Accident analysis of Qinglong natural gas pipeline explosion based on fault tree and Bayesian network

作者：Bowen Shao, Jian Guo, Shuaipeng Ji, Weikang Han, Baikang Zhu, Bingyuan Hong · 发表于：Fifth International Conference on Traffic Engineering and Transportation System (ICTETS 2021) · 年份：2021 · DOI：10.1117/12.2620166 · 被引用次数：2 · 研究领域：Risk and Safety Analysis、Occupational Health and Safety Research、Structural Integrity and Reliability Analysis

An accident analysis method based on Fault Tree and Bayesian network is proposed to analysis the accident cause for the explosion of Qinglong natural gas pipeline. The method maps the fault tree to the Bayesian network and combines the causal modelling ability of Fault Tree analysis with the probabilistic dynamic updating ability of Bayesian network. The results demonstrate that the most influential factor of the explosion event is the "natural gas leakage", which is much higher than other factors such as "fire source" in the site of accident. Four key failure causation chains were identified by probabilistic updating. The identified key risk factors and failure causation chains can provide guidance for on-site hazard source identification and risk prediction.