Evaluation of Bayesian Networks for Risk Assessment of Major Hazards in Oil and Gas Facilities

Abstract

The research focuses on applying the principles of BN for the most common equipment in the industry namely; loss of containment scenarios of oil and gas separator, hydrocarbon pipelines, Floating roof and Cone roof hydrocarbon storage tanks and centrifugal compressors. newlineWhat are the aspects that makes BN attractive for risk assessment? Fundamentally, it is the ability of BN to describe causal mechanisms (cause and effect) in a clear and visually understandable way that has made it the prime choice. BN can describe the complex interactions and inter relationships of cause and effect at various levels quite easily. Further, it can incorporate static probability numbers or more realistic probability distributions for failure rates. newline