Quantitative reliability assessment in the safety case of computer-based automation systems

An essential issue in the construction of new or in the replacement of the old analogue automation applications in nuclear power plants is the reliability of computer-based systems, and especially the question of how to assess their reliability. The reliability issue is particularly important when the system under assessment is considered as a safety-critical system, such as the reactor protection system. To build sufficient confidence on the reliability of computer-based systems appropriate reliability assessment methods should be developed and applied. The assessment methods should provide useful and plausible reliability estimates, while taking the special characteristics of the reliability assessment of computer-based systems into consideration.The Bayesian inference has proved to be an efficient methodology in the reliability assessment of computer-based automation applications. Practical implementation of Bayesian inference, Bayesian networks, allow the combination of the different safety arguments concerning the system and its development process to a unified reliability estimate. Bayesian networks are also a convenient way to communicate on the safety argumentation between various participants of systems design and implementation as well as between the participants in the licensing processes of computer-based automation systems. This study is a part of the research project “Programmable Automation System Safety Integrity assessment (PASSI)”, belonging to the Finnish Nuclear Safety Research Programme (FINNUS, 1999–2002). The project aimed to provide support for the authorities and utilities in the licensing problems of computer-based automation systems. Particular objective of the project was to acquire, develop and test new and more cost-effective methods and tools for the safety and reliability assessment, and to gather practical experience on their use in order to achieve a more streamlined licensing process for the computer-based automation systems.The project is financed together by the Radiation and Nuclear Safety Authority (STUK), the Ministry of Trade and Industry (KTM) and the Technical Research Centre of Finland (VTT)