Life-Cycle Failure Risk Assessment of Subsea Shells Using AcciMap and Dynamic Object-Oriented Bayesian Networks

Abstract Shell structures play a critical role in marine energy development, storage, and utilization. This paper presents a life-cycle failure risk assessment approach for subsea shells, implemented using AcciMap and the dynamic object-oriented Bayesian network (DOOBN). The method enables complex systems to be represented in a modular and hierarchical manner. Based on life-cycle assessment, subsea shell’s life cycle is divided into four stages: design, manufacturing, deployment, and service. AcciMap model systematically identifies risk factors throughout the design-to-service process. The causal logic and interactions among these risk factors are analyzed to construct dynamic Bayesian network models for failure scenarios at each life-cycle stage. These models are integrated into a DOOBN model for life-cycle risk assessment of subsea shell. Dynamic failure probabilities of subsea shell failure are evaluated, and the critical contributing factors are identified. The approach is illustrated through a case study, which can serve as a useful tool for risk assessment and maintenance decision-making for subsea shells.