Application of machine learning to predict safeguards parameters for irradiated salts from a molten salt reactor concept

The importance of implementing effective safeguards measures for the ongoing development of Molten Salt Reactors (MSRs) is well acknowledged. A lot about the nature of the spent fuel from these reactors remains to be studied. In light of these prevailing research gaps, in this study, we investigate the use of machine learning analysis of NDA signatures from the irradiated salt for the verification of MSR spent fuel. The techniques will involve the application of machine learning models trained on a dataset of simulated signatures to predict the safeguards-centric parameters such as BIC (Burnup, Initial enrichment, and Cooling time) of irradiated salts from MSRs. Such techniques are expected to be valuable for MSRs where such methods have not been used before. The study is expected to provide much-needed insight into the nature of such irradiated salts and provide the international community a basis for safeguards measures for such spent fuel. • Modeling and simulation of molten salt reactor fuel. • Highlighting key safeguards challenges associated with the molten salt reactor fuel cycle. • Use of novel techniques and new safeguards-relevant signatures to verify spent fuel from molten salt reactors.