Proceedings of the Eighth International Conference on Probabilistic Safety Assessment & Management (PSAM)
264 Next Generation Probabilistic Risk Assessment (PRA) Tools and Methods (PSAM-0393)
Download citation file:
- Ris (Zotero)
- Reference Manager
This paper presents the Electric Power Research Institute (EPRI) strategic vision for advanced, integrated risk assessment and management tools for current and future nuclear power plants (NPPs). It also describes research and development (R&D) activities underway at EPRI to support this vision.
The fault tree / event tree approach to developing and solving Probabilistic Risk Assessment (PRA) models was developed in the 1970's. The methods and the tools (e.g., computer software) gave shape to the current modeling efforts. Modern applications of risk technology have produced issues that are not easily solved using these traditional modeling techniques and the associated tools. Some of these issues include time phased modeling, multiple models (e.g., internal events, seismic and fires), incorporation of recoveries, dependent human actions, component aggregation and others. While modifications to the methods and tools, such as rule-based methods for incorporation of recoveries, have ameliorated some of these problems, each modification has some level of issue or difficulty with incorporation. In addition, the potential for the construction of new plants as well as the development of the next generation of reactors both of which need to be risk-informed, provide additional impetus to develop new methods and tools.
Any solution must be “evolutionary” as opposed to “revolutionary”. This is a result of the simple fact that the majority of the first users of the new technology will be the current set of operating reactors. Significant changes in the methodology or tools would result in an increase in the short-term resources, which are currently not available. They are many secondary issues such as the need to train new personnel, investments in existing technology, and others.
The solution must also be holistic in its approach. One significant issue associated with the current risk technology is its fragmentation. The fragmentation issue is evidenced in the maintenance of separate model structures for various hazards as well as maintenance of documentation that is separate from the model.
In the case of the next generation of reactors, the new methodology and tools must be able to quickly generate results of competing alternatives such that a risk-informed design can be achieved. This need requires almost a simulation type environment where designs, reliability, and human performance are the variables than can be manipulated and the results compared.
Several emerging technologies are in development at EPRI for the next generation of risk technology methods and tools including: (1) Declarative modeling which allows attributes associated with modeling elements to be specified and used to alter model structure, change event probabilities or assess recovery and dependence; (2) Binary Decision Diagram approach to quantification which eliminates the rare event approximation and other simplifications; (3) Object oriented user interface which allows for building of PRA structures based on drawings or templates; (4) Documentation assistance tools which provide a method for documenting the analysis with a seamless interface between the model and documentation.