Investigation into the causes and network evolution of fire and explosion accidents in oil depots
To address significant issues such as the insufficient systematic investigation of hidden dangers and the lack of quantitative risk management and control methods in the current oil and gas storage and transportation industry,this paper employs complex network theory and methodology.It focuses on a typical historical case of a fire and explosion accident at an oil depot,analyzing the causes of the accident and the evolution of the network involved.Based on the establishment of the accident-causing network,the risk factors and final consequences of the historical fire and explosion accident at the oil depot are treated as the nodes of the network model.The accident chain is then analyzed and extracted.Additionally,the number of accidents is used as the weight for the edges between the nodes,allowing for the creation of a directed weighted accident-causing network model.Finally,important topological characteristics—such as degree and cumulative degree distribution,clustering coefficient,and betweenness centrality—are selected to quantitatively analyze the network from both overall and local perspectives.To avoid bias in assessing the importance of network nodes based solely on a single topological index,we combine the total degree value,clustering coefficient,betweenness centrality,and entropy method of the three topological indices to calculate the weights.This approach enables us to derive and rank the importance scores of the risk factor nodes.Key risk points in the accident evolution process have been identified,and corresponding risk control measures are provided for oil depot enterprises.Simultaneously,based on complex network theory,an analysis of the average shortest path length and network diameter reveals that the network exhibits a smaller average shortest path length,indicating that each risk factor is closely interconnected and demonstrates small-world characteristics.When a risk factor occurs,it propagates more rapidly within the network.The establishment of the accident-causing network model enhances our understanding of the mechanisms and dynamics underlying accidents,providing valuable insights for effective risk management.
safety engineeringoil depot fire explosioncomplex networkaccident evolutionnetwork topology
万方数据
维普
