Title: Challenging Fuel Cycle Modeling Assumptions: Facility and Time Step Discretization Effects

Abstract: Due to the diversity of fuel cycle simulator modeling assumptions, direct comparison and benchmarking can be difficult. In 2012 the Organisation for Economic Co-operation and Development completed a benchmark study that is perhaps the most complete published comparison performed. Despite this, various results from the simulators were often significantly different because of inconsistencies in modeling decisions involving reprocessing strategies, refueling behavior, reactor end-of-life handling, etc. This work identifies and quantifies the effects of selected modeling choices that may sometimes be taken for granted in the fuel cycle simulation domain. Four scenarios are compared using combinations of fleet-based or individually modeled reactors with monthly or quarterly (3-month) time steps. The scenarios approximate a transition from the current U.S. once-through light water reactor fleet to a full sodium fast reactor fuel cycle. The Cyclus fuel cycle simulator's plug-in capability along with its market-like dynamic material routing allow it to be used as a level playing field for comparing the scenarios. When under supply-constraint pressure, the four cases exhibit noticeably different behavior. Fleet-based modeling is more efficient in supply-constrained environments at the expense of losing insight on issues such as realistically suboptimal fuel distribution and challenges in reactor refueling cycle staggering. Finer-grained time steps enable more efficient material use in supply-constrained environments resulting in lower standing inventories of separated Pu. Large simulations with fleet-based reactors run much more quickly than their individual reactor counterparts. Gaining a better understanding of how these and other modeling choices affect fuel cycle dynamics will enable making more deliberate decisions with respect to trade-offs such as computational investment vs. realism.