A molten salt community framework for predictive modelling of critical characteristics

Used nuclear fuel (UNF) for both existing and advanced nuclear fuel cycles currently under consideration in many countries, including the US and UK, is usually processed in molten salt (e.g., LiCl-KCl), and fission product and (activated) corrosion product elements that are more active than uranium accumulate in the molten salt as dissolved ions (e.g., Co2+, Nd3+, Pr3+, and Cs+). These elements need to be periodically removed to ensure process safety, recycle the salt, and minimize nuclear waste generation. However, there is a critical knowledge gap associated with the thermodynamics, physical and chemical properties, and in-service behaviour of these salts as a function of their evolution in-service - for example, fission product and actinide content, combined with evaporation / corrosion behaviour, can dynamically alter, in real-time, the composition and therefore critical properties of these molten salts. Indeed, even simple phase relations are poorly understood in fission product / actinide containing salts. For adequate treatment, recycle, and disposal of this nuclear waste stream plus recovering valuable elements such as uranium and transuranic metals, thermochemical and physical knowledge of molten salts is required. However, this knowledge is limited, especially for complex salts, such as LiCl-KCl salts involving lanthanide and actinide elements, oxide contamination, and alkali (Na, Rb, Cs) and alkaline earth (Sr, Ba) elements.

In the proposed project, the extended LiCl-NaCl-KCl-RbCl-CsCl-SrCl2-BaCl2-LnCl3 (Ln = La, Ce, Pr, Nd, Sm, and Y) system has been selected as the primary model system to demonstrate accelerated exploration of structural, thermodynamic, and physical properties using our open-source platform including ML models, high throughput DFT-based first-principles calculations, MD and AIMD simulations, and high throughput CALPHAD modeling using UNIQUC and MQM models. The UK partners will also aim to further study LiCl-Li2O-(RbCl, CsCl, SrCl2, and BaCl2), representative of electroreducer pyro-processing salts, and (LiF-NaF-KF and LiF-BeF2), representative of the US and UK candidate molten salt reactor (MSR) fuel / coolant salts.

Our proposed US/UK collaborative research will develop a computational framework by implementing advanced thermodynamic models to systematically explore critical molten salt characteristics, addressing a critical global knowledge gap in molten salts relevant to future fast nuclear reactors and advanced future nuclear fuel cycles. The proposed research is not only directly relevant to the US NEUP Call Workscope FC-1.2, but also to the corresponding UK EPSRC Call and the wider UK civil nuclear programme. This project will be truly significant for the global molten salt community by providing various open source, high throughput computational approaches and tools, and underpinning datasets, to model and design advanced molten salts.