A new approach for quantitative risk assessment of gas explosions on FPSO

Abstract

Gas explosions on floating, production, storage and off-loading units (FPSO) cause catastrophic consequences. It is vital to perform an accurate risk assessment for preventing these hazards. This paper presents a comprehensive quantitative risk assessment approach for evaluating the whole gas explosion process from initial leakage to final explosion. A stochastic sampling method is proposed to determine the explosion scenarios by using historical statistics data. Secondly three operational barriers: gas detection, emergency shutdown and ignition prevention are arranged to prevent the development from leakage to explosion. The sequential Bayesian Network (BN) predictive model is constructed to describe the hazard development by considering Human and organizational factors (HOFs). As low as reasonably practicable (ALARP) principle is performed to assess the explosion risks. The proposed approach has been applied to a case study focusing on estimating the risk levels for humans and constructions. It turns out that HOFs have significant effect on the normal functions for operational barriers, if ignoring the HOFs effect the accident risks will be underestimated. This approach is especially good at updating the results whenever new data becomes available. Several safety measures can be recommended based on the diagnostic analysis of BN.