Identifying and modeling the causes of trench collapse accidents in gas pipeline excavation and piping operation using Bayesian Network (BN) and Fuzzy DEMATEL

Introduction: Identifying and modeling the root causes of accidents can play an important role in preventing them. The purpose of this study is to identify and model the causes of gas pipeline excavation and piping operation accidents using the Bayesian network (BN) and fuzzy DEMATEL. Material and Methods: In this study, industrial accidents during gas pipeline excavation and piping operations were analyzed using the Bowtie method. The fuzzy DEMATEL method was employed to determine relationships between accident root causes, and the fuzzy AHP method was used to compare pairs of causes and determine their weight. Finally, Bowtie and DEMATEL outputs were mapped in Bayesian networks to determine the important risk factors for accidents. Results: The most important risk factors for trench collapse accidents were as follows: risk management (16% impact weight), competency assessment (14.2% impact weight), supervision (13.8% impact weight), work permit system (13.7% impact weight), compliance with requirements and guidelines (13.4% impact weight), training (11.4% impact weight), HSE system (9.5% impact weight), and contractor management (8% impact weight). Conclusion: Based on the results, it was demonstrated that risk management and competency assessment, having the highest weight percentages, play the most significant roles in the occurrence of trench collapse accidents. The findings of this study can inform the prioritization of corrective measures to prevent trench collapse accidents in gas pipeline excavation and piping operations.