Abstract
Natural gas processing is always a high-risk operation in the Floating liquefied natural gas facility (FLNG). This facility is considerably prone to unfavourable accidents, especially fires. This paper is aimed at developing an integrated model to conduct a comprehensive fire risk analysis in the FLNG platform. Fault Tree (FT) tool was employed to represent the root causes of LNG fire and the FT diagram was then converted into the Bayesian network (BN) to perform the dynamic analysis. Using BN model, a novel procedure was proposed to conduct the criticality analysis and identify the most critical BEs leading to the fire risk. The results indicated that the Human and Organizational errors (HOEs) are the most contributing factors to the LNG fire occurrence. The diagnostic and sensitivity analyses were implemented to figure out the most hazardous units to LNG release as well as the most probable paths of the fire evolution. The analyses revealed that the fire probability is highest in pipe network and liquefaction unit. Based on the findings, some appropriate safety procedures were proposed for lowering the fire probability in the FLNG processing facility.