Analysis of Key Risk Factors in the Thermal Coal Supply Chain

Abstract

The thermal coal supply chain serves as core infrastructure for ensuring the safe and stable supply of electricity in China. Effective risk management and control of this supply chain are therefore critical to national energy security and socio-economic development. However, the thermal coal supply chain involves multiple complex risk dimensions, including cross-regional multi-entity coordination, a complex network structure, and a dynamic policy environment. Traditional risk analysis methods often fall short in depicting the concurrent events and dynamic propagation characteristics inherent to such a system. This necessitates systematically investigating the thermal coal supply chain within the Coal–Electricity Joint Venture (CEJV) operational framework, which primarily involves equity-based consolidation and long-term contractual coordination between coal producers and power generators, to comprehensively analyze its critical risk factors and transmission mechanisms. Initially, based on the integration of coal-fired power joint operation policy evolution and industry characteristics, 28 risk factors were identified across three dimensions: internal enterprise, external environment, and overall structure. These encompassed production fluctuation risks, thermal coal transport process risks, and insufficient supply chain flexibility. A dynamic behavior model for the thermal coal supply chain was constructed by analyzing the causal relationships among these risk factors, based on the operational processes of each link. Utilizing Petri net simulation technology enables a quantitative analysis of supply chain risks, facilitating the identification of bottleneck links and potential risk points. Through model simulation, 18 key risk factors were determined, providing a theoretical basis for optimizing supply chain resilience within CEJV enterprises. The limitations of traditional methods in dynamic process modeling and industrial applicability were addressed through a Petri net-based methodology, thereby establishing a novel analytical paradigm for risk management in complex energy supply chains.

1. Thermal Coal Supply Chain Overview and Identification of Risk Factors

A stable electricity supply is a critical foundation for national economic development, societal advancement, and the maintenance of civilian livelihood. As evidenced by statistics from the National Energy Administration (2020–2024), China’s coal-fired power capacity has demonstrated persistent growth, with annual increments illustrated in Figure 1.

By 2024, the national total installed power generation capacity reached approximately 3320 GW, including 1440 GW of thermal-power capacity. Notably, coal-fired power constituted 1190 GW of this thermal capacity (a 2.6% year-on-year increase), representing 35.7% of the nation’s total installed power generation capacity [1]. Despite constituting less than 40% of installed capacity, coal-fired power undertakes over 70% of peak load regulation responsibilities during supply–demand crises. This operational paradigm confirms its irreplaceable role as both a strategic pillar for grid stability and a regulatory ballast for energy security, safeguarding China’s socioeconomic development imperatives and residential electricity demands.

China’s thermal coal supply chain is characterized by a multi-province geographical distribution and a complex network structure involving governmental agencies, enterprises, and other stakeholders. The operational dynamics of this system are susceptible to heightened uncertainties arising from two primary sources: abrupt natural disasters and policy interventions initiated by regulatory bodies. Exogenous risk factors can propagate through the supply-chain network, thereby potentially jeopardizing the operational steadiness and efficacy of the thermal coal supply chain. Subject to the bullwhip effect, these stochastic events propagate within the supply chain and may even precipitate a comprehensive disruption of the entire chain’s functionality [2].

To effectively mitigate challenges within the thermal coal supply chain and guarantee the stability and sustainability of electricity provision, the National Development and Reform Commission and the National Energy Administration have promulgated a series of policy documents regarding coal–power integration since 2005. The coal–electricity production joint venture (CEJV) refers to a strategic business model that integrates coal mining, transportation, and power generation through equity linkage, long-term contracts, or strategic collaboration between coal producers and power utilities. This integration internalizes the supply chain from mining to generation, distinguishing it from arm’s-length market transactions. Consequently, coal–power integration has evolved into the prevailing development paradigm for enterprises in the coal–power industry, inducing substantial alterations in the topological structure of the thermal coal supply chain.

The advent of the dual-carbon policy, the volatility of the international political milieu, and the expedited reconfiguration of the global energy supply chain [3] have engendered novel risks for the thermal coal supply chain of CEJV enterprises. These newly emerged risks posed a significant threat to the stable operation of the thermal coal supply chain.

Therefore, the systematic identification of the critical risk factors in the thermal coal supply chain and the construction of a scientific analytical model have emerged as exigent practical issues demanding immediate resolution.

