Application of Bayesian approach to the assessment of mine gas explosion. (July 2018)

• Record Type: Journal Article
• Authors: Tong, Xing; Fang, Weipeng; Yuan, Shuaiqi; Ma, Jinyu; Bai, Yiping
• Is Part Of: Journal of loss prevention in the process industries. Volume 54(2018)
• Page Start: 238
• Page End: 245
• Publication Date: 2018-07
• Subjects: Mine gas explosion -- Dynamic assessment -- Emergency response -- Bayesian network -- Delphi method; Chemical industries -- Safety measures -- Periodicals; 660.2804
• DOI: 10.1016/j.jlp.2018.04.003
• Languages: English
• ISSNs: 0950-4230

Abstract

Abstract: Frequent mine gas explosion accidents in recent years have caused catastrophic casualties and economic loss in China. In this paper, based on expert knowledge with treatment by the Delphi method to determine conditional probabilities, a Bayesian network (BN) has been developed to investigate the factors influencing mine gas explosion accidents. Based on case analysis of typical mine gas explosion accidents and further evaluation by experts, twenty BN nodes are proposed to represent mine gas explosion process from occurrence causes to explosion impacts, and final consequences. The results of case studies and Sensitivity Analysis (SA) with the proposed Bayesian model indicate that the integration of Bayesian network and Delphi method is an effective framework for dynamically assessing mine gas explosion accident, which could provide a more realistic assessment for emergency decision-making on mine gas explosion disaster response and loss prevention. Highlights: A Bayesian model for assessing mine gas explosion is proposed. Twenty nodes for representing mine gas explosion from causes to consequences are given. The assessment framework is of significance for emergency response decision-making.