Nuclear fuel cycle optimization: identifying deployment strategies under multiple constraints

The growing need for low-carbon, dispatchable electricity and increasing concerns over energy security of supply have renewed interest in nuclear energy worldwide. While nuclear power provides reliable baseload generation with very low greenhouse gas emissions, conventional once-through fuel cycles lead to inefficient use of nuclear materials and increased quantities of material requiring disposal, as about 95% of recyclable materials remain in the waste alongside other long-lived nuclear waste, leading to long-term radiotoxic nuclides. These limitations motivate the development of Advanced Nuclear Fuel Cycles (ANFCs), which enable multiple recycling of spent fuel, separation and transmutation of minor actinides, and deployment of fast reactors and other advanced systems aimed at reducing waste burdens and optimizing geological disposal. Determining how a transition towards ANFCs should actually be carried out (i.e. which facilities to build, when, and how to route material between them over under logistical and geopolitical constraints) is a highly-complex planning and optimization problem.

NFC modeling has been the subject of many studies over the past decades, and has led to the development of numerous simulation software which have progressively evolved towards increasingly complex and realistic modeling. However, state-of-the-art simulators remain focused on evaluating predefined scenarios, with facility deployment and material routing, at best, optimized one time step at a time. Assessment is therefore limited to comparing a handful of hand-built scenarios through sensitivity and uncertainty analyses, rather than searching the deployment space for optimal fuel-cycle configurations. This challenge is compounded by long planning horizons, multiple technological pathways, complex facility modelling, and competing objectives such as resource utilization, waste minimization, proliferation resistance, cost, and societal acceptance. As a result, no study has yet identified optimal fuel-cycle transition strategies under realistic constraints.

This PhD project addresses this gap by developing a methodology and software framework for ANFC optimization. Building on SCK CEN's ANICCA fuel-cycle simulator, ALEPH2 depletion code, and associated physics data, facility models, and material-flow descriptions, the project will develop an optimization framework based on graph theory and artificial intelligence techniques. This approach will enable systematic exploration of transition pathways while incorporating constraints and parameters identified through collaboration with international experts, notably within the OECD/NEA Nuclear Science Committee. The research will initially focus on brownfield scenarios involving existing infrastructures, industrial capabilities, and material inventories, before extending the methodology to greenfield systems as idealized reference cases but taking into account the multiple constraints.