Dynamic flowgraph methodology used in reliability analysis of digital instrumentation and control system in NPP

Software and microprocessor based digital techniques have been widely applied in digital instrumentation and control system (DI and CS) of nuclear power plant (NPP). Many under construction or operation NPPs have utilized DI and CS to monitor and control some critical processes/parameters in a decade. The implementation of full DI and CS has greatly improved the performance of control system. Tradition Fault-Tree Analysis (FTA) cannot describe the dynamic interaction in DI and CS and dynamic flowgraph methodology (DFM) can describe these features. Therefore, DFM is adopted to assess the reliability of automatic power control system (APC) in NPP DI and CS. Given three initial conditions, namely sensor failure, SRB and main processor together fails and main/backup processor fails at present, analyzed inductively by DFM, A deviation of control rod position shift (CSP) is obtained by corresponding failure conditions. Given a small deviation of CSP state, analyzed deductively by DFM, the most probably basic event/fault combinations (Prime Implicants) contribute to the top event can be derived, namely the combination failure of SRB (output latch) switch off, output of Main processor failure low, the others are normal at present. (authors)