A DISCUSSION OF UNCERTAINTY QUANTIFICATION FOR CHF AS PERFORMED IN COBRA-IE

摘要

An assessment of the predictive capability of COBRA-IE for Critical Heat Flux (CHF) using the 2005 Groeneveld CHF Look-Up Table is presented. The assessment was performed against 13 different open literature CHF experiments which were conducted over a wide range of conditions in various internal flow geometries. Overall, approximately 1300 data points were evaluated. Different methodologies to quantify the uncertainty inherent in the CHF models are discussed in this paper. The simulation techniques, uncertainty methods, and the results of two of the methods are provided. A discussion of the appropriate use of the various CHF uncertainty methods is included. The results indicate that for the method associated with the largest uncertainty, the average Measured/Predicted value in CHF is 1.19 and the standard deviation is 0.62. For the second method, similar to the critical power ratio used for boiling water reactors, the average ratio is 0.98 and the standard deviation is 0.13. Finally, a method to translate between the various methods is proposed and shown to be accurate. The use of this transformation could permit significant time and cost savings by allowing a single uncertainty assessment to serve two very different analytical needs.