Scenario Deduction on Fire Accidents for Oil­–Gas Storage and Transportation Based on Case Statistics and a Dynamic Bayesian Network

Abstract

In order to solve the problem of unclear evolution paths after oil–gas storage and transportation accidents, which lead to a lack of targeted accident emergency, delayed disposal measures, and further deterioration of the accident, based on the existing accident scenario-response theory, this article summarizes 17 basic scenarios in oil–gas storage and transportation by analyzing 116 accident cases. Moreover, the accident scenario expressions are given in time and space dimensions, and general scenario evolution paths of fire accidents for oil–gas storage and transportation are constructed. On this basis, a dynamic scenario deduction network model is established by using a dynamic Bayesian network. In this model, key scenario nodes and their final scenario probabilities are determined in consideration of statistical probability obtained by the actual accident cases, empirical probability given by domain expert’s scoring, and state probability calculated by joint probability formula of a dynamic Bayesian network. As a case study, the scenario deduction process and result of a Dalian 7.16 accident are analyzed to verify the rationality and effectiveness of the proposed method. According to the proposed scenario deduction method, it can help decision-makers to make more targeted emergency disposal measures.