A Novel Fuzzy Bayesian Network-HFACS (FBN-HFACS) model for analyzing Human and Organization Factors (HOFs) in process accidents

Abstract

Human and organizational factors (HOF) play a significant role in the accident occurrence in chemical process industries (CPI). Human Factors Analysis and Classification System (HFACS) is a comprehensive framework widely used for analyzing HOFs involved in accidents. HFACS, however, has been criticized due to limitations such as the lack of quantitative analysis and interdependencies consideration among causal factors, and reasoning under uncertain conditions. This paper presents a novel accident analysis model incorporating Bayesian network (BN) and fuzzy Best Worst Method (fuzzy-BWM) into the HFACS framework to overcome the mentioned limitations. In the proposed model, BN is used to promote the ability of HFACS in providing both quantitative assessment and consider conditional dependencies among causal factors, while fuzzy-BWM is applied to relax the difficulties related to uncertainties and insufficient data on human errors and organizational failures. Application of the model was tested for analysis of HOFs in a real accident. The results revealed the capability of the model to quantify the failures and to provide an HFACS framework characterized by a flexible and dynamic analytical capability. The model was also able to identify key safety measures for development of effective intervention strategies in order to prevent future similar accidents.