LAUSR

ABSTRACT Reliability of the digital reactor protection system (RPS) is intensively researched as it is designed and installed to ensure the safety and economy which can be measured respectively by the probability of failure on demand (PFD) and probability of spurious trip (PST). Meanwhile, by analyzing the failure modes of the digital RPS, the failure on demand and spurious trip are the two main modes that should be evaluated for the reliability of digital RPS. Therefore, this paper develops the PFD and PST calculation formulas considering the module repair time as the repair takes some time, and during the repair duration, the digital system is operated in the degraded configuration and the common cause failure (CCF) which would severely impact the system in the event of occurrence. Considering the failure phenomenon of the digital RPS, the binomial failure rate (BFR) model is adopted for CCF. And the fault-tolerance techniques and their fault coverage are considered when calculating the PFD and PST. The quantitative results show that, in the example, CCF dominates the PFD while CCF is one of the major factors that result in PST but the main contributor is the independent failure. Also it can be concluded that the discovery time for the undetected failures dominates the PFD and PST when it costs long time to discover the failures even though the uncovered failures are very few. Thus, the failures should be covered by the fault-tolerance techniques as much as possible when designing the digital RPS.