Dynamics, Neutronics, and the Safeguarding of Molten Salt Reactors from Plutonium Diversion

Molten Salt Reactors (MSRs) face unique challenges and requirements for material accountancy and control. Current techniques such as item counting and serial numbers are suited for the current fleet of solid-fueled reactors. However, as the next generation of reactors enter the market, new means of material accountancy and safeguarding will need to be introduced. This research takes a novel approach to assessing the plutonium concentration in the liquid fuel medium found in MSRs. When a reactor is subjected to a periodic reactivity transient, it can have drastically different responses depending on frequency of the transient. This frequency response is sensitive to parameters, such as the delayed neutron fraction, that are highly linked to what isotope is currently fissioning (e.g. U-235, Pu-239, etc.). Therefore, if plutonium is diverted, a characteristic change in neutronic parameters are observable in a reactor frequency response. Using this discrepancy, the frequency response can be very low latency, nondestructive measurement that does not disturb reactor operations. This research examines the frequency response while plutonium is being diverted in fast and thermal spectrum MSRs. In doing so, the paper reveals how key neutronic parameters evolve with burnup and change when plutonium is diverted. The means to predicting the frequency response of an MSR is also detailed. Finally, the frequency response is quantified for both normal burnup conditions as well as when a significant quantity of plutonium has been diverted. The final results show that for a small modular reactor sized thermal MSR, a diversion of plutonium is very noticeable in the frequency response. However, for the large fast spectrum MSRs, there is little to no change in the parameters. Therefore, the frequency response signature is not recommended for large MSRs. All custom tools created to perform this study have been uploaded on a public repository. They can be quickly adapted for evaluation of any liquid fuel reactor design. While this study used these specific tools, the method is not married to any tool (i.e. either burnup code or dynamic model) can be replicated using the steps presented here.