A hybrid approach based on the HFACS-FBN for identifying and analysing human factors for fire and explosion accidents in the laboratory. (February 2022)

• Record Type: Journal Article
• Authors: Ma, Laihao; Ma, Xiaoxue; Xing, Pengfei; Yu, Fanyong
• Is Part Of: Journal of loss prevention in the process industries. Volume 75(2022)
• Publication Date: 2022-02
• Subjects: Fuzzy bayesian network -- HFACS -- Human factors -- Lab fire and explosion accidents; Chemical industries -- Safety measures -- Periodicals; 660.2804
• DOI: 10.1016/j.jlp.2021.104675
• Languages: English
• ISSNs: 0950-4230

Abstract

Abstract: Fire and explosion accidents are the most important crises in the lab at colleges and universities, and human factors (HFs) are widely regarded as the highly contributing factors to the occurrence of lab accidents. Therefore, it is necessary to use efficient and reliable methods to identify and monitor the key HFs that cause and affect lab fire and explosion accidents. In this paper, to identify the most critical and highly contributing HFs exposed in the lab fire and explosion accidents, a hybrid method integrating Human Factors Analysis and Classification System (HFACS), Fuzzy set theory (FST), and Bayesian network (BN) is applied, which can compensate for the static nature of conventional methods in HFs analysis and its inability to deal with uncertainty. The hybrid model was tested on 39 lab fire and explosion accidents from 2008 to 2020 in China and the United States, and the sensitivity analysis was also conducted to recognize the top 10 most critical root events associated with HFs leading to lab fire and explosion accidents. The results demonstrated that organizational influences are the leading contributors to the top 10 most highly contributing root events for lab accidents. Highlights: A hybrid method containing FBN-HFACS is applied for analysis of human factors in lab fire and explosion accidents. BN and FST were integrated into HFACS to address the uncertainties on human errors and organizational failures. The analytical power of HFACS in considering conditional dependencies among causal factors was promoted by using BN.