Unavailability model for demand-caused failures of safety components addressing degradation by demand-induced stress, maintenance effectiveness and test efficiency

The reliability, availability and maintainability (RAM) modelling of safety equipment has long been a topic of major concern. Some RAM models have focused on explicitly addressing the effect of component degradation and surveillance and maintenance policies, searching for an optimum level of the safety component RAM by adjusting surveillance and maintenance related parameters. As regards the reliability contribution, these components normally have two main types of failure mode that contribute to the probability of failure on demand (PFD): (1) by demand-caused and (2) standby-related failures. The former is normally associated with a demand failure probability, which is affected by the degradation caused by demand-related stress. Surveillance testing therefore not only introduces a positive effect, but also an adverse one, which it compensates by performing maintenance activities to eliminate or reduce the accumulated degradation. This paper proposes a new model for the demand failure probability that explicitly addresses all aspects of the effect of demand-induced stress (mostly test-induced stress), maintenance effectiveness (PAS or PAR model) and test efficiency. A case study is included on an application to a typical motor-operated valve in a nuclear power plant.