Emergency Core Cooling System (ECCS) analyses using Loss of Coolant Accident (LOCA) codes utilize two-phase Reactor Coolant Pump (RCP) performance models formulated on the basis of data from tests conducted on the Semi-scale pump (Reference 1) operating at 60 Hz frequency. In some PWRs, the RCPs operate at a frequency of 50 Hz. This paper presents the results of an evaluation performed to determine the applicability of RCP two-phase performance models developed on the basis of data from the Semi-scale tests for analyzing ECCS performance of new generation PWRs. The evaluation addressed two major issues: (1) the applicability of the two-phase RCP performance model developed using the data from the Semi-scale pump tests (Reference 1) for full scale Pressurized Water Reactor (PWR) LOCA simulations, and (2) the relevance of the two-phase RCP performance model developed on the basis of test data for the Semi-scale pumps running at 60 Hz frequency to PWR RCPs running at 50 Hz frequency with higher specific speeds.
Reviews of pump performance test data available in the open literature identified two-phase performance data appropriate for use in substantiating the validity of current PWR pump performance models. These data supported the conclusion that the two-phase head performance degradation for the Semi-scale Mod-1 pump is conservative compared to the two-phase pump performance data generated from testing of pumps representative of full scale PWR RCPs. A review of ECCS analyses results available in the literature determined that the use of the current RCP two-phase performance model (developed using the Semi-scale Mod-1 pump test data) for a typical PWR plant resulted in about a 100 °F increase in the Peak Clad Temperature (PCT) for a Large Break LOCA (LBLOCA) in comparison to the PCTs calculated using the two-phase pump performance model developed on the basis of test data for pumps representative of full scale PWR RCPs.
It was determined from the current study that the frequency (50 Hz vs. 60 Hz) of the electrical power that drives the pump motor is not of much consequence for two-phase RCP performance modeling, since (1) the RCP performance model is characterized via normalized pump performance parameters, and (2) for the LBLOCA analysis of interest, the RCPs are assumed to lose power at the start of the event.