Modelling of DIgital I&C, MODIG: Interim report 2015

Stefan Authén, Ola Bäckström, Jan-Erik Holmberg, Markus Porthin, Tero Tyrväinen

    Research output: Book/ReportReport

    Abstract

    The NKS-project MODIG (MODelling of DIGital I&C) aims to get a consensus approach for a reliability analysis of a plant design with digital I&C, improved integration of probabilistic and deterministic approaches in the licensing of digital I&C, improved failure data collection including software failure probability quantification, and a practical application of probabilistic safety assessment (PSA). A survey of the defence-in-depth (DiD) framework and PSA's role in it has been made. The assessment of DiD and diversity is in principle straightforward for PSA, e.g., risk metrics can be used to evaluate DiD levels 3, 4 and 5. A PSA model always includes uncertainties, which needs to be accounted for especially when comparing with deterministic assessment. Regarding digital I&C, the focus of the assessment is on the DiD levels 1, 2 and 3. In addition the logic model of PSA can be used in the assessment of deterministic failure criteria. Spurious actuation is a functional failure mode when a component performs a function without a real demand. Spurious actuations are of special interest for I&C due to complex effects via system dependences and due to a huge number of possible failure locations. There is a need to develop a reasonable but comprehensive approach both for deterministic and probabilistic analyses. Analysis requirements have been compiled, and a generic failure modes taxonomy and an analysis approach have been outlined. The software reliability task has been working on the confidence building in the method to estimate application software failure probability. The impact of pooling data from high and low demand systems is discussed. The principle of the probability estimation has been adjusted from the approach developed in the DIGREL project. A solution for the software complexity assessment has been prepared. I&C failure data is one of the information sources needed for the assessment of I&C reliability. Vendors have data sources as they typically have access to experience data from many plants, have needed insight on the software development processes and are capable to analyse the causes of the detected failures. International collaboration and discussions are still needed in order to forward the use of I&C failure data in PSA.
    Original languageEnglish
    Number of pages85
    Publication statusPublished - 2016
    MoE publication typeD4 Published development or research report or study

    Publication series

    SeriesNKS Reports
    NumberNKS-361

      Fingerprint

    Keywords

    • Digital I&C system
    • probabilistic safety assessment
    • reliability
    • nuclear power plant safety

    Cite this

    Authén, S., Bäckström, O., Holmberg, J-E., Porthin, M., & Tyrväinen, T. (2016). Modelling of DIgital I&C, MODIG: Interim report 2015. NKS Reports, No. NKS-361