Comparison of critical flow models for SBLOCA tests of the ATLAS facility

Abstract

A comparison of critical flow models between the Trapp–Ransom and Henry–Fauske models for all SBLOCA (small break loss of coolant accident) scenarios of the ATLAS (advanced thermal–hydraulic test loop for accident simulation) facility was performed using the MARS-KS code. For the comparison of the two critical models, the accumulated break mass was selected as the main parameter for the comparison between the calculations and tests. Four cases showed the same respective discharge coefficients between the two critical models, e.g., 6″ CL (cold leg) break and 25%, 50%, and 100% DVI (direct vessel injection) breaks. In the case of the 4″ CL break, no reasonable results were obtained with any possible Cd values. In addition, typical system behaviors, e.g., PZR (pressurizer) pressure and collapsed core water level, were also compared between the two critical models. From the comparison between the two critical models for CL breaks, the Trapp–Ransom model predicted quite well with respect to the other model for the smallest and larger breaks, e.g., 2″, 6″, and 8.5″ CL breaks. In addition, from the comparison between the two critical models for DVI breaks, the Trapp–Ransom model predicted quite well with respect to the other model for the smallest and larger breaks, e.g., 5%, 50%, and 100% DVI breaks. In the case of 50% and 100% breaks, the two critical models predicted the test data quite well.