Development of a 3D APOLLO3® Neutron Deterministic Calculation Scheme for the CABRI Experimental Reactor

CABRI is an experimental reactor at the French Alternative Energies and Atomic Energy Commission (CEA) used to study fuel behavior during reactivity insertion transients. As these transients have a high level of multiphysics, it is important to develop suitable modeling and simulation tools to simulate them in order to be able to optimize testing and control of experimental conditions. This paper focuses on the development and validation of the neutron deterministic APOLLO3® calculation scheme that is included in the CABRI neutronic/thermal-hydraulic multiphysics coupled simulation tool; it represents the first stage of a stepwise validation process for the CABRI multiphysics simulation tool. The neutron calculation scheme is based on a classical two-step approach. The first step consists of a 281-energy-group flux calculation with the TDT-MOC (Method of Characteristics) solver for nuclear data space and energy collapsing for the different CABRI assembly clusters. The biases on a two-dimensional (2D) core neutron calculation due to self-shielding correction and collapsing on a restricted pattern are investigated by means of comparison with a direct full 2D calculation on a quarter core. The second step relies on a three-dimensional (3D) core calculation. Two approaches are presented. The first one consists of a best-effort approach corresponding to a 3D pin-scale description of the core, performing a transport calculation with the SN solver MINARET. And, the second one, a best-estimate approach, which will be implemented for kinetics calculations, relies on solving a simplified transport SPN equation in the solver MINOS with an exact 3D cell description of the core. The best-estimate calculation scheme is then used to analyze three experimental CABRI transients. A stepwise validation process is followed to quantify the calculation biases on physical parameters such as reactivity, reaction rates, and total core power at each step using static reference calculations with the stochastic code TRIPOLI4® or transient experimental data. The next development stage toward a multiphysics calculation scheme will be implementation and validation of coupling with a core thermal-hydraulic model.