Impact of radial void fraction distribution on boiling water reactor lattice physics calculations: Application to AREVA's next generation BWR fuel assembly, the ATRIUM™ 11 design

In current Boiling Water Reactor (BWR) methodologies for neutron cross sections generation, a homogeneous mixture of steam and liquid water is assumed to exist. However in reality, strong gradients in the vapor-liquid distribution in both axial and radial direction exist, which create a non-uniform water (coolant/moderator) density distribution inside the BWR fuel assembly (channel). Variations in sub-channel water densities are reinforced by variations in pin powers, themselves due to uranium enrichment, gadolinium-bearing rods, partial-length rods, and control blades. In this study, the assumption of using an assembly-uniform water density for a coupled neutronics/thermal-hydraulic simulation is evaluated on AREVA's next generation BWR assembly, the ATRIUM™ 11. The effects of using a non-uniform sub-channel moderator density feedback versus using a radially uniform moderator density feedback on lattice infinite multiplication factor (k-infinity) and pin powers are discussed for different scenarios, including changes in gadolinium concentration, depletion effects, and presence of control blades. Without the control blade inserted and after the gadolinium is depleted, the difference in k-infinity are within 100 pcm and the differences in the radial peaking factor and the maximum linear heat generation rate (LHGR) are within 0.5%. The sub-channel water densities cause the gadolinium and control blade worth to increase. The maximum decrease in k-infinity occurs at the beginning of cycle (BOC) by 513 pcm due to high gadolinium concentration and by 806 pcm due to both high gadolinium concentration and the control blade. The maximum increase in k-infinity occurs when the gadolinium is depleted by 650 pcm due to high gadolinium and by 219 pcm due to high gadolinium and the control blade.