Advanced BWR criticality safety part I: Model development, model benchmarking, and depletion with uncertainty analysis

Abstract

Listen

Translate article

Due to the increased amounts of BWR spent fuel that will need to be transported and stored, additional research is needed in the area of BWR criticality safety and burnup credit. This study performs an advanced depletion analysis for a BWR lattice capturing various BWR operating complexities simultaneously. This is done to address the compounding effects of these complexities on lattice reactivity, k∞ uncertainty, and isotopic inventory. A set of BWR lattice models is developed with evolving complexity, these complexities include presence of a gadolinium absorber, control rod modeling, variable radial enrichment, a nonuniform axial burnup profile, non-uniform axial coolant density, control rod partial insertion, variable axial enrichment, part-length rods, and control rod movement during operation. These models have been developed rigorously and benchmarked using different codes to ensure modeling accuracy. The lattice configuration, coolant density profile, control blade history, and other operating data were based on real-world data collected from literature. It was found that averaging radial enrichment had minimal impact on the k∞ value and reactivity peak location. The effect of axial burnup and coolant density profiles was significant on the time of peak reactivity, making the lattice reaching the peak reactivity earlier in cycle. The 3D models show slower U-235 and Gd-155 depletion compared to the 2D cases, making the 3D lattices less reactive in general. The k∞ uncertainty for the studied design is driven by the uncertainties of U-235, U-238, Gd-155, and Gd-157 at beginning of cycle, these are replaced later by plutonium isotopes after depletion. The effect of variable axial enrichment and part-length rods showed a significant impact on U-235 and Gd-155 depletion, making those design complexities important for criticality safety considerations. It was found that BWR modeling require many complexities which make the depletion calculations very expensive even for a single lattice. Furthermore, this adds more difficulty on the brute-force sampling-based uncertainty analysis.