Performance evaluation of subsea blowout preventer systems with common-cause failures

Subsea blowout preventer (BOP) stack configurations are not determined under the present international standards, and both the retrievable pods and non-retrievable pods are used in current projects. This work investigates the effects of stack configurations and mount types on subsea BOP systems from the viewpoint of reliability. A model based on the Markov method for performance evaluation is presented. The subsea BOP system is split into seven independent modules because of its complexity. With respect to common-cause failures, the corresponding Markov models are subsequently developed. The availability and reliability of the system are evaluated by merging the independent Markov models with the Kronecker product approach. The results show that one more subsea ram preventer does not greatly improve the performances of the subsea BOP system, whereas one less subsea annular preventer reduces the performances in the long run. Retrievable control pods markedly improve the performances. Therefore, the subsea BOP system with two annular preventers, four ram preventers, and retrievable pods should be selected preferentially. In addition, the effects of varying of failure rates and mean time to repair for subsea control pods and control stations on the availability and reliability are greater than those for other components, which suggest that the two components should be given more attention when designing subsea BOP systems.

► A Markov model for performance evaluation of subsea BOP system is presented.
► One more subsea ram cannot markedly improve the performances of subsea BOP.
► The mounts for subsea control pods markedly affect the performances.
► The failure rate and MTTR of pods and control stations have significant effects.