Considering the uncertainties in empirical correlations for vertical countercurrent flow limitation (CCFL) with TRACE

Abstract

Abstract TRACE is used to calculate the thermal-hydraulic sequence in nuclear power plants for accident analysis. In some postulated accidents, countercurrent flow limitation (CCFL) phenomena can occur. This phenomenon is calculated by means of empirical relationships in TRACE. Usually, not all empirical relationships used in TRACE are accessible to the user who develops the computational model for accident analysis, but in the case of CCFL, the user must specify the parameters of the empirical relationship. Additional preliminary work is needed for fitting the parameters so that experimental data matches the computational calculations and the accident analysis can be performed. Furthermore, uncertainties in measurement errors from experimental data lead to user-based variations in the parameters of the empirical relationship for CCFL. Therefore, an alternative procedure in which accident analysis is performed with consideration of the uncertainties in the empirical relationships is presented. The uncertainty was quantified by means of a stochastic linear regression model for the Wallis correlation and the Bankoff correlation. An uncertainty analysis was subsequently performed to demonstrate the influence of uncertainty in certain cases of CCFL.