Fission matrix methods for temperature and control feedback of small nuclear systems
Open Access
- Author:
- Rau, Adam J
- Graduate Program:
- Nuclear Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 03, 2020
- Committee Members:
- William J Walters, Dissertation Advisor/Co-Advisor
William J Walters, Committee Chair/Co-Chair
Elia Merzari, Committee Member
Amanda M. Johnsen, Committee Member
Thomas Litzinger, Outside Member
Arthur Thompson Motta, Program Head/Chair - Keywords:
- Fission Matrix
Reactor Physics
Neutronics
PSBR
TRIGA
Serpent
Monte Carlo
TRACE
Breazeale - Abstract:
- Understanding and modeling reactor physics is essential to the design and operation of fission nuclear power plants. Although Monte Carlo simulations provide the most accurate models of reactor physics, running them can be time consuming. This can prohibit the use of Monte Carlo in situations that require sequences of neutronics solutions, such as transient and multiphysics calculations. The present work demonstrates a fission matrix methodology that calculates the fission source distribution and multiplication factor in the Penn State Breazeale Reactor (PSBR) while accounting for control rod movement and fuel temperature feedback, as well as coupling with a D2O tank. These methods are also tested on a nuclear thermal propulsion (NTP) reactor. Given an arbitrary state, a new fission matrix can be obtained by interpolating from a database of fission matrices that are tallied using Monte Carlo. Different configurations of the initial database and ways to estimate the fission matrix are discussed, and their effect on accuracy compared to Monte Carlo is quantified. For the nuclear thermal propulsion reactor and the PSBR case where control rods are moved together, keff error is within ±25 pcm, normalized root-mean-square average (RMS) 3D fission source error is less than 1.3%, and maximum normalized 3D fission source error is within 6%. When control rods are moved individually in the PSBR, keff error is within ±70 pcm, normalized RMS 3D fission source error is less than 1.9%, and maximum 3D fission source error is within 18.1%. Depending on the method used, simulating fuel temperature, control rod position, and coupling with the D2O tank requires a database of 16 - 31 Monte Carlo calculations. However, after these calculations are performed, individual fission matrix calculations take around 0.2 - 0.3 seconds, instead of the minutes / hours required of a Monte Carlo calculation. Additionally, a Trace thermal-hydraulic simulation of the PSBR is developed and coupled to the fission matrix method. Validation is performed on the coupled Trace-fission matrix model for core power levels ranging from 50 kW to 1 MW. For these cases, the RMS average error with experimental data is 12.7 K for instrumented element fuel temperature and $0.07 for reactivity change. Agreement is generally within estimated experimental uncertainty. The potential impact of fuel temperature, control rod positions, and coupling with the D2O tank on fuel management are also considered, as well as effects from xenon and temperature changes in other materials. Finally, recommendations for updating the fuel management code are discussed.