Operating and maintenance experience with computer-based systems in nuclear power plants - A report by the PWG-1 Task Group on Computer-based Systems Important to Safety

This report was prepared by the Task Group on Computer-based Systems Important to Safety of the Principal Working Group No. 1. Canada had a leading role in this study. Operating and Maintenance Experience with Computer-based Systems in nuclear power plants is essential for improving and upgrading against potential failures. The present report summarises the observations and findings related to the use of digital technology in nuclear power plants. It also makes recommendations for future activities in Member Countries. Continued expansion of digital technology in nuclear power reactor has resulted in new safety and licensing issues, since the existing licensing review criteria were mainly based on the analogue devices used when the plants were designed. On the industry side, a consensus approach is needed to help stabilise and standardise the treatment of digital installations and upgrades while ensuring safety and reliability. On the regulatory side, new guidelines and regulatory requirements are needed to assess digital upgrades. Upgrades or new installation issues always involve potential for system failures. They are addressed specifically in the 'hazard' or 'failure' analysis, and it is in this context that they ultimately are resolved in the design and addressed in licensing. Failure Analysis is normally performed in parallel with the design, verification and validation (V and V), and implementation activities of the upgrades. Current standards and guidelines in France, U.S. and Canada recognise the importance of failure analysis in computer-based system design. Thus failure analysis is an integral part of the design and implementation process and is aimed at evaluating potential failure modes and cause of system failures. In this context, it is essential to define 'System' as the plant system affected by the upgrade, not the 'Computer' system. The identified failures would provide input to the design process in the form of design requirements or design changes for the new installation or the upgrade. Procedures, training, and other practices, as well, may be affected by the failure analysis because administration controls, periodic calibration and surveillance procedures may all be used to provide defence against potential failures. An important input to the failure analysis activities comes from the feedback of operating and maintenance experience. Feedback of operating experience in nuclear power plants has long been recognised as a valuable source for improving system design, procedures or human performance to achieve safety and to prevent recurrence of failures. This is particularly true in the case of complex systems such as computer-based systems. The process of feedback would provide designers with information on systems failures, unforeseen scenarios, or unanalysed configurations. The review of operating experience and the identification of causes of failures is also essential for the regulators in performing their safety assessments. Currently, the NRC reviews the electromagnetic compatibility (EMC), software reliability, and the human-machine interface when it performs a safety evaluation of digital upgrades to ensure that the digital system failures resulting from the identified causes are within the acceptable level of a system's reliability. Operating experience with computer-based system is one of the topics raised in the SESAR report. The CSNI Bureau of the OECD has requested NEA Principal Working Group No. 1 (PWG1) to review this topic. A task group led by Canada was therefore formed within PWG1, including France, Japan, U.K., and U.S.A. to address the related issues. The purpose of this report is to summarise the observations and some findings related to the operating and maintenance experience, based on contributions from France, U.S.A., U.K., Japan, and Canada. Additional information from the review of the open literature is also included. A number of the operational incidents, selected as case studies, are presented as examples in Appendix 1. In addition, this report presents an example of an evaluation by the US NRC of the digital upgrades introduced at an operating nuclear power plant