Heat Transfer in a CANDU-Type Fuel Bundle During a LOCA Experiment

Abstract

Pre- and post-test simulations with the thermalhydraulics code CATHENA [1] (Canadian Algorithm for THErmalhydraulic Network Analysis) MOD-3.5c/Rev O of a LOCA (Loss Qf Coolant Accident) experiment in the RD-14M test facility at AECL, Whiteshell Laboratories, Canada were performed. The simulations were performed to assess the heat transfer calculations by CATHENA in a CANDU (CANada Deuterium Uranium) type fuel bundle under conditions expected in the unlikely event of a LOCA. The experiment was a 15-mm diameter inlet-header break with a primary-pump exponential ramp-down, and no emergency-coolant injection (ECI). Two channels in the RD-14M loop were used, one per pass, with a simulated power pulse in the channel of the broken pass. In the pre-test simulation results, average maximum sheath temperatures were 15% higher than measured temperatures. Analysis of the post-test simulation results identified uncertainties in the heat transfer correlations, flow regime, and discharge conditions as the most important factors contributing to the discrepancy between measured and simulation results. These uncertainties and their impact on the calculated fuel sheath temperatures are presented and discussed in detail in this paper.