Validation of thermal–hydraulic system code improved by advanced drift-flux correlation using bundle data at moderate pressure conditions

Abstract

The accurate prediction of the interfacial drag force is essential to predict the void fraction in a reactor core during the transient and accident phases of nuclear reactors. The key to accurate prediction is to provide reliable drift-flux parameters in calculating the interfacial drag force. The authors previously developed a drift-flux correlation to improve the interfacial drag force prediction accuracy under low-flow and low-pressure conditions. In the present study, the authors demonstrated the validity of the interpolation scheme at a moderate pressure range in the drift-flux correlation and quantified its uncertainty using the experimental data measured in a 5 × 5 rod bundle for the pressures of 0.25, 0.41, 2, 5, and 7 MPa at the multi-purpose Hitachi utility steam test leading facility (HUSTLE). The validated drift-flux correlation was implemented into the RELAP5/MOD3 code and numerical analyses were also performed to reproduce the HUSTLE experimental data. The numerical calculations using the modified RELAP5/MOD3 code showed improved prediction accuracy compared to the calculations by the original RELAP5/MOD3 with the Chexal-Lellouche correlation (e.g., EPRI correlation) in the tested pressure range. The uncertainty of the distribution parameter in the drift-flux correlation was estimated using the HUSTLE database. Next, a numerical analysis was done for the differential pressure in the reactor core in a small break LOCA integral test conducted with the ROSA/large scale test facility (LSTF) by the former Japan Atomic Energy Research Institute. This analysis showed that the predictions with the improved RELAP5/MOD3 agreed with the ROSA/LSTF data within the uncertainty at a 95 % confidence level. It was concluded that the improved RELAP5/MOD3 with the new drift-flux correlation could predict the water level of a reactor core during accident scenarios.