4)) and use the PULSE function in Vensim DSS to realize the simulation of disruption (equation (5)).
If the inventory of a product exceeds the market demand, there will be an inventory surplus. If the inventory of the product is less than the market demand, there will be a shortage phenomenon, and the number of outstanding orders is related to the current demand and current inventory (equation (6)). At the same time, the service level of the supply chain is defined as the ratio between the fulfilled order and the demand (equation (7)). The service level reflects the operational efficiency of the supply chain.
4.2.3. Structure of Inventory and Order
We assume that each level of supply chain members adopts a continuous inventory review system, which is a common inventory control strategy in reality. According to the characteristics of a continuous inventory review system, the target inventory of supply chain members at each level is a function of lead time, predicted demand (or predicted order), and safety inventory (equation (8)). Smooth procurement strategies are often used to predict demand or order, which can minimize inventory fluctuations. The expected demand faced by the manufacturer is a prediction of product demand (equation (9)), while the expected orders faced by the first and second tier suppliers are predictions of order from upstream supply chain members (equation (10)). The SMOOTH function in Vensim DSS can achieve this smooth visualization (equation (11)). In addition, the safety stock is determined by the safety stock coefficient, and a higher safety stock coefficient means a higher safety stock (equation (12)).
Each level of the supply chain members determines the upstream order based on the difference between their target inventory and the existing inventory. It should be noted that due to the functional substitutability of traditional products and green products, when the demand for one product cannot be fulfilled, consumers will flow to the market for another product. Therefore, the manufacturer will increase the order after observing this increased demand. The manufacturer’s orders for traditional products and green products are also related to the unfulfilled demand for both products (equation (13)). Inventory is a state variable, with each level of supply chain member’s inventory having an inflow rate variable and an outflow rate variable (equation (14)).
Assembling traditional products and green products requires a certain amount of assembly time, as well as some production time for the critical components and intermediate components in the production of these products. Therefore, the inflow rate of each inventory has a delay. The DALAY function in Vensim DSS can visualize these equations (equation (15)).
The outflow rate variable represents the quantity of products or components sold, and each outflow rate is a function of inventory and demand (or order) (equation (16)).
4.2.4. Structure of Cost, Revenue, and Profit
We model the cost, revenue, and profits of all supply chain members as state variables, which means that the simulation model calculates the cumulative cost, revenue, and profits. The profit of each member is the difference between revenue and cost (equation (17)). The revenue of all supply chain members is equal to their sales revenue, which includes the revenue from first tier suppliers selling intermediate components, second tier suppliers selling critical components, and the manufacturer selling two types of products (equation (18)). The costs of the two suppliers of traditional products include inventory cost and procurement cost for intermediate and critical components, while the costs of the two suppliers of green products include inventory cost, procurement cost, and green technology R&D cost for intermediate and critical components (equation (19)).
Figure 2 shows the complete SD model based on Vensim DSS, which includes three supply chain levels: manufacturer level, first tier supplier level, and second tier supplier level.
5. Model Validation
A model is only an abstraction and approximation of a real system, and whether the constructed model can effectively represent the real system directly determines the quality of model simulation and strategy analysis. Therefore, before conducting simulation experiments, we verify the effectiveness of the model. The verification method of the SD model is different from the general simulation model. SD model mainly focuses on the behavior of the system rather than the quantitative analysis of data. Extreme test and reality test are two commonly used methods to check the rationality of the model structure. We developed two different scenarios with extremely high demand and extremely low demand (0 demand) and benchmark models to compare their behavior in different situations.
5.1. Test in Inventory
Figures 3 and 4 show the behavior of inventory in three scenarios: extremely high demand, extremely low demand, and base case. We capture the inventory of intermediate components, critical components, and two products.
From the simulation results of inventory, it can be seen that the inventory levels of both first tier suppliers are higher than those of their downstream manufacturers, and the inventory levels of both second tier suppliers are also higher than those of their downstream first tier suppliers. Due to the bullwhip effect in the supply chain, the demand information is gradually amplified, so this result is in line with reality. In the context of extremely low demand, the inventory of critical components from all first tier suppliers, intermediate components from second tier suppliers, and the manufacturer’s product inventory are all in a straight line. This is because without demand, it means that there are no sales or orders. Therefore, the state variable “inventory” only has an initial inventory value, and their inflow and outflow rates are both 0. In the benchmark model, the inventory levels of all supply chain members are no longer straight lines and exhibit fluctuations, which is in line with the regular pattern of supply chain inventory changes during normal operation. It should be explained that the inventory level of the supply chain in the baseline scenario is sometimes lower than the extremely low demand scenario: this is because the supply chain in the baseline scenario has sales and order fulfillment behavior, and all members of the supply chain will adjust the inventory according to the demand, and the initial inventory value we set is higher than the demand in the base case. Finally, the inventory level in the supply chain under extremely high demand scenarios is much higher than the other two scenarios, which is also in line with reality, as all suppliers and manufacturers require more inventory to fulfill the extremely high demand.
