Markov analysis of redundant standby safety systems under periodic surveillance testing

摘要

In modern applications of probabilistic safety assessment (PSA), maintenance planning and changes to technical specifications play an important role, not least due to regulatory requirements. In particular, standby safety systems under periodic surveillance testing are at the center of this issue. Since traditional PSA techniques impose limitations when complex maintenance and repair strategies are to be taken explicitly into account, we introduce continuous time Markov models to discuss various strategies for organizing repair and testing of two-train standby safety systems, which have the potential to replace traditional system models based on fault tree techniques in PSA. Besides a conventional steady state analysis of these Markov models, we provide a general numerical method which allows the calculation of the probability of exceeding allowed outage times of equipment in Markov models of safety systems, and we apply it to the models introduced in the present paper.