Dynamic Bayesian networks based performance evaluation of subsea blowout preventers in presence of imperfect repair

Abstract This paper presents a quantitative reliability and availability evaluation method for subsea blowout preventer (BOP) system by translating fault tree (FT) into dynamic Bayesian networks (DBN) directly, taking account of imperfect repair. The FTs of series system and parallel system are translated into Bayesian networks, and extended to DBN subsequently. The multi-state degraded system is used to model the imperfect repair in the DBN. Using the proposed method, the DBN of subsea BOP system is established. The reliability and availability with respect to perfect repair and imperfect repair are evaluated. The mutual information is researched in order to assess the important degree of basic events. The effects of degradation probability on the performances are studied. The results show that the perfect and imperfect repairs can improve the performances of series, parallel and subsea BOP systems significantly, whereas the imperfect repair cannot degrade the performances significantly in comparison with the perfect repair. To improve the performances of subsea BOP system, eight basic events, involving LWHCO, LLPR, LCC, LLICV, SLPSV, LRPIL, PIHF and SVLPLE should given more attention, and the degradation probability of basic events, especially the ones with high sensitive to system failure, should be reduced as much as possible.