Improving the Resolution and Computational Efficiency of Depletion Simulations on a TRISO-Fueled Microreactor

Abstract

Modeling individual TRi-structural ISOtropic (TRISO) particles is possible using constructive solid geometry in Monte Carlo neutron transport codes (such as MCNP); however, the billions of surfaces required incurs considerable computational expense. Depletion simulations, in conjunction with Monte Carlo transport, are also inherently slow, requiring many individual criticality calculations to approximate time-dependent behavior. Additionally, choosing acceptable time step sizes, spatial zoning, and other depletion parameters involves prior knowledge and/or iteration by the user. The slow and complex nature of depletion calculations greatly increases the cost and duration of the design iteration process. The importance of tracked isotopes, spatial zones, and time stepping during depletion simulations is evaluated globally using neutron multiplication factor, total isotope inventory, and burnup. Different depletion schemes are compared using the TRISO-fueled Snowflake microreactor. Woodcock Delta Tracking and a Reactivity-equivalent Physical Transformation (RPT) are investigated as potential approaches to increase the speed of Monte Carlo neutron transport calculations involving TRISO fuel. The Monte Carlo N-Particle® Code (MCNP®) Version 6.3 is used to obtain the transport solution, and depletion calculations are performed using its internally coupled version of CINDER90. A prototype MCNP delta tracking module under development at Los Alamos National Laboratory is also used. Although a single spatial depletion region is sometimes acceptable, large errors in core lifetime and total 238Pu, 241Pu, and 242Pu mass are possible. The prototype delta tracking module significantly speeds up criticality calculations; however, depletion calculations are not always faster than those performed with surface tracking. The RPT method is effective at reducing CPU times in both criticality and depletion calculations, without significantly impacting neutron multiplication factor and total isotope inventory. The necessary number of time steps to converge total isotope inventory was also explored; the required number and spacing varied greatly depending on the objective of the depletion simulation.