Proceedings of the Eighth International Conference on Probabilistic Safety Assessment & Management (PSAM)
Information Technology: Challenges in Risk and Reliability Modeling of Digital Systems
Download citation file:
- Ris (Zotero)
- Reference Manager
This paper discusses the conceptual framework and the project progress for a benchmarking task defined under the framework of a collaborative agreement between the U.S. Nuclear Regulatory Commission (NRC) and the Ohio State University (OSU). The objective of this task is to develop both policies and methods for inclusion of reliability models for digital systems into the current generation nuclear power plant Probabilistic Risk Assessments (PRAs), thus addressing the need to include the digital system contribution to the overall risk within a plant PRA framework.
This pilot study expands on the deductive analysis and inductive analysis features of the Dynamic Flowgraph Methodology (DFM) demonstrated in the past projects, and augment them with a software conditional risk approach and a quantification scheme to:
- Estimate the risk contribution of the digital system, and
- Integrate the results into the plant PRA framework.
DFM is an analytical process that was initially demonstrated in pilot Nuclear Regulatory Commission and NASA applications. It combines multi-valued logic modeling/analysis capabilities and time transition modeling that specifically address systems with non-coherent and dynamic features. The deductive analysis module of DFM explores the causality and the temporal behavior of the system model in reverse and generates prime implicants that can be thought of as multi-valued logic equivalent of minimal cut sets across time transitions. On the other hand, the inductive module of DFM follows the causality and the temporal behavior of the system model, and it is capable of producing automated Failure Modes and Effects Analyses (FMEA).