Locating urban CNG stations using quantitative risk assessment: using the Bayesian network

Due to the increasing need for fuel and the construction of compressed natural gas (CNG) refuelling stations in urban areas, the importance of accurate risk management in these stations has increased. This study is locating urban CNG stations using quantitative risk assessment in Hamedan, Iran using the Bayesian Network. In this study, the causes–consequences analysis of methane gas leakage from CNG storages was investigated by the Bowtie (BT) method. To quantitate this diagram, the Bayesian network (BN) was used. Finally, the quantitative risk was calculated and evaluated in a case study. Findings of the BT diagram showed that the probabilities (in year) of occurrence of VCE, Jet fire and Flashfire were 1.82 × 10−2, 1.532 × 10−2 and 1. 213 × 10−2, respectively. These values were calculated as 2.9 × 10−3, 2.2 × 10−2 and 1.9 × 10−3 in year, respectively, in the BN. The mortality area affected by these consequences was also calculated as 14.2, 44.5 and 1930 m2, respectively. The risk assessment results estimated the risk as unacceptable in most affected areas. Accurate risk management in process accidents requires high-precision risk assessment in the urban design phase. The findings showed that the use of the Bayesian network and PHAST modelling can be used in this risk assessment.