A comparative assessment of independent thermal-hydraulic models for research reactors: The RSG-GAS case

• Increased use of thermal-hydraulic codes requires assessment of important phenomena in RRs.
• Three independent modeling teams performed analysis of loss of flow transient.
• Purpose of this work is to examine the thermal-hydraulic codes response.
• To perform benchmark analysis comparing the different codes with experimental measurements.
• To identify the impact of the user effect on the computed results, performed with the same codes.

This study presents the comparative assessment of three thermal-hydraulic codes employed to model the Indonesian research reactor (RSG-GAS) and simulate the reactor behavior under steady state and loss of flow transient (LOFT). The RELAP5/MOD3, MERSAT and PARET-ANL thermal-hydraulic codes are used by independent research groups to perform benchmark analysis against measurements of coolant and clad temperatures, conducted on an instrumented fuel element inside RSG-GAS core. The results obtained confirm the applicability of RELAP5/MOD3, MERSAT and PARET-ANL on the modeling of loss of flow transient in research reactors. In particular, the three codes are able to simulate flow reversal from downward forced to upward natural convection after pump trip and successful reactor scram. The benchmark results show that the codes predict maximum clad temperature of hot channel conservatively with a maximum overestimation of 27% for RELAP5/MOD3, 17% for MERSAT and 8% for PARET-ANL. As an additional effort, the impact of user effect on the simulation results has been assessed for the code RELAP5/MOD3, where the main differences among the models are presented and discussed.