Probabilistic risk assessment for inert gas system on oil tanker ships using system theoretic accident model and process (STAMP) and Bayesian belief network (BBN)

As shipboard systems increasingly prioritize automation over human intervention, risk assessment models must adapt to this paradigm shift. The proposed framework addresses this need by focusing on software, hardware, and external factors including human factors, aligning with modern technological dependencies. This research conducts an extensive risk analysis of the inert gas system (IGS) on oil tankers by adopting system theoretic accident model and process (STAMP) and Bayesian belief network (BBN). While the STAMP identifies failure scenarios through a hierarchical control and feedback structure, BBN quantifies the failure probabilities based on STAMP outcomes. The study identifies critical failure pathways through STAMP's systemic hazard analysis and BBN's probabilistic quantification and calculates the system failure probability as 1.29E-01. The results indicate that the most critical failures in IGS are “Flame instability or burner failure”, “Inert gas blower fan failure”, and “Insufficient pressure during operation” from the hardware component. The methodology enhances predictive accuracy and provides actionable strategies for mitigating risks in increasingly automated maritime operations. The research outcomes are expected to provide valuable insights for maritime safety managers, safety inspectors, technical inspectors, HSEQ managers, and ship crews to improve operational safety as well as prevent potential risks for inert gas incidents on-board oil tankers.