Numerical assessment of LOCA experiments of RD-14M using MARS-KS code

Abstract

The code assessment on the design basis events has been of fundamental importance to the safety analyses of the nuclear power plants. When a system analysis code is released to the new version by improving the existing models and correcting the found errors, it is essential to make sure that the updated code can predict and simulate reasonably the important phenomena possibly observed during the transient. This work has focused on strengthening the predictability and reliability of the system analysis code by benchmarking the invaluable loss-of-coolant-accident (LOCA) data obtained from an integral effect test facility (RD-14M). In this study, the verification tests of the inlet-header break simulating the LOCA of the CANDU-6 plants were selected as a benchmark case for MARS-KS code evaluation. First, a proper analysis model to benchmark the RD-14M experiment was envisioned into the MARS-KS input model and its steady-state conditions were achieved successfully. Subsequently, through the transient analysis experimental data of void fraction, header pressure, differential pressure between headers, and sheath temperature were compared with the code results. The code input models predicted the experimental results of the transient conditions with a good agreement. Through this study, it is expected that the MARS-KS code validation for the CANDU safety analysis could be improved significantly. In addition, this study is expected to contribute to developing practical assessment problems and providing a meaningful example for direction to code improvement.

Keywords

• PHWR
• CANDU-6
• LOCA
• MARS-KS
• RD-14M