Dynamic Bayesian networks based approach for risk analysis of subsea wellhead fatigue failure during service life

摘要

Subsea wellhead is a critical component of the drilling and production system in offshore oil and gas industry. Excited by cyclical fatigue loadings due to environmental forces, the wellhead is prone to fatigue failure, which could lead to the loss of well integrity and even catastrophic accidents. Although fatigue failure probability of the well-head carries an elevated uncertainties, it will definitely increase with the accumulation of fatigue in wellhead. This paper presents a fatigue failure risk analysis approach based on dynamic Bayesian Networks, aiming to predict the fatigue failure probability of the wellhead during service life. The proposed model can use the previously accumulated fatigue of the wellhead to probabilistically predict the present failure risk under present dynamic conditions. The practical application of the developed model is demonstrated through a case study. Adopting the predictive, diagnostic analysis techniques in the Bayesian inference, the dynamic fatigue failure probability of the wellhead at any time slices was achieved, and the most influential factors were figured out. Finally, the corresponding safety control measures are proposed to effectively mitigate the fatigue failure risk of subsea wellhead during service life.