Spatial analysis of unprotected transients in heterogeneous sodium fast reactors

Abstract

The main purpose of this benchmark paper is to study and compare point and spatial neutronic approaches used to calculate ULOF and UTOP transients in sodium cooled fast reactors. A second objective is to compare deterministic and Monte Carlo calculations with two different calculation codes. The first one is based on a deterministic (discrete ordinate S N) approach, using tabulated self-shielded cross sections, where the core reactivity and the power shape distribution are evaluated at each time step of the transient calculation. The second model relies on the Transient Fission Matrix (TFM) approach, condensing the response of a Monte Carlo neutronic code in time dependent Green functions characterizing the local transport in the reactor. This second approach allows a fast estimation of the reactivity and of the flux redistribution in the system during the transient with a precision closed to that of the Monte Carlo code. Both models have been coupled to the thermalhydraulics and applied on an ASTRID representative assembly. This application case is supposed to be sensitive to power redistributions. A second comparison between spatial kinetics and point kinetics calculations has been led to study this point. Finally we obtain a good agreement between spatial and point kinetics on ULOF and UTOP calculations, while some discrepancies are observed between the TFM and the S N approaches on the power level stabilization, due to difference on the feedback estimation in both models.

Uncertainty propagation for the design study of the PETALE experimental programme in the CROCUS reactor

“…In order to estimate the absorption rate inside the foils located in the metal reflector, a variance reduction is required in order to increase the number of thermal neutrons simulated in the metal reflector plates. This work has been performed with a modified version of Serpent2 code v2.1.21 used for previous studies [11] and where the correlated sampling technique has been implemented [12,13]. Even if some variance reduction methods have already been implemented in recent developments [14] of Serpent2, a specific variance reduction has been developed in this work dedicated to applications for detectors near to a reactor with specific treatments according to the neutron energy.…”