Traditional methodologies for supply-chain risk analysis, such as Fault Tree analysis [4], the Extended Exergy Accounting (EEA) method [5] and the SWARA method [6], encounter limitations in modeling dynamic interactions and concurrence events. In contrast, Petri nets, as a formalized modeling instrument, have progressively emerged as a crucial approach for supply-chain risk analysis, attributed to their remarkable descriptive capacity for concurrent, asynchronous, and distributed systems.

Previous research has leveraged fuzzy Petri nets to quantify risk transmission pathways within the supply chain, unveiling the cascading influence of node failures on the overall system [7]. Furthermore, by integrating stochastic Petri nets with Markov chains, a diagnostic model for supply-chain reliability has been proposed to quantitatively assess the impact of node failure probabilities on the system [8]. By analyzing the games between the supply chain members, a cost-sharing contract in the context of coal supply chain coordination under a linear demand function considering low carbon was preferred [9].

These scholarly investigations provide theoretical foundations for the risk modeling of the thermal coal supply chain. However, most existing research is centered on the manufacturing industry or general logistics scenarios. Consequently, inquiries into the industry-specific idiosyncrasies of the thermal coal supply chain remain insufficient.

Furthermore, recent studies have begun to explore the impact of evolving electricity market mechanisms on traditional energy supply chains. For instance, the emergence of shared energy storage and other flexibility resources can significantly alter demand patterns for thermal coal, introducing new dimensions of volatility [10]. While these studies highlight the evolving context, their integration into dynamic risk models for industry-specific coal supply chains, particularly under CEJV arrangements, remains limited.

To bridge this research gap, this study aims to address the following core scientific question: How do risks dynamically propagate within the thermal coal supply chain under the CEJV framework, and what are the key risk factors that determine the system’s resilience? To answer this, we employ a systematic approach that combines risk factor identification with dynamic modeling using Petri nets.

Hence, this study systematically dissected the operational procedures within the thermal coal supply chains of CEJVs, combining this with an analysis of their structural peculiarities to discern critical risk factors. A causality model for risk factors in thermal coal supply chains was constructed using Petri net simulation. The causal intensity among risk factors was quantitatively scrutinized, thereby pinpointing key risk determinants and offering a theoretical basis for risk assessment and management.

1.1. Analysis of Thermal Coal Supply Chain Operation Process

The thermal coal supply chain constituted a complex network system spanning from coal extraction to power transmission. It principally encompasses three primary segments: coal supply (upstream), transportation (midstream), and power-plant operation (downstream). In the upstream segment, coal-mining enterprises can be classified as self-owned mines and outsourced coal mines based on the extraction methodology. The midstream segment pertains to thermal coal logistics and transportation enterprises, which are categorized by transportation mode into road transportation, inland maritime and inland-sea shipping, imported coal shipping, railroad transportation, and others. The downstream thermal power planted serve as the consumption terminus of thermal coal and represent the ultimate link within the thermal coal supply chain. The upstream, midstream, and downstream links of the thermal coal supply chain and their interactive relationships are visually presented in Figure 2.

1.1.1. Upstream Overview of the Thermal Coal Supply Chain and Operation Process

(1)

Domestic coal supply

Domestic coal supply pertained to the upstream coal-production echelon, encompassing both self-owned coal mines and procured coal mines. The procurement of domestic coal within the context of coal–power integration was chiefly bifurcated into two components: issuance of thermal coal orders to self-owned coal mines, and placing orders with other domestic coal mines. The differences in procurement procedures between self-owned coal mines and outsourced coal mines are detailed in Figure 3, including key nodes such as order issuance, production scheduling, and quality inspection:

(2)

Imported coal supply

Imported coal supply pertained to the external sourcing of coal from overseas coal mines. Despite substantial domestic coal resources in China, a certain volume of coal is still imported annually, with the import volume showing a growth trend year by year. In 2024, for instance, 540 Mt of coal were imported [11] representing a 14.4% year-on-year increase. Among these imports, 240 Mt originated from Indonesia (accounting for 44.35%), and 95.22 Mt originated from Russia (accounting for 17.52%).

Within the framework of coal–power integration, imported coal was predominantly procured through the spot trading of thermal coal. The operational process of CEJV entities in procuring imported thermal coal is depicted in Figure 4.

