Feasibility Assessment of Adopting the ADDER Computer Code for the MIT Research Reactor Fuel Management
Author(s)
Garanzini, Maurane
DownloadThesis PDF (6.540Mb)
Advisor
Hu, Lin-Wen
Forget, Benoit
Terms of use
Metadata
Show full item recordAbstract
The Massachusetts Institute of Technology Reactor (MITR) is a 6 MW research reactor currently operating with highly enriched uranium (HEU) plate-type fuel. Fuel management calculations for this reactor are performed using MCODE, which allows for the coupling of a neutron transport code and a depletion code. As part of the low-enriched uranium (LEU) fuel conversion program, the Advanced Dimensional Depletion for Engineering of Reactors (ADDER) software is being developed at Argonne National Laboratory to provide a more flexible and performant approach to fuel management. This study evaluates the feasibility of transitioning from MCODE to ADDER for MITR fuel management by carrying out a code-to-code comparison. Analyses for a full MITR cycle (70 days) for a 22-element fresh HEU core and fresh LEU core were completed, and the impact of simplified in-core experiments with various materials was also evaluated. Calculations with mid-cycle restart were performed, in which reactor power was reduced to 100 kW for 7 hours to evaluate Xe-135 poison reactivity effects. The parameters selected for comparison include control blades height, cumulative fission density, integral neutron flux and nuclide inventory (for selected actinides and neutron poisons). The study showed satisfactory agreement between ADDER and MCODE results, with control blades worth differences within the 200 pcm range that corresponds to ± 100 pcm critical search tolerance, and U-235 mass differences remaining below 0.5 g per fuel element at end of cycle. Differences for other result types remain low enough to show the potential of transitioning to ADDER, with higher differences located near control blades when using the predictor-corrector method for depletion since the codes rely on different algorithmic definitions for predictor-corrector as well as different critical blade search schedules. Closer agreement between results is obtained when switching to the predictor method but still indicates some potential differences in power normalization. The two software also present good agreement on control blades height and Xe-135 core inventory results for mid-cycle restart calculations. Further study is recommended to assess depletion factors such as neutron flux normalization and predictor-corrector schemes. Before ADDER is implemented for MITR fuel management, future work is required to evaluate good agreement for equilibrium cores with depleted HEU fuel element compositions, and analyze fuel elements shuffling in between cycles.
Date issued
2024-05Department
Massachusetts Institute of Technology. Department of Nuclear Science and EngineeringPublisher
Massachusetts Institute of Technology