RELAP5/MOD3 simulation of the loss of residual heat removal during midloop operation experiment conducted at the ROSA-IV/ Large Scale Test Facility

Abstract

The modeling of the complex thermal hydraulics Of reactor systems involves the use Of experimental test systems as well as numerical codes. A simulation of the loss of residual heat removal (RHR) during midloop operations was performed using the RELAP5/MOD3 thermal hydraulic code. The experiment was conducted at the Rig of Safety Assessment (ROSA)-IV/ Large Scale Test Facility (LSTF). The experiment involved a 5% cold leg break along with the loss of the RHR system-The transient was simulated for 3040 seconds. The ROSA-1-V/]LsTF is one of the largest test facilities in the world and is located in Japan. It is a volumetrically scaled (1/48) full height, two loop model of a Westinghouse four loop pressurized water reactor (PWR). The facility consists of pressure vessel, two symmetric loops, a pressurizer and a full emergency core cooling system (ECCS) system. The transient was run on the CRAY-YMP supercomputer at Texas A&M university. Core boiling and primary pressurization followed the initiation of the transient. The time to core boiling was overpredicted. Almost all Primary parameters were predicted well until the occurrence of the loop seal clearing (LSC) at 2400 seconds. The secondary side temperatures were in good agreement with the experimental data until the LSC. Following the LSC, the steam condensation in the tubes was not calculated. This resulted in the overprediction of primary pressures after the LSC. Also, the temperatures in the hot and the cold legs were overpredicted. Because there was no significant condensation in the U-tubes, the core remained uncovered. Moreover, the LSC did not recover. Consequently, secondary side temperatures were underpredicted after the LSC. This indicated the deficiency of the condensation model. The core temperature excursion at the time of the LSC was not predicted, though there was good agreement between the experimental and calculated data for the rest of the transient. Severe oscillations were calculated throughout the course of the transient. Overall, there was reasonable qualitative agreement between the measured and the calculated data.