SBIR-STTR Award

Changes in plant status, either before or during an accident sequence can considerably affect its outcome; both the actual outcome and that predicted by probabilistic risk assessment. In order to accurately represent the possible accidents and their probabilities, the PRA model must consider the dynamic character of the plant, it's systems and operators. A living, dynamic PRA is one in which all types of changes must be periodically accounted for. For instance, the effects of aging and plant configuration changes must be periodically accounted for by updating the PRA data and model. Existing PRAs will be examined to identify scenarios in which the use of dynamic PRA methods could make a significant difference in the resulting core damage frequency. Conventional results for these scenarios will be compared to those obtained using selected representative dynamic PRA methods as they exist now. One combination of methods may be chosen on it's relative merits compared to conventional methods. The chosen combination will be examined to determine improvements needed to make it applicable to situations which require dynamic analysis. These improvements will be implemented during Phase II where a prototype dynamic PRA computer program will be developed and tested with representative dynamic scenarios.

Anticipated Results:
Phase I will identify classes of situations where dynamic effects need to be modeled in PRAs. In addition, existing dynamic methodologies will be reviewed and their applicability to these situations assessed. Phase I will also define the overall characteristics of a comprehensive dynamic methodology. Phase II will focus on methodological improvements and on development and testing of a prototype computer code for comprehensive dynamic PRA analysis. It is anticipated that positive results from Phase II will motivate federal agencies and the commercial sector particularly the nuclear, chemical process, and defense industries to fund the full scale development of the computer code in Phase lil.