BEPU analysis of a CANDU LBLOCA RD-14M experiment using RELAP/SCDAPSIM

A key element of the safety analysis is Loss of Coolant Analysis (LOCA) which must be performed using system thermal-hydraulic codes. These codes are extensively validated against separate effect and integral experiments. RELAP/SCDAPSIM is one such code that may be used to predict LBLOCA response in a CANDU reactor. The RD-14M experiment selected for the Best Estimate Plus Uncertainty study is a 44 mm (22.7%) inlet header break test with no Emergency Coolant Injection. This work has two objectives first is to simulate pipe break with RELAP and compare these results to those available from experiment and from comparable TRACE calculations. The second objective is to quantify uncertainty in the fuel element sheath (FES) temperature arising from model coefficient as well as input parameter uncertainties using Integrated Uncertainty Analysis package. RELAP calculated results are found to be in good agreement with those of TRACE and with those of experiments. The base case maximum FES temperature is 335.5 °C while that of 95% confidence 95th percentile is 407.41 °C for the first order Wilk's formula. The experimental measurements fall within the predicted band and the trends and sensitivities are similar to those reported for the TRACE code. •A large break loss of coolant accident is simulated using RELAP/SCDAPSIM in RD-14M experiment.•The model is compared with a channel voiding test B0106 for a 44 mm break with no emergency coolant injection.•RELAP/SCDAPSIM predictions are compared to the publicly available TRACE results.•Uncertainty in the temperature is quantified.•Both input parameters and the source code model coefficients are considered for perturbation.•Using the Wilks criterion, the 95/95 confidence interval on peak temperature lies within 100 °C of the best-estimate case.