Search

Within the European SFR – Safety Measures Assessment and Research Tools (ESFR-SMART) project, steady-state coupled simulation of the ESFR core has been performed using several core analysis packages, with the objective of quantifying the coupling effect. Focus is on the fuel Doppler effect and coolant expansion effect. Standalone neutronics calculations in TRACE/PARCS (PSI), DYN3D (HZDR) and WIMS (Jacobs) showed superb agreement with the reference Serpent power distribution (root mean square (rms) discrepancies of 1.3%, 1.5% and 0.7% respectively). Results for COUNTHER (CIEMAT) were also in reasonable agreement, but with a somewhat higher discrepancy of 3.7%. Temperature distributions from thermal-hydraulic calculations were also compared and are found to be in good agreement. The effect of Doppler and coolant density feedback on core power distribution was predicted by TRACE/PARCS, DYN3D and WIMS to be between 0.4% and 0.8% rms difference in assembly powers. Reactivity coefficients for perturbations in the inlet temperature, flow rate and core power were shown to be negative for these three codes, with values of roughly -0.5 pcm/°C, -0.3 pcm/°C and -3.5 pcm/% respectively. A preliminary investigation of differential thermal expansion effects indicates that this may have a significant effect on core power distribution of a few %, greater than anticipated a priori and may warrant inclusion in coupled core analysis to ensure the accurate calculation of power distributions.

The paper presents a transient simulation phase of the new benchmark on a large sodium fast reactor core. This phase of the benchmark was devoted to the modelling of selected operational transients performed during start-up tests of French SFR Superphénix in order to evaluate its reactivity characteristics and core response to certain perturbations. Six operational transients were selected for the analysis. The specification of a simplified thermal hydraulic model equipped with point kinetics reactivity data and boundary conditions for the selected transients are given in the paper and the results of simulations with several system codes are reported. The study uses the results of the reference Serpent 2 Monte Carlo solution obtained during the first phase of the benchmark related to static neutronic characterization of the core. The developed model contains necessary thermal hydraulic description of primary system components and assumptions to account for thermal expansion reactivity feedbacks from in-reactor structures, neutron kinetics parameters, power distribution and reactivity coefficients. Thus the transient benchmark appears as an effective tool for validation and cross comparisons of system codes including approbation and comparison of different modelling features for thermal expansion of the in-reactor structures, giving a reference core behaviour with use of relatively simple models. The results of the modelling demonstrate a reasonable agreement between all solutions and with the experimentally measured evolution of the core parameters. Particular discrepancies with experimental data could not be resolved using the simplified benchmark model and available experimental data reconstructed from the published analysis of start-up tests. Because of that, the future steps for achieving the improved agreement between the simulated results and the experimental data were proposed.

This study presents an approach to the selection of optimal energy group structures for multi-group nodal diffusion analyses of Sodium-cooled Fast Reactor cores. The goal is to speed up calculations, particularly in transient calculations, while maintaining an acceptable accuracy of the results. In Part I of the paper, possible time-savings due to collapsing of energy groups are evaluated using 24-group energy structure as a reference. Afterwards, focusing on energy structures with a number of groups leading to significant calculation speedups, optimal grid configurations are identified. Depending on a number of possible energy grid configurations to explore, the optimization is conducted by either a direct search or applying the simulated annealing method. Speedup and optimization studies are performed on a selected case of the Superphénix static neutronic benchmark by using the nodal diffusion DYN3D code. The results demonstrate noticeable improvements in DYN3D performance with a marginal deterioration of the accuracy.

This study presents an approach to the selection of optimal energy group structures for multi-group nodal diffusion analyses of Sodium-cooled Fast Reactor cores. The goal is to speed up calculations, particularly in transient calculations, while maintaining an acceptable accuracy of the results. In Part I of the paper, possible time-savings due to collapsing of energy groups are evaluated using 24-group energy structure as a reference. Afterwards, focusing on energy structures with a number of groups leading to significant calculation speedups, optimal grid configurations are identified. Depending on a number of possible energy grid configurations to explore, the optimization is conducted by either a direct search or applying the simulated annealing method. Speedup and optimization studies are performed on a selected case of the Superphénix static neutronic benchmark by using the nodal diffusion DYN3D code. The results demonstrate noticeable improvements in DYN3D performance with a marginal deterioration of the accuracy.

ATHLET is a thermal-hydraulic (TH) system code developed at the GRS for the modeling of Light Water Reactors (LWRs). To extend the applicability of ATHLET to the analyses of Sodium Fast Reactors (SFRs), the code was recently upgraded with the thermal-physical properties of the liquid sodium. The new extension is still under verification and validation phases. The present work contributes to the verification efforts. This study investigated the perfor-mance of the extended version of ATHLET as applied to the transient analysis of a set of start-up tests conducted at the Superphénix SFR. The specifications of the corresponding tests such as the simplified SPX reactor core models and the set of reactivity coefficients were adopted primarily from a previous dedicated study performed at PSI and at KIT. The reactivity effects accounted for by ATHLET included fuel Doppler effect and thermal expansion effects of sodium, fuel, diagrid, control rods driveline, strongback, and reactor vessel. The results obtained by ATHLET for main stationary TH parameters, power evolutions, and reactivity feedback components were benchmarked against the reference solutions provided by TRACE. Employing an identical set of reactivity coefficients, either in steady-state or transient calculations, the codes produce consistent and close results.

The thermal-hydraulics (TH) code ATHLET has been upgraded to be capable of sodium flow modeling. Its new extension is under verification and validation phase. The presented study aimed to demonstrate ATHLET capability in Sodium-cooled Fast Reactor (SFR) transient predictions, through the comparison against TRACE TH code, this last being more established and tested for SFR applications. Calculations were performed on a set of start-up tests on Superphénix (SPX) SFR, and compared with TRACE results, which were used as a reference. It has been shown that given a specific set of reactivity coefficients, ATHLET and TRACE give consistent and close results.