An integrated risk assessment modelling for cargo manifold process on tanker ships under FMECA extended Dempster-Shafer theory and rule-based Bayesian network approach

Risk assessment is one of the top requirements in maritime transportation due to the complexity of ship oper-ations, particularly on tanker vessels. By carrying hazardous liquid cargo, tankers pose significant risks for life, property and the marine environment, which should be assessed in detail. This paper focuses on the manifold process during tanker cargo operation since it involves various risks that may lead to severe consequences such as human injury, gas poisoning, hull damage, cargo spill or explosion. The process is modelled via Bayesian network, and marine experts evaluate 12 failure modes with respect to Failure Mode, Effect and Criticality Analysis parameters. To fuse the expert judgment, Dempster-Shafer theory is applied with a rule-based approach in the Bayesian model. The highest crisp risk value is found 48.85 for Failure Mode (FM) 1.1 (Improper arrangement of valves and pipelines to be connected). It is followed by FM 1.2 (Improper connection between the line and the manifold) with a crisp risk value of 45.83, and FM 4.1 (Failure in earthing & bonding condition of equipment to be used) with 38.26 crisp risk value. According to results, control actions are presented to reduce the risks during the manifold process. Beside its technical background, the paper provides utmost contributions to maritime safety inspectors, tanker ship crew, tanker ship operators and safety researchers to improve safety process and minimize the operational risks of cargo manifold process.

Automated summary

This paper focuses on risk assessment in tanker cargo operation and uses Bayesian network modeling to evaluate 12 failure modes. The results show that control actions can effectively reduce operational risks.