Investigation of the bias coming from spectrum corrections in the simulations of nucler reactor transients

This paper investigates the effect introduced by the deviation from criticality in the generation of two energy group macroscopic data (diffusion coefficients and cross-sections) typically used by core simulators. Such tools are based on data adjusted to mimic a critical system. This adjustment is usually performed using a B1 homogeneous method. This method is also used in the present paper to generate neutron energy spectra deviating from criticality, spectra thereafter utilized to condense macroscopic data to two energy groups. For deviations from criticality in the range of ±20xBeta_eff for a system representative of Pressurized Water Reactor conditions, it was found that the relative deviation from critical conditions amounts to nearly ±10% in the thermal group flux. In terms of condensed macroscopic data, the largest devia-tions are observed on the isotropic and anisotropic scattering cross-sections, with the largest deviations reaching ±4000 pcm. For the data used in two-group diffusion-based core simu-lators, the largest deviations occur for the removal cross-section, amounting to ±3000 pcm. The deviations from the criticality-adjusted macroscopic data are significant, and might have a large impact in reactor transients largely deviating from criticality. Criticality-adjusted material data should thus be used with care in such situations. This study also demonstrates the need in developing a methodology in cross-section preparation and core simulations to account for this effect