Neutronic evaluation of a PWR with fully ceramic microencapsulated fuel. Part I: Lattice benchmarking, cycle length, and reactivity coefficients

Highlights: • Fully ceramic microencapsulated (FCM) fuel was studied. • Sensitivity and benchmark calculations were performed on lattice level for FCM fuel. • A non-linear reactivity model was vital to represent the reactivity during burn-up. • Reactivity coefficients were determined for FCM versus conventional UO_X fuel. - Abstract: The fully ceramic microencapsulated (FCM) fuel concept is based on the tri-isotropic (TRISO) carbon coated fuel particles. These particles were developed and demonstrated for use in high temperature gas reactors. It has been proposed to use these particles in light water reactors to provide potential operational and safety benefits. The reference fuel in this case assumes TRISO-like particles with a ∼20%-enriched uranium-nitride kernel embedded in a silicon carbide (SiC) matrix. The fuel particles are contained in a “compact” which is then inserted into a cladding. The fuel assembly features the same dimensions as a standard 17 × 17 Westinghouse fuel assembly. FCM fuel requires fission products to traverse several barriers in the proposed fuel design before reaching the cladding. FCM fuel may also reduce fuel-cladding interaction and fuel pellet swelling while enabling higher fuel burn-up. This study is a neutronic evaluation of the use of FCM fuel in an advanced pressurized water reactor (PWR). On the lattice level, the SERPENT Monte Carlo and TRITON deterministic tools were used, while the whole core simulation was based on the three-dimensional PARCS nodal code. This paper presents the results of the lattice-level neutronic study of doubly heterogeneous FCM fuel. Strong agreement was found between the SERPENT and TRITON codes in terms of k-infinity as a function of burn-up, actinide build-up, and “pin” powers. The impact of several simplifying geometric assumptions was considered, such as the use of a square particle lattice within the FCM fuel pins. It was determined that the linear reactivity model does not provide a good estimate of the fuel cycle length, due primarily to non-linear reactivity behavior at high burn-up (>800 effective full power days). To determine cycle length, higher order reactivity models were applied to the lattice results. The calculated cycle lengths are slightly reduced versus a reference uranium oxide case. Finally, the assembly-level reactivity coefficients were calculated as a function of burn-up. The fuel and moderator temperature coefficients were negative for FCM fuel, but reduced in magnitude by approximately 50% versus a reference uranium oxide case