Uncertainty quantification in modeling and simulation is an important step in increasing confidence in the safety analysis of a nuclear power plant. The goal of this research was to improve the multiphysics uncertainty quantification capabilities of RELAP5-3D, ultimately decreasing over-conservatisms during safety margins assessment.
The research improved uncertainty analysis capability in RELAP5-3D, the INL best-estimate nuclear thermal-hydraulic computer code, focusing on the reflooding phase of large break loss-of-coolant accident. Relevant closure laws of RELAP5-3D were modified to allow codes user to perturb their main parameters using assigned probabilities distribution functions. Idaho National Laboratory's risk-analysis tool RAVEN, coupled with RELAP5-3D code, were used to demonstrate phase of the research project, and applied to the analysis of two reflood experiments. The methodology was widely used in other projects in the Risk-Informed Systems Analysis pathway.
For more information, contact