1.1.2. Overview of the Midstream of the Thermal Coal Supply Chain and Operation Process

The midstream logistics of China’s thermal coal supply chain constitute a multimodal transportation system that transfers extracted resources from mining pits to demand centers. Current transportation modalities include roadway, railway, and maritime networks. Railway transport dominates the national coal logistics framework, supplemented by coastal/riverine shipping, while roadway transport serves primarily as last-mile connectivity. The main operational processes of thermal coal road transportation are shown in Figure 5.

(1)

Maritime Transportation

Shipping offers advantages such as high efficiency, low transportation costs, environmental friendliness, and strong safety. Its primary drawback lies in long transit times. The thermal coal shipping link encompasses self-owned shipping and third-party shipping. In the case of third-party shipping, CEJV entities entrust third-party shipping companies to complete the transportation of thermal coal.

(2)

Railway Transportation

Railway transportation is the most crucial means of thermal coal transportation in China. It ensures punctual delivery, features relatively low transportation costs, has a large one-time carrying capacity, and is less affected by environmental and climate changes. According to data released by the National Bureau of Statistics and the China Coal Industry Association, in 2024, the national railway freight volume reached 3.99 G [12], and the national raw coal production was 4.759 Gt [13]. The annual coal railway transportation volume accounted for 70% of the total railway freight volume and 58.8% of the annual coal production, indicating that railway transportation is the irreplaceable dominant mode of transportation for coal.

In the context of CEJV operations, some entities possess their own railways, which can achieve rapid transportation from coal-producing areas in the northwestern region to northern ports. Owing to rapid information communication and high-efficiency dispatching, the transportation efficiency of private railways was often higher than that of state-owned railways. However, CEJV operations are on a large scale, with substantial thermal coal material flows. Due to capacity constraints, private railways cannot fully meet the transportation requirements of coal–power integration. Despite its pivotal role, railway transportation is susceptible to significant vulnerabilities. Extreme weather events, such as blizzards in northern corridors or flooding in central regions, can cause track damage and signal failures, leading to prolonged delays. Structural fleet shortages of specialized coal wagons create bottlenecks, particularly during seasonal demand peaks. These vulnerabilities underscore that railway dominance does not equate to infallibility, and its risks must be critically accounted for in supply chain models.

(3)

Roadway Transportation

Road transportation is capable of implementing point-to-point direct conveyance, holding a significant position in short-distance thermal coal transportation from the mine pit. Nevertheless, road transportation of thermal coal gives rise to severe environmental pollution and is plagued by issues such as limited primary capacity and shortages of petroleum-based fuels. Consequently, relevant state departments vigorously advocate the shift in coal transportation from “road to rail.” As a result, the proportion of thermal coal road transportation has been diminishing steadily. Currently, it mainly serves the short-distance transportation from coal mines to nearby pithead power plants.

1.1.3. Downstream Overview and Operation Process of Electricity and Coal Supply Chain

The downstream operational architecture of the thermal coal supply chain constitutes a demand-driven procurement system. Within this system, coal–power integrated operators formulated strategic coal procurement plans through power market load forecasting models, followed by material flow execution encompassing supplier selection and inventory management. This system bifurcated into captive coal procurement and third-party coal sourcing mechanisms, each exhibiting distinct operational paradigms (as shown in Figure 6).

1.2. Thermal Coal Supply Chain Risk Factor Identification

Risk identification represents the most fundamental and pivotal step in supply-chain risk analysis, and scholars have extensively delved into this area, as presented in

Table 1. Risk factors affecting the stability of the thermal coal supply chain that emerged from the literature survey.

Author	Thermal Coal Supply Chain Risk Factors
Kuang Yang [14]	Policy change risk, inventory management risk, financial security risk, lower degree of informationization risk, demand fluctuation risk
Wang Qianqian [15]	Natural environment risk, economic environment risk, national policy risk, thermal coal production risk, thermal coal transportation risk, thermal coal inventory risk, capital operation risk
Liu Dongliang [16]	Price fluctuation risk, inventory risk, supply and demand side risk, transportation risk, policy risk, market environment risk, environmental risk
Yu Guoliang [17]	Planning process risk, raw material procurement risk, thermal coal production risk, thermal coal transportation process risk, return process risk
Mo Congying [18]	Mining process risk, processing risk, transportation process risk, storage risk, thermal coal sales risk
Tan Haiyan [19]	Produced electricity sales risk, thermal coal transportation risk, supply security risk, government regulation risk, economic and natural environment risk
Hallikas [20]	Demand-related risk, financial-related risk, pricing risk, delivery fulfillment capability

. The risk indicators of the thermal coal supply chain have been relatively comprehensively summarized in prior research. However, these were predominantly deliberated from the vantage point of coal enterprises.

