Dynamic Analysis of the Consequences of Gas Release in Process Industries Using Event Tree Technique and Bayesian Network

In process industries, storage tanks play an important role in the storage of a wide range of chemicals, compressed gas and other hydrocarbons. Gas release from these tanks can lead to catastrophic events with huge financial, human, and environmental consequences. In this study, a compressed gas tank was chosen as the unit of case study. Gas release as the top event was taken into consideration for quantitative and qualitative analysis of the probable consequences using the Event Tree Analysis and Bayesian network model. Analyzing the consequences of gas release from the tank by Event Tree Analysis (ETA), 6 safety barriers were detected to prevent the top event. The success and failure of these barriers could lead to 10 final consequences. Among the identified consequences, “near misses” were known to be the most probable consequence. The results of the study showed that safety barriers could significantly reduce the consequences of the occurrence of the top event. Bayesian networks can fix the static problem of quantitative risk analysis and provide the capability to determine the most probable consequences of the top event.