5.2. Test for Profit
Figure 5 shows the behavior of profit in the three scenarios. From the simulation results of profit, it can be seen that in the scenario of extremely low demand, the profits of all members of the supply chain are negative, which is in line with reality. They have no sales revenue and no order fulfillment, only inventory costs, so they are all at a loss. On the contrary, in the scenario of the basic model, the supply chain achieves positive profit, and especially in the case of extremely high demand, all supply chain members have higher profit.
5.3. Test of Service Level
Figure 6 shows the behavior of service level in the three scenarios. From the simulation results of service level, it can be seen that when the market demand is extremely low, there is no transaction generated, so the service level is 0. When the market demand suddenly becomes extremely high, the existing supply chain inventory cannot fulfill the extremely high demand, resulting in a large number of out-of-stock situations. Therefore, the service level curve in the extremely high demand scenario is much lower than the benchmark scenario in the early simulation time.
In summary, the behavior of various key variables in the model in different contexts is in line with reality, and it can be considered that the model we have constructed is effective.
6. Risk Analysis of Different Disruption Modes
In this section, we use the SD model to analyze the research problem. We set the simulation duration to 200 weeks. Looking back at the third section, our research focuses on the impact of different disruption scenarios on the behavior of the supply chain system. Therefore, we used three indicators to evaluate the behavior of the supply chain: inventory (which includes the inventory of all supply chain members), profit (which includes the profit of all supply chain members), and service level (demand fulfillment rate).
6.1. Setting of Simulation Scenarios
We have set up 15 simulation scenarios that represent different disruption modes, and their specific forms are shown in Table 3.
Table 3 shows the design of our simulation scenarios. Supplier 1 and supplier 3 provide critical and intermediate components for traditional products, while supplier 2 and supplier 4 provide critical and intermediate components for green products. We have set up 15 different disruption modes, which cover all possible scenarios. Among them, the symbol “-” represents that there has been no disruption. When a disruption occurs, the start time of the disruption is set to 50th week and the duration of the disruption is set to 10 weeks for each supplier. For example, S1 represents that supplier 1 experienced 10 weeks disruption from the 50th week, while S5 represents that all suppliers experienced 10 weeks disruption from the 50th week. S1, S2, S3, and S4 all belong to situations where only one supplier experiences disruption, S5, S6, S7, S8, S9, and S10 all belong to situations where two suppliers experience disruption simultaneously, S11, S12, S13, and S14 all belong to situations where three suppliers experience disruption simultaneously, and S15 represents that all suppliers have experienced disruption.
6.2. Only One Supplier Is Disrupted
In this section, we mainly analyze the situation where there is only one supplier experiencing disruption. The suppliers experiencing disruption may be a primary supplier producing critical components or a second tier supplier producing intermediate components. Existing research about assembly supply chain mostly focuses on the disruption of primary suppliers, while the disruption problem of second tier suppliers is rarely mentioned. In reality, the disruptions of primary suppliers are usually caused by the disruption of their upstream second tier suppliers. Therefore, it is necessary to compare the performance of the supply chain system with that of primary suppliers when the intermediate components of second tier suppliers are disrupted. We use S1, S2, S3, and S4 to capture these characteristics. S1 and S2 represent disruption from primary suppliers of traditional and green products, while S3 and S4 represent disruptions from second tier suppliers of traditional and green products. The simulation results of inventory, profit, and service levels are as follows.
Figure 7 shows the inventory of the manufacturer’s traditional and green products under different simulation scenarios. From the perspective of the manufacturer’s inventory, when there is a disruption of traditional products in the 50th week, the inventory of the manufacturer’s traditional products drops sharply. However, compared to S3, the inventory curve of manufacturer’s traditional product in S1 is significantly lower than that in S3; that is to say, the manufacturer’s traditional product inventory is more sensitive to disruption from supplier 3 which provides critical components, and disruption from second tier suppliers will have a higher impact on the manufacturer’s traditional product inventory. In the S2 and S4 scenarios, after the disruption in the 50th week, the inventory of green products has also sharply decreased, the manufacturer’s green product inventory curve in the S2 scenario is lower than that in the S4 scenario, which means that the manufacturer’s green product inventory is more sensitive to the disruption of supplier 4 which provides a critical component. Therefore, it can be concluded that whether it is a traditional or green product, the disruption of second tier suppliers has the highest impact on the manufacturer’s inventory.
