Exploring hazardous chemical explosion accidents with association rules and Bayesian networks

•Hazardous chemical explosion accidents are analyzed according to a hybrid approach.•Fault tree model is constructed using the association rule approach.•The influence of subjective factors in Bayesian network analysis is reduced by the accident data.•The process design and operation events have the...

Full description

Saved in:
Bibliographic Details
Published in:Reliability engineering & system safety Vol. 233; p. 109099
Main Authors: Li, Xin, Chen, Chao, Hong, Yi-du, Yang, Fu-qiang
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.05.2023
Subjects:
Association rule
Bayesian network
Explosion accidents
Fault tree
Hazardous chemical
Hazardous chemical
Explosion accidents
Association rule
Fault tree
Bayesian network
ISSN:0951-8320, 1879-0836
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•Hazardous chemical explosion accidents are analyzed according to a hybrid approach.•Fault tree model is constructed using the association rule approach.•The influence of subjective factors in Bayesian network analysis is reduced by the accident data.•The process design and operation events have the greatest impact on the occurrence of accident.•Improper temperature control is the most likely to provoke accident directly. Hazardous chemical accidents pose a severe threat to human and economic activities. To explore the causes and paths of hazardous chemical explosion accidents (HCEAs) in China, a qualitative and quantitative model was proposed in this study. Firstly, 129 accident cases were collected from the official website of the Ministry of Emergency Management of the People's Republic of China, and the accident factors were screened based on association rule (AR). Then, the fault tree (FT) structure of HCEA was developed by the AR, and the qualitative model was transformed into a quantitative Bayesian network (BN) model. Finally, the accident path, the importance size of the basic events, and the sensitivity of the direct cause of the accident were uncovered based on the BN model. The results show that improper temperature control due to operating procedures is the most likely route to accidents. Among the basic events, process management and operating procedures issues ranked the first in importance, with the most significant impact on HCEA. The sensitivity of improper temperature control far exceeds other direct causes and is most likely to induce accidents directly. According to the study results, some feasible measures are put forward to improve the safety management of hazardous chemical production.
ISSN:0951-8320
1879-0836
DOI:10.1016/j.ress.2023.109099