Evaluation of the Sub-Channel Code COBRA-TF for Prediction of BWR Fuel Assembly Void Fraction Distribution

Abstract

Good quality experimental data is needed to refine the thermal hydraulic models for the prediction of rod bundle void distribution and critical heat flux (CHF) or dry-out. The Nuclear Power Engineering Corporation (NUPEC) has provided a valuable database to evaluate the thermal hydraulic codes [1]. Part of this database was selected for the NUPEC BWR Full-size Fine-Mesh Bundle Tests (BFBT) benchmark sponsored by US NRC, METI-Japan, NEA/OECD and Nuclear Engineering Program of the Pennsylvania State University (PSU). Twenty-five organizations from ten countries have confirmed their intention to participate and will provide code predictions to be compared to the measured data for a series of defined exercises within the framework of the BFBT benchmark. This benchmark data includes both the fine-mesh high quality sub-channel void fraction and critical power data. Using a full BWR rod bundle test facility, the void distribution was measured at mesh sizes smaller than the sub-channel by using a state-of-the-art computer tomography (CT) technology [1]. Experiments were performed for different pressures, flow rates, exit qualities, inlet sub-cooling, power distributions, spacer types and assembly designs. There are microscopic and sub-channel averaged void fraction data from the CT scanner at the bundle exit as well as X-ray densitometer void distribution data at different elevation levels in the rod bundle. Each sub-channel’s loss coefficient was calculated with using the Rehme method [2,3], and a COBRA-TF sub-channel model was developed for the NUPEC facility. The BWR assembly that was modeled with COBRA-TF includes two water rods at the center. The predicted sub-channel void fraction values from COBRA-TF are compared with the bundle exit void fraction values measured using the CT-scanner void fraction from the BFBT benchmark data. Different plots are used to examine the code prediction of the void distribution at a sub-channel level for the different sub-channels within the bundle.