It is worth noting that when one of the traditional or green products is disrupted, the inventory of the other product will experience a delayed increase. For example, in S2 and S4, in which suppliers of green products were disrupted, the manufacturer’s traditional product inventory curve did not show a downward trend but instead experienced a high fluctuation around the 70th week. This is because the disruption in the green product caused some consumers to switch to traditional products due to the substitutability between green products and traditional products. The manufacturer observed an increase in demand for traditional products in the current period, which increased the order volume of traditional products in the next period. Therefore, there is a lag oscillation in the curve. For the green products, the manufacturer’s green product inventory experienced a decrease after disruption in the S2 and S4 scenarios. However, the manufacturer’s green product inventory curve in the S2 scenario is lower than that in the S4 scenario, which means that the manufacturer’s green product inventory is more sensitive to the disruption of supplier 4, which provides critical components. The disruption of the second tier suppliers will have a higher impact on the manufacturer’s green product inventory.
Figure 8 shows the inventory of first tier suppliers’ critical components for traditional products and green products under different simulation scenarios. From the perspective of inventory at the first tier supplier level, unlike manufacturers, in S1 and S2, when supplier 1 and supplier 2 are disrupted, their own inventory shows an upward trend after the disruption event due to the inability to ship the critical components and the presence of intermediate components provided by the upper tier supplier in transit. In addition, the inventory of the other parties also showed an upward trend after the disruption event, but the upward trend was delayed by about 20 weeks due to demand transfer caused by the shortage of competitive products, while fulfilling new orders has a lag. In S3 and S4, when the superior suppliers of supplier 1 and supplier 2 experienced a disruption, their inventory sharply decreased until it reached zero, as they both had to fulfill the manufacturer’s order and could not obtain upstream components to replenish their inventory.
Figure 9 shows the inventory of second tier suppliers’ intermediate components for traditional products and green products under different simulation scenarios. From the perspective of inventory at the second tier supplier level, the curve indicates that regardless of whether it is their own disruption or the disruption of their downstream primary suppliers, their inventory will increase due to the disruption. Although any disruption event at any level will cause the supply chain’s orders to stop, the inventory fluctuations of second tier suppliers are more sensitive to the disruption of their own level suppliers, as they can still ship intermediate components to reduce inventory pressure during the disruption of their lower level suppliers.
Figure 10 shows the profit of each supply chain member service level. From the simulation results of profit, it can be seen that supplier 1 and supplier 3, as traditional product suppliers, have the highest profit in S2 and the lowest profit in S3, while supplier 2 and supplier 4, as green product suppliers, have the highest profit in S3 and the lowest profit in S2. Like the suppliers of green products, the manufacturer also has the highest profit in S3 and the lowest profit in S2. This means that the disruption of intermediate component suppliers (upstream suppliers) in traditional products causes the highest profit losses to traditional product suppliers, while the disruption of critical component suppliers (downstream suppliers) in green product causes the highest profit losses to green product suppliers and manufacturers. This enlightens us that if the enterprise’s goal is to pursue profit maximization, then traditional product suppliers should focus on improving the supply capacity of their upstream critical component suppliers, while green product suppliers and manufacturers should focus on improving the supply capacity of their upstream intermediate component suppliers. For manufacturers, the disruption of green products brings higher losses than the traditional products.
From the simulation results of service level, it can be seen that in the first 50th week, there was no disruption, and the service level in all four scenarios was the same. However, after the disruption in the 50th week, the service level in all four scenarios became different. The service level in S4 was the highest, while the service level in S1 was the lowest. Service level represents the degree of fulfillment of demand; it can be seen that the disruption of critical components in traditional product causes the highest risk of out-of-stock, while the risk of out-of-stock in intermediate component in green products is the smallest. If the manufacturer’s goal is to better serve demand, then they should prioritize the investment in improving supply capacity at supplier 1, followed by supplier 3 and supplier 2 and supplier 1 at last.
6.3. Multiple Suppliers Are Disrupted
The previous section analyzed the scenario of individual disruption, but in reality, many suppliers often have similar geographical locations and transportation capabilities. Therefore, simultaneous disruption by multiple suppliers is also inevitable. This section focuses on analyzing the impact of simultaneous disruption by multiple suppliers on the supply chain system. We use S5, S6, S7, S8, S9, S10, S11, S12, S13, S14, and S15 to capture these characteristics, which cover all possible disruption scenarios. Among them, S5, S6, S7, S8, S9, and S10 represent two suppliers experiencing disruption at the same time, S11, S12, S13, and S14 represent three suppliers experiencing disruption at the same time, and S15 represents all suppliers experiencing disruption. Due to the plethora of variables and scenarios, we will only discuss the manufacturer’s inventory, profit, and service level here.
