KAERI has been operating an integral effect test facility, Advanced Thermal–Hydraulic Test Loop for Accident Simulation (ATLAS), for accident simulations of advanced pressurized water reactors. As an integral effect test database for major design basis accidents has been accumulated, a domestic standard problem (DSP) exercise using ATLAS was proposed in order to transfer the database to domestic nuclear industries and to contribute to improving the safety analysis technology for pressurized water reactors (PWRs). As the third DSP exercise, a double-ended guillotine break of the main steam-line at an 8% power without loss of off-site power was decided as a target scenario. Seventeen domestic organizations joined this DSP exercise. They include universities, government, and nuclear industries. The participants of DSP-03 were classified into three groups and each group has focused on the specific subject related to the enhancement of the code assessment; (1) scaling capability of the ATLAS test data by comparing with the code analysis for a prototype, (2) multi-dimensional thermal–hydraulic phenomena anticipated during the steam-line break transient, (3) effect of various models in the one-dimensional safety analysis codes.

Loop seal clearing (LSC) is an important phenomenon for the safety of a pressurized water reactor (PWR) during a small-break loss-of-coolant accident (SBLOCA). The investigation on an LSC phenomenon of 4″, 6″, and 8.5″ break cold leg SBLOCAs simulated by Advanced thermal–hydraulic Test Loop for Accident Simulation (ATLAS) is performed using a Multi-dimensional Analysis of Reactor Safety-KINS Standard (MARS-KS) code. The LSC triggers earlier for larger break sizes during tests and calculations. LSCs occur during the simultaneous sudden decrease of steam condensation rate and the sudden increase of the break volumetric flow rate while the core volumetric flow rate increases slowly in calculation. The five phases of an SBLOCA transient are blowdown, pressure plateau, LSC, boil-off, and core-recovery phase, which can be identified by observing the volumetric flow rate and the time-dependent pressure variation. Loop seal refilling (LSR) occurs due to the trivial steam flow rate to the crossover leg inlet in calculation. The sensitivity analysis shows that the combination of countercurrent flow limitation (CCFL) model option for hot leg and steam generator (SG) inlet (Kutateladze, c = 1.36, m = 1), crossover legs (Kutateladze, c = 1, m = 1), and SG U-tubes (Wallis, c = 1, m = 1) provide good prediction of the LSC phenomenon and thermal-hydraulics behaviors in an SBLOCA transient by MARS-KS code calculation.