A Study About Software-Implemented Fault Injection Strategy for Digital RPS in Nuclear Power Plant

With the widely use of digital instrumentation and control (I&C) system in nuclear power plant (NPP), the safety and economic efficiency of NPP could be affected directly by the performance (reliability, fault tolerance etc.) of software, an effective evaluation technology plays critical role in ensuring the safety operation of NPP. At present, the main evaluation method for digital Reactor Protection System (RPS) is uncomprehensive that only pay attention to the correctness of logic function. In order to improve the evaluation method, this paper make a study about bringing in Software-Implemented Fault Injection technology, which inject fault inside the software artificially, the reliability and fault tolerance of software can be evaluated while the inside errors are activated. Based on the practice about Yang-Jiang units 5 and 6 projects, which is a generation II+ pressurized water reactor named CPR1000, this study analyze the RPS architecture, network transmission model, CPU design, data types, and software code, build the fault model with types, positions, and test items. As a result, achieves a software-implemented fault injection method for RPS, provides a comprehensive and advanced evaluation strategy for equipment in NPP.