Quantifying Software Safety in Nuclear Power Plants: A Framework for Requirements Phase Assessment

Boyuan Li,Duo Li,Jianghai Li,Chao Guo,Huasheng Xiong,Shuqiao Zhou,Xiaojin Huang

DOI: https://doi.org/10.1115/icone31-135382

2024-01-01

Abstract:Abstract With the increasing dependence on digital instrumentation and control (I&C) systems in nuclear power plants, software has become a significant determinant of system safety assurance. To expand the use of digital technology in the nuclear industry, systematic methods are required for quantifying the safety of software-based I&C systems in safety critical applications. A software safety quantification model limited to the requirements phase is built in this paper based on the causal mechanisms that challenge safety. A preliminary mathematical method was developed to assess the number of requirements faults, their sub-types, their hazard rates and a control gate that allows determination of when to move beyond the requirements phase. A case study is conducted on a function of a reactor protection system to verify the validity of the quantification model. The framework allows us to assess software safety quantitatively.