Advanced method for neutronic simulation of control rods in sodium fast reactors: Numerical and experimental validation

• DOI: 10.1016/j.anucene.2019.01.042
• Publication Date: 2019-02-10T00:27:12Z
• Authors: H. Guo; E. Garcia; B. Faure; L. Buiron; P. Archier; P. Sciora; G. Rimpault
• Keywords: [PHYS.NUCL]Physics [physics]/Nuclear Theory [nucl-th]; control rods; Neutronics; 0202 electrical engineering, electronic engineering, information engineering; fast reactor; 02 engineering and technology; [PHYS.NEXP]Physics [physics]/Nuclear Experiment [nucl-ex]; 7. Clean energy

Abstract High accuracy neutronic simulations are required for the development of Generation-IV sodium cooled fast reactors (SFR). This paper therefore focuses on the development and validation of an improved calculation method in APOLLO3® deterministic code to predict the reactivity control system in SFR. This method relies on the “lattice-core” paradigm and the generation of few group effective cross-sections. Those are computed over small lattice geometries with detailed descriptions of structures but no heterogeneity correction method. Two models are considered at the core level: the first one relies on homogeneous descriptions for all assemblies (fuel, control rods, reflector, diluent, etc.), whereas the second one takes into account the heterogeneity of absorber pins (semi-heterogeneous description). In order to assess the robustness of calculation methods, a validation work is achieved for three distinct SFR cores. A numerical validation is first performed for the SFR-3600-MOX core taken from the international WPRS benchmark. Then an experimental validation work is conducted for the control rods measurements from SUPERPHENIX start-up experiments. The last case of application is the ASTRID CFV core. At the cross-sections generation stage, the results show a good coherence with reference Monte Carlo calculations for both spatial and energy distributions. At the core level, it is shown that the homogeneous description of the control rods induces an overestimation of the reactivity worth while the semi-heterogeneous model significantly improves the results.