An integrated methodology for assessing accident probability of natural gas distribution station with data uncertainty

Natural gas distribution station (NGDS) is an important facility established for providing natural gas to urban users along long-distance pipeline. The reasons, such as corrosion, material aging, misoperation, etc., lead to the leakage of the NGDS, and may cause serious fire or explosion hazards. Bow-tie analysis (BTA) is an effective and popular technique to identify hazards, evaluate risk and develop mitigative strategies in process industries. However, it is often difficult to acquire precise failure data for the BTA of the NGDS due to incomplete knowledge and insufficient statistical data. In this paper, considering the failure data uncertainty, an integrated methodology, which incorporates the BTA technology, fuzzy set theory and Bayesian updating technique, is developed for evaluating accident occurrence probability of the NGTS system. Fuzzy importance measure method is further recommended to find out the weakest links of system. A case analysis is carried out to verify the feasibility and effectiveness of the proposed methodology. This study gives support the risk management decision-making for the NGDS and also provides methodological reference for evaluating accident probabilities of other types of oil and gas stations. •A bow-tie accident model for natural gas distribution station (NGDS) is constructed.•NGTS accident spectrums are identified and accident minimal cut set (MCS) equations are established.•An integrated methodology is proposed to evaluate NGTS accident probabilities with data uncertainties.•Bayesian updating models for evaluating running failure rates and demand failure probabilities are clearly presented.