3-D Deterministic Burnup With High Performance Computing

Current 3-D reactor burnup simulation codes typically utilize either transport-corrected diffusion theory or Monte Carlo methods to perform flux calculations necessary for fuel depletion. Monte Carlo codes, particularly the Monte Carlo N-particle Transport Code (MCNP) from Los Alamos, have become increasingly popular with the growth of parallel computing. While achieving a criticality eigenvalue is relatively straight forward, run times for large models requiring converged fission sources from proper burnup computation quickly becomes very time consuming. Additionally, past analyses have shown difficulties in source convergence for lattice problems using Monte Carlo [1]. To invoke an alternative means of computing core burnup and decreasing computation time for large models, a deterministic tool such as the PENTRAN/PENBURN suite is necessary. PENTRAN is a multi-group, anisotropic Sn code for 3-D Cartesian geometries; it has been specifically designed for distributed memory, scalable parallel computer architectures using the MPI (Message Passing Interface) library. Automatic domain decomposition among the angular, energy, and spatial variables with an adaptive differencing algorithm and other numerical enhancements make PENTRAN an extremely robust solver with a 0.975 parallel code fraction (based on Amdahl’s law). PENBURN (Parallel Environment BURNup), a recently developed fuel depletion solver, works in conjunction with PENTRAN and performs 3-D zone based fuel burnup using the direct Bateman chain solution method. The aim of this paper is to demonstrate the capabilities and unique features of the PENTRAN/PENBURN suite through a fuel burnup study on a 3 wt% enriched UO₂ fuel pin and 17×17 Westinghouse OFA assembly.