Markov/CCMT modeling of the benchmark system and incorporation of the results into an existing PRA

D. Mandelli, T. Aldemir, P. Bucci, L. A. Mangan, J. Kirschenbaum, M. Stovsky, E. Ekici, S. A. Arndt

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In the reliability modeling of digital control systems, conventional methodologies based on the event-tree/fault-tree (ET/FT) methodology may not represent adequately the statistical dependence between failure events in certain digital systems. Dynamic methodologies can be considered as an important alternative to overcome this limitation. The Markov/CCMT (Cell-to-Cell Mapping Technique) has been proposed as a dynamic methodology for the probabilistic risk assessment (PRA) of digital control systems. The application of Markov/CCMT is illustrated using the digital control system for a feedwater system of a pressurized water reactor (PWR). Discrete hardware/software/firmware states are defined and transitions between these states are deduced from the control logic of the system, as well as from the failure modes and effects analysis (FMEA) performed on each component. Markov/CCMT is used to represent the dynamics of the system as the probability of transitions between process variable magnitude intervals (cells) that partition the state space. The resulting Markov model is converted into dynamic event trees which then are incorporated into an existing ET/FT based PRA of a PWR using the SAPHIRE code.

Original languageEnglish
Title of host publication9th International Conference on Probabilistic Safety Assessment and Management 2008, PSAM 2008
Pages94-103
Number of pages10
StatePublished - 2008
Externally publishedYes
Event9th International Conference on Probabilistic Safety Assessment and Management 2008, PSAM 2008 - Hong Kong, China
Duration: May 18 2008May 23 2008

Publication series

Name9th International Conference on Probabilistic Safety Assessment and Management 2008, PSAM 2008
Volume1

Conference

Conference9th International Conference on Probabilistic Safety Assessment and Management 2008, PSAM 2008
Country/TerritoryChina
CityHong Kong
Period05/18/0805/23/08

Keywords

  • Digital systems
  • Dynamic PRA
  • Markov
  • Probabilistic risk assessment

Fingerprint

Dive into the research topics of 'Markov/CCMT modeling of the benchmark system and incorporation of the results into an existing PRA'. Together they form a unique fingerprint.

Cite this