Reliability assessment of marine floating structures using Bayesian network

Marine floating structures are widely used in various fields of industry from oil and gas to renewable energy. The predominant dynamic responses of these structures are controlled by mooring lines. In recent years, a number of high-profile mooring failures have highlighted the high risk of this element in floating structures. A reliable design of mooring lines is necessary to improve the safety of offshore operations. This paper proposes a novel methodology to conduct reliability analysis of moored floating structures using Bayesian network (BN). The long-term distributions of extreme responses of the floating object are estimated using analytical frequency domain method, while mooring failure probability is estimated using limit state function in the proposed BN framework. Application of the methodology is demonstrated by estimating the failure probabilities of a floating cylinder with tensioned mooring system. The proposed study also explains how the hydrodynamic and reliability analysis could be integrated with BN to assess the overall safety of the offshore structures. The methodology presented can be employed to mitigate associated risk with marine structures brought about by stochastic hydrodynamic loads.