Within the context of coal–power integration, the internalization of the supply chain fosters more intimate synergistic effects. In contrast, an ordinary coal company might be solely involved in a single segment of the supply chain, for instance, engaging exclusively in coal mining, transportation, or power generation. As a result, the supply-chain linkages are more frequently accomplished through external collaborations. Stakeholders within coal–power integration are consequently more concerned with internal-process risks, such as those associated with the planning process and production coordination, compared to other aspects.

Therefore, from the perspective of coal–power consortia, in addition to the general risks inherent in the thermal coal supply chain, risks within the power-generation link and risks related to the overall supply-chain structure must also be taken into account. These aspects were specifically analyzed for coal–power consortia in this study.

Through a comprehensive review and synthesis of existing research, the internal operational risks impacting the thermal coal supply chains within CEJVs were systematically examined. Particular emphasis was placed on the dimensions of logistics, capital flow, and information flow.

By synthesizing the risk factors identified in prior studies (Table 1), this study further incorporated the industry-specific characteristics of coal–power integration (e.g., internalized supply chain collaboration, cross-link coordination between coal mining and power generation). Specifically, we supplement risks related to power generation operations (e.g., poor operation in the power generation business) and supply chain structural rigidity (e.g., insufficient flexibility in the thermal coal supply chain), which were rarely addressed in previous research focused on single-segment coal enterprises. This integration results in a comprehensive list of 28 risk factors, categorized into internal operational risk, external environmental risk, and overall structural risk, as illustrated in Figure 7.

2. Petri Net-Based Supply Chain Risk Modeling

2.1. Causality Analysis of Supply Chain Risk Factors

The thermal coal supply chain inherently constituted a complex and dynamic system, wherein the origin of risks is neither fixed nor static. A perturbation induced by a specific risk factor has the potential to precipitate the occurrence of other supply-chain risks. Subsequent to the implementation of relevant national environmental protection policies, enterprises within each link of the supply chain adapt their production and operational strategies in compliance with these national policies.

Taking into account the operational processes of each link within the thermal coal supply chain, the causal relationships among the risk factors are analyzed and presented in Appendix A.

2.2. Petri Net-Based Supply Chain Risk Modeling

Petri net components, as defined in [21], consist of the following elements: tokens, which symbolize system resources; places (p), serving as repositories for resources; transitions (t), representing events that initiate the flow of resources; and directed arcs, depicting the pathways through which resources move upon the activation of transitions.

The initial token distribution for the Petri net model was calibrated through a structured qualitative process based on expert judgment, necessitated by the lack of publicly available, system-wide quantitative data for all risk factors. A panel of industry experts from coal mining, logistics, and power generation sectors within CEJVs was convened.

Experts were presented with the list of risk factors and asked to score the relative frequency and potential initial impact of each factor on a standardized scale (e.g., Low = 1, Medium = 2, High = 3). The scores were aggregated, normalized, and mapped to a discrete token scale (0 to 3). A token value of 3 was assigned to factors consistently rated as having both high frequency and high immediate impact. Applying this method, place P17 (‘Risk of failure of coal mine production equipment’) was assigned an initial three tokens. This assignment reflects the consensus expert view that equipment failure is a prevalent and acutely disruptive issue in daily mining operations, with a high propensity to trigger initial cascading effects within the integrated supply chain. The initial token assignment rules are detailed in Appendix B, ensuring the transparency of the model’s parameter settings. Transitions encapsulated the activities or events occurring within the supply chain, including coal extraction and transportation processes. Arcs illustrated the relationships between states and activities.

Drawing upon the causal analysis of the risk factors, a dynamic-behavior Petri net model for the thermal coal supply chain was developed, as depicted in Appendix C.