From the simulation results in Figure 11, it can be seen that the manufacturer’s profit under S6 is the highest. That is to say, from a profit perspective, the manufacturer’s profit under S1, S2, S3, and S4 that only disrupted one supplier is not the highest. This conclusion is counterintuitive, which means that multisupplier disruption is not necessarily inferior to single supplier disruption. The reason for this result is that in the S6 scenario, all traditional product suppliers have been disrupted and consumers in the market are unable to purchase any traditional product. Therefore, consumers have entered the market for green products, and selling the green products can increase the manufacturer’s profit.
From the perspective of service level, the service level with the lowest is not S15, which has the highest number of disrupted suppliers, but S5 and S8. This is because in S5 and S8, there is a higher accumulation of order and midway shipments, which leads to the ripple effect of higher inventory fluctuations in the supply chain and cannot better restore the capacity and serve demand after the disruption disappears. Although all suppliers have been disrupted under S15, there is no logistics difference between them and they can recover faster after the disruption.
From the inventory fluctuations of the manufacturer’s two products, it is not S15 that causes a significant impact on inventory. The reasons for this result are the same as the analysis of service levels.
7. Conclusions
Uncertain events such as earthquakes, tsunamis, epidemics, and wars have increased the risk of supply chain disruption. Faced with the need for carbon reduction policies and environmental protection, a large number of enterprises have begun to produce both traditional and green products. Studying the issue of supply chain disruption for such enterprises has significant practical significance. Previous research has rarely focused on the disruption of green products in substitutable supply chains, and there is also little literature on the disruption of assembly supply chains in two-stage production. Considering these, we developed a three-level SD model to analyze the behavior of the supply chain under different disruption modes. The main conclusions drawn in this article are as follows:(1)Supply chain disruption can cause fluctuations in the manufacturer’s inventory curve, and disruptions from second tier suppliers have a higher impact on the manufacturer’s inventory than those from primary suppliers. In addition, the disruption of traditional products can cause consumers to flow to the green product market, leading to a sudden increase in orders for green products and components in a short period of time, causing a delayed impact on the inventory of various suppliers and manufacturers of green products. Therefore, after observing the disruption of competitive enterprises, enterprises should promptly increase their inventory to cope with the impact of delay.(2)The disruption of the intermediate component suppliers (upstream suppliers) in traditional products causes the highest profit losses for all traditional product suppliers, while the disruption of the critical component suppliers (downstream suppliers) in green products causes the highest profit losses for all green product suppliers and manufacturers. If the enterprise’s goal is to pursue profit maximization, then all traditional product suppliers should focus on improving the supply capacity of their upstream critical component suppliers, while the manufacturer and all green product suppliers should focus on improving the supply capacity of their upstream intermediate component suppliers.(3)From the perspective of service level, compared to other components, disruptions in the critical components of traditional products pose the highest risk of out-of-stock in the supply chain, while the risk of being out of stock in the intermediate component of the green product is minimal. If the manufacturer’s goal is to better serve demand, they should prioritize investing in improving supply capacity of critical component suppliers of the traditional products.(4)Common sense may suggest that the more the suppliers disrupted, the higher the loss. However, due to the ripple effect, this article finds that from the perspectives of profit, inventory, and service level, multisupplier disruption is not necessarily inferior to single supplier disruption.
Based on our study, we have distilled key managerial implications that can be valuable for practitioners. Allocating investment towards enhancing the supply capacity of upstream intermediate component suppliers enables manufacturers to attain higher profitability. Conversely, directing investment towards improving the supply capacity of critical component suppliers for traditional products can lead to higher service levels. Therefore, the decision on which supplier to prioritize for investment hinges on the manufacturer’s equilibrium between profit and service level objectives.
The research in this paper can be applied to many manufacturing enterprises that assemble and produce both green and traditional products. The constants’ value of the SD model in this paper is randomized and cannot be consistent with the actual data of each enterprise. However, the research simulation experiments and conclusions in this paper are based on qualitative analysis, so the specific values of these constants will not affect our conclusions. Despite the valuable insights gained from our research, it is essential to acknowledge certain limitations in our model. For instance, our model lacks constraints on inventory capacity, and there is no consideration of the interaction between price and demand. Future research endeavors could explore and expand upon these aspects to provide a more comprehensive understanding of the dynamics within the studied context.
Data Availability
All data that support the findings of this study can be obtained from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
This research was funded by the National Natural Science Foundation of China (71861028 and 72263023) and Jiangxi Provincial Graduate Innovation Special Fund (YC2022-s748).