Assessing the Risk of Proliferation via Fissile Material Breeding in ARC-class Fusion Power Plants

Abstract: Construction of a nuclear weapon requires access to kilogram-scale quantities of fissile material, which can be bred from fertile material like U-238 and Th-232 via neutron capture. Future fusion power plants, with total neutron source rates in excess of $10^{20}$ n/s, could breed weapons-relevant quantities of fissile material on short timescales, posing a breakout proliferation risk. The ARC-class fusion reactor design is characterized by demountable high temperature superconducting magnets, a FLiBe liquid immersion blanket, and a relatively small size ($\sim$ 4 m major radius, $\sim$ 1 m minor radius). We use the open-source Monte Carlo neutronics code OpenMC to perform self-consistent time-dependent simulations of a representative ARC-class blanket to assess the feasibility of a fissile breeding breakout scenario. We find that a significant quantity of fissile material can be bred in less than six months of full power operation for initial fertile inventories ranging from 5 to 50 metric tons, representing a non-negligible proliferation risk. We further study the feasibility of this scenario by examining other consequences of fissile breeding such as reduced tritium breeding ratio, extra heat from fission and decay heat, isotopic purity of bred material, and self-protection time of irradiated blanket material. We also examine the impact of Li-6 enrichment on fissile breeding and find that it substantially reduces breeding rate, motivating its use as a proliferation resistance tool.