Design and implementation of a modular modeling framework for the ANICCA nuclear fuel cycle simulation code

This thesis presents the design and implementation of a new modular simulation framework for ANICCA, the nuclear fuel cycle simulation code developed at the Belgian Nuclear Research Center (SCK CEN). The new framework introduces a three-layer architecture separating graph topology and mass dispatch, facility inventory management, and physical process execution. A nuclear fuel cycle scenario is formally represented as a directed multigraph. The currently implemented mass dispatch algorithm resolves mass flows through a two-pass sequential modular approach: a backward pass propagating demand upstream in reverse topological order, and a forward pass physically transferring packages in topological order. Recycling loops, which break the acyclic assumption required by this algorithm, are handled through a graph-tearing strategy. Four facility cycle types, fixed demand, on demand, variable demand, and none, define how facilities interact with the solver. Radioactive decay and fuel irradiation are both solved using the Chebyshev Rational Approximation Method (CRAM) with LU factorization caching, and the irradiation process is built around a multi-batch reactor core model driven by pre-computed fuel libraries. The framework is validated and demonstrated through four case studies, which also reveal several modeling errors present in the legacy code which are corrected in the new version.