Validation of RELAP5/Mod3.3 Against the PACTEL SBL-50 Benchmark Transient

Paper in proceeding, 2013

ABSTRACT The PWR PACTEL test facility has recently been designed to support the safety studies of EPR type nuclear reactor thermal-hydraulics. The facility is located at the Lappeenranta University of Technology (LUT) in Finland. It is essentially important to understand the system behavior under natural circulation conditions during a Loss of Coolant Accident (LOCA). With this objective in mind, an international benchmark transient was conducted at the LUT in 2010-2011. The SBL-50 test was a SB-LOCA with a 1 mm break in the cold leg. The continuous inventory loss led to core dry-out. This project gave unique opportunities for several organizations to build and validate their models for the PWR PACTEL, as well as to simulate the transient by using various computer codes. Chalmers University of Technology participated with RELAP5/Mod3.3 calculations both in the pre-test and post-test phases of the project. The pre-test simulation included a simplified steam generator model, with the description of the heat exchange by a single characteristic U-tube. This coarse nodalization resulted in reasonably good agreement with the measured data. As the test data became known in the post-test, minor modifications contributed to achievement of better results. The changes were related to the upper plenum nodalization and the critical discharge flow parameters at the break assembly. Even if the post-test model provided better agreement for most of the parameters, it still had difficulties to predict the temperatures in the longest tubes in the steam generators (SGs). A careful examination of the measured data indicated that discrepancies might originate from a flow reversal in the SG primary side. Thus, an advanced SG model was prepared with application of a multi-channel system. Altogether 51 heat exchanger tubes are arranged into 5 bundles in the PWR PACTEL SG, according to 5 different lengths. These bundles were individually modeled in the refined input. The most recent multi-channel SG model has not only confirmed the presence of reverse flow and internal circulation in the SG primary, but it has also contributed to a much better prediction of the fluid temperature distribution.