| There are many chemical enterprises in China,the chemical industry is large in volume,and the inherent safety risks of enterprises with major hazards are high and difficult to control.The State Council has put forward the requirements of improving the ability to control major safety risks of hazardous chemicals,promoting the implementation of the main responsibility of enterprises and strengthening the ability of safety supervision in the three-year action of special rectification of hazardous chemical safety.In this context,based on complex networks and stakeholder-related theories and methods,this paper uses chemical accident cases as the basic data material to study the causal mechanism and network propagation evolution process of chemical accidents.(1)Construction and analysis of a directional weighted causal network model for chemical accidents.The causal factors and accident outcomes in the chemical accident cases were used as network nodes,the causal relationships between factors as edges,and the multiple edges between nodes as edge weights.The results show that the causal network of chemical accidents has scale-free characteristics.Based on the results of the topological indicators of the causal network,key nodes with global influence,node pairs with strong cascade relationships and nodes that promote cascade,and nodes that locally influence risk transmission in the network were identified.(2)Analysis of chemical accident causation networks considering cascade effects.The cascade effect in the causation network of chemical accidents is described by the edge weights between nodes,and the potential key risk factors with significant cascade effect in accidents are analysed.Based on the cascade transfer effect between nodes,the performance of the chemical accident causation network is compared and evaluated under different attack strategies.The results show that the cascading paths of key risk factors with significant cascading effects have a significant impact on the occurrence of chemical accidents,such as non-compliance/violation and equipment quality defects,and their failure causes the failure of 39.58% of nodes and 81.97% of edges in the network.The cascade transmission effect accelerates network collapse,and the connectivity efficiency of the network decreases most significantly when key nodes with strong global influence or edges carrying cascade transmission relationships in the network fail.(3)Stakeholder-based analysis of chemical accident cohesion networks.To identify and classify the stakeholders in chemical accidents based on the complex relationship between stakeholders and accident causation,the risk factor nodes in the accidents are coalesced according to the accident causation theory,so as to build a chemical accident cohesion network.A 2-mode network was constructed based on the co-occurrence relationship between stakeholders and causes of chemical accidents.The results show that there is a significant causal relationship between the unsafe behaviour of people in chemical accidents,which leads to an unsafe state of the equipment and an unsafe state of the environment(29%),and the unsafe state of the environment,which leads to a particularly serious accident and a serious accident(27%).The operators of the identified and expected stakeholders in a chemical accident are significantly responsible for the occurrence of chemical accidents. |