Verification and Validation of the EUCLID/V2 Integrated Code’s HEFEST-FR Module

Abstract

For substantiating liquid metal cooled reactor plants, the EUCLID/V2 integrated code is being developed, and its verification and validation are carried out for certifying it at the Scientific and Technical Center for Nuclear and Radiation Safety (NTC NRB). One of the integrated code’s main parts is the severe accident block, which includes the SAFR module for calculating the destruction of fuel pins, fuel assemblies (FAs) and the entire core, as well as the HEFEST-FR module for calculating the melt retention and cooling down in the sodium-cooled reactor core catcher. The HEFEST-FR module implements the possibility to perform 2D simulation of the structural elements and fuel melt behavior in liquid metal cooled reactors. In accordance with the NTC NRB requirements, for the HEFEST-FR module to be used as part of the EUCLID/V2 code for analyzing the safety of fast reactors, it must be validated with the use of available experimental data; the validation shall be accompanied with an uncertainty and sensitivity analysis and assessment of the calculation result error. The article presents the results obtained from verification of the EUCLID/V2 integrated code HEFEST-FR module through solving the analytical problem of settling a stationary temperature of a homogeneous bounded cylinder uniformly heated from below with boundary conditions of the third kind and through solving the Stefan single-phase problem, as well as the results of validating the HEFEST-FR module based on the SCARABEE BF1 experiment. It is shown that the average absolute value by which the numerical calculation deviates from the analytical solution of the problem of settling a stationary temperature of a homogeneous bounded cylinder uniformly heated from below with boundary conditions of the third kind makes approximately 1.1 K. The maximum relative deviation of the results of calculations carried out using the computer program from the results of analytical solution of the Stefan problem (determination of the melt front) makes 0.46%. An assessment of the errors of modeling using the melt retention module as part of the EUCLID/V2 code (HEFEST-FR) based on the BF1 test of the SCARABEE experiment has shown that the temperature calculation error lies in the interval [‒82.3; 182.5] K, and the error of calculating the radial heat flux lies in the interval [‒55.2; 31.2] kW/m².