Steady-state analysis of ROSTOV-2 benchmark problem using SERPENT-2 and PARCS codes

Abstract

Recently, a new benchmark problem has been introduced by the OECD/NEA on the Rostov-2 reactor to address the verification and validation of novel High-Fidelity Multi-Physics tools. In line with the first phase of this benchmark, the current research was dedicated to the steady state reactor physics analysis, using “traditional multi-physics” codes for further and future comparisons. A set of homogenized two-group cross-section libraries were developed using SERPENT-2 Monte Carlo code. Some sophisticated techniques have been employed for detailed consideration of radial as well as axial reflectors, spacer grids and assembly discontinuity factor evaluation. Since PARCS code requires the group constants in PMAXS format, the generated libraries were re-formatted using the GenPMAXS interface. The PMAXS library verification, as well as the steady state analysis have been conducted in three operational states of the reactor using PARCS nodal diffusion code, independent SERPENT-2 Monte Carlo simulation and some available reference data. A set of criticality calculations have been conducted to evaluate the total reactivity worth of different control rod groups, the group #10 integral reactivity curve, fuel temperature coefficient and the critical boron concentration in Hot Zero Power (HZP) and Hot Full Power (HFP) states. The maximum reactivity difference between PARCS and SERPENT-2 simulations was less than 100 pcm in all criticality calculations. In HFP state, the relative power distribution in radial and axial directions were calculated and compared. The maximum Power Peaking Factors (PPFs) and the location of the “hot assembly” from PARCS and SERPENT-2 are in good agreement with each other. Finally, the burn-up calculation was conducted with PARCS code to evaluate neutronic parameters at some specific burn-up values. The calculated reactor physics parameters including the critical boron concentration, the power peaking factors, the power and burn-up distributions at the Beginning of Experiment (BOE) were consistent with available reference data.