The simulation experiments were conducted over 1000 independent runs to ensure statistical significance. A burn-in period of 100 runs was implemented to allow the model to stabilize. The system was considered to have reached a steady state when the variance in the token distribution across all places fell below a threshold of 0.01 for 50 consecutive runs. These parameters were set to guarantee the reproducibility and robustness of the simulation outcomes.

3. Analysis of Supply Chain Risk Factors Based on Petri Net

Based on the developed Petri net model, the critical nodes and risk factors within the thermal coal supply chain were discerned. By analyzing the loops and variances in the model, the bottleneck links and latent risk points in the supply chain were identified. Employing Petri net simulation techniques, supply-chain risks were quantitatively evaluated.

A token was introduced into the place representing each third-level risk factor. The model was simulated for 10,000 independent runs to ensure statistical significance. A burn-in period of 1000 runs was used to allow the system to stabilize, and a steady state was assumed when the coefficient of variation for the token counts in all key places fell below 2% over a moving window of 100 consecutive runs. Through the statistical analysis of the average number of tokens in each place across multiple simulation experiments, it was observed that a larger average number of tokens indicates that, during the system-operation process, the corresponding risk factor was more significantly influenced by other factors. The simulation results are presented in Figure 8.

Notably, the ‘clean energy substitution risk’ (P25) exhibits a high average token value (0.2673), indicating its significant susceptibility to influence from other factors. This risk is primarily triggered under conditions such as stringent carbon emission policies, technological breakthroughs in renewable energy storage, and seasonal peaks in hydropower or wind generation, which collectively suppress the demand for thermal power and propagate financial and operational pressures upstream through the supply chain.

In the Petri net model, the smaller the average token number of a particular library, the higher the degree of independence of the risk factors represented by that place. Such risk factors convey a greater amount of information in the context of risk assessment.

On the premise of maintaining the second-level risk factors, the third-level risk factors were filtered by retaining those with an average token number below the 25th percentile of all simulated values. This threshold was selected to identify factors with high independence and primary influence within the risk network. While preserving the hierarchical structure of risk factors in coal–power linkage and based on the simulation outcomes of the Petri net, the indices of the third-level factors that were less affected by other factors were selected. That is, among the third-level risk factors under each second-level factor, those with a small average token number were chosen. Subsequently, the key risk factors of the thermal coal supply chain were determined, as presented in Figure 9. Compared with traditional risk analysis methods such as Fault Tree analysis [4], which focuses on static logical relationships, the Petri net model in this study effectively captures dynamic risk propagation (e.g., the cascading effect of ‘Natural disaster risk’ on transportation and production links). This advantage aligns with the findings of Qin [7] who emphasized the value of Petri nets in simulating asynchronous events, but our study further tailors the model to the thermal coal supply chain by incorporating industry-specific risks (e.g., coal mine rectification policy risk) that were overlooked in general logistics research [9]. Additionally, unlike previous studies on coal supply chains [14,16], which focused on single-risk dimensions (e.g., price fluctuations), our identification of 18 key factors (including structural risks such as ‘Insufficient flexibility in the supply chain’) provides a more holistic perspective for resilience optimization.

4. Conclusions

Through a comprehensive and systematic review of CEJV operation policies and in-depth examination of industrial characteristics, this study identifies a total of 28 risk factors from three distinct dimensions: internal operations, external environments, and structural integrity. These risk factors encompass production volatility risks, transportation process risks, and supply chain rigidity risks, among others.

By leveraging the operational processes across various segments of the supply chain, the causal relationships among these risk factors were meticulously analyzed. Subsequently, a dynamic behavioral model was constructed to explore and elucidate the risk transmission mechanisms within the thermal coal supply chain.

Petri net simulation technology was employed to quantitatively analyze the supply chain risks. The analysis revealed that risks related to policy changes, production safety, and supply chain inflexibility were among the most independent and influential. As a result, 18 key risk factors were effectively screened out based on their relative independence and influence within the network.

These key risk factors offered a solid decision-making foundation for coal–power integration enterprises to formulate risk-coping strategies. They also assisted enterprises in optimizing the resilience of the supply chain by focusing on these key risk factors, thereby enhancing the enterprises’ capabilities to navigate and adapt to the complex market environment. These findings provide CEJV managers with a prioritized focus for resource allocation and resilience planning, particularly in mitigating policy-driven and operational disruptions.