Research On Reliability Analysis Methods Of Digital I&C System For Nuclear Power Plant

In the nuclear power industry, the use of digital instrumentation and control(I&C) system is an important feature of advanced reactor. Digital I&C systems offerthe potential to improve plant safety through features such as increased hardwarereliability and stability, reduced the human errors and improved failure detectioncapability. Currently, operating NPPs are replacing some of their analog I&C systemwith digital ones. New NPPs under construction and consideration also will integratedigital systems into their design.At the same time, some new problems raised in the process of using digital I&Csystem. A large number of I/O cards, microprocessors and supporting software areneeded to support the use of digital I&C system, and the software and hardware willbe linked together by the logic design to achieve the system’s function, which may befailed due to shortcomings in design or the failure trigger under result of particularhybrid input. Therefore, digital I&C systems are believed to improve the safety ofplant generally, but this belief has yet to be validated by a systematic assessment ofthe impact of these systems on the safety of NPP.There are presently no universally accepted methods for modeling digitalsystems in current-generation PSAs. This paper will focus on the feasibility of usingtraditional and dynamic reliability assessment methods for digital systems. In thestudy, several example system models have been constructed and have been evaluatedby using of conventional fault tree (FT) approach, Dynamic Flowgraph Methodology(DFM) and Markov/CCMT model approach respectively, and the results arediscussed in details too.The conventional FT analysis method has been widely used in the analysis ofNPPs PSA, which combine with system components’ failure modes in order tomodeling the digital system’s failure. However, conventional FT approach appears tobe too conservative and inadequate to model the time mechanism of event sequenceexplicitly, etc. Therefore it is necessary to establish some dynamic approach to modeling of digital I&C systems.DFM approach can model the relationship between variables and time sequenceof system properly because of its dynamic nature, therefore DFM can be used toevaluate the impact of software failure, hardware failure and environmentalconditions on the behavior of system synthetically. The Markov/CCMT modelapproach is able to integrate software’s ability to mask hardware faults. A completeMarkov/CCMT model contains the transfer chains of all the discrete state of system,and an integral picture of the structure of system can be built through the teansferprocess of these discrete states.Finally, a review of the-state-of-the-art of the methods and tools that canpotentially be used to model software reliability has been performed in this paperand the application and principle of fault injection technology (FIT) for the digitalsystem software analysis has been studied especially..