Analysis And Modeling Of Coupling Patterns Between Risks And Scenarios In Emergency Processes For Oil Storage System Fire

Oil storage system safety involves many aspects of the economy,people’s livelihood,national defense,the environment,etc.In recent years,there have been frequent oil storage system fire accidents.In multiple accidents,secondary accidents often occur due to improper emergency response,leading to an increase in the number of casualties,an increase in the scale of the accident,and an increase in environmental pollution.It indicates that the potential risk factors and their mutual coupling effects in the emergency process are important reasons for the recurrence of accidents.The interaction of these factors will generate a series of scenarios in the evolution of the accident,and the occurrence of these scenarios will lead to the recurrence of the accident.At present,the research on the oil storage system risk factors is mainly oriented to the risk factors of the first accident.There are few research results reported in the public for the risk factors that may lead to secondary accidents and possible accident scenarios.Analyzing the coupling pattern of potential risk factors in the emergency process and establishing the correlation model between emergency risk and accident scenarios will provide the theoretical basis for revealing the evolution mechanism of accident development.Through the statistical analysis of 252 typical oil storage system fire and explosion accidents in the past 50 years,17 main emergency process risk factors were summarized based on the 3 types of hazard source theory and from four aspects of human-physical-environmentmanagement;6 main accident scenarios were summarized from the perspective of possible consequences of accidents.The risk factors of the emergency process were divided into five major categories: human factor,physical factor(equipment factor,fire Fighting facilities factor),environmental factor,and management factor;and the coupling types of these risk factors were divided into single-factor,two-factor,and multi-factor coupling types according to the different numbers of factors involved in the same attribute in the risk coupling process.The risk coupling values under each coupling type were calculated and ranked using the N-K model,and the strength of coupling effects under different factor coupling was obtained.The more risk factors involved in coupling the greater the coupling degree value,and management factors and physical factors are more likely to be coupled with other factors.Based on this,a coupled effect analysis model based on Bayesian network and N-K model is constructed.Then,the constructed analytical model was applied to analyze the risk coupling effect,the changing characteristics of the risk coupling expectation under each coupling type,the influence of each risk factor,and the dependency relationship between risk factors on the accident risk coupling effect were obtained.The results showed that: management risk factors in single-factor coupling,equipment-management coupling type in the two-factor coupling,equipment-fire-management(three-factors)in the multi-factor coupling,equipment-fireenvironment-management(four factors),and people-equipment-fire-environment-management(five factors)have the greatest impact on the risk expectation of each coupling node.In order to quantify the relationship between risk factors and accident scenarios in the emergency response process,the basic laws of accident development and evolution are revealed.The correlations between risk factors and accident scenarios in the emergency process are analyzed with specific accident cases,and a structural equation model between them is established.The results show that human factor,physical factor,environmental factor and management factor have significant direct effects on both the fire expansion scenario and the fire scene damage scenario.The largest combined effect on the fire expansion scenario is the human factor,followed by the management factor,environmental factor,equipment factor,and fire factor;the largest combined effect on the fire scene damage scenario is the management factor,followed by the fire factor,environmental factor,human factor,and equipment factor.The risk coupling effect analysis is based on Bayesian network and N-K model,and the sensitivity analysis results are combined with Bayesian network.Using "Netica" software,we investigate how to minimize the probability of a risk coupling event among similar risk factors with a joint coupling relationship,and then analyze the prevention and control strategy.Based on the results,reliable and specific suggestions,and measures are proposed in five aspects:human behavior awareness,inherent equipment inspection,fire protection facility supervision,bad environment rectification,and improvement of management capability.The research results have theoretical guidance and practical countermeasure reference value for the emergency safety management of oil storage system fire accidents.