Validation of Fire Models Applied to Nuclear Power Plant Safety

Abstract

The paper highlights key components of a fire model validation study conducted by the U.S. Nuclear Regulatory Commission and the Electric Power Research Institute. These include the selection of fire phenomena of interest to nuclear power plant safety, the selection of appropriate models, the selection of relevant experimental data, and the selection of appropriate evaluation criteria. For each model and each quantity of interest, there are two metrics of accuracy. The first is a bias factor, which indicates the extent to which the model tends to over or under-predict the given quantity. The second is a relative standard deviation, which indicates the degree of scatter in the predicted quantity when compared with experimental measurements. While the study is motivated by nuclear power plant safety, the general procedure and results are appropriate for most industrial applications.