DEVELOPMENT OF A PRACTICAL FUEL MANAGEMENT SYSTEM FOR PSBR BASED ON ADVANCED THREE-DIMENSIONAL MONTE CARLO COUPLED DEPLETION METHODOLOGY
Open Access
- Author:
- Tippayakul, Chanatip
- Graduate Program:
- Nuclear Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 29, 2006
- Committee Members:
- Kostadin Nikolov Ivanov, Committee Chair/Co-Chair
C Frederick Sears, Committee Member
Yousry Azmy, Committee Member
Ludmil Tomov Zikatanov, Committee Member - Keywords:
- PSBR
Monte Carlo coupled depletion
Research reactor modeling - Abstract:
- The main objective of this research is to develop a practical fuel management system for the Pennsylvania State University Breazeale research reactor (PSBR) based on several advanced Monte Carlo coupled depletion methodologies. Primarily, this research involved two major activities: model and method developments and analyses and validations of the developed models and methods. The starting point of this research was the utilization of the earlier developed fuel management tool, TRIGSIM, to create the Monte Carlo model of core loading 51 (end of the core loading). It was found when comparing the normalized power results of the Monte Carlo model to those of the current fuel management system (using HELIOS/ADMARC-H) that they agreed reasonably well (within 2% – 3 % differences on average). Moreover, the reactivity of some fuel elements was calculated by the Monte Carlo model and it was compared with measured data. It was also found that the fuel element reactivity results of the Monte Carlo model were in good agreement with the measured data. However, the subsequent task of analyzing the conversion from the core loading 51 to the core loading 52 using TRIGSIM showed quite significant difference of each control rod worth between the Monte Carlo model and the current methodology model. The differences were mainly caused by inconsistent absorber atomic number densities between the two models. Hence, the model of the first operating core (core loading 2) was revised in light of new information about the absorber atomic densities to validate the Monte Carlo model with the measured data. With the revised Monte Carlo model, the results agreed better to the measured data. Although TRIGSIM showed good modeling and capabilities, the accuracy of TRIGSIM could be further improved by adopting more advanced algorithms. Therefore, TRIGSIM was planned to be upgraded. The first task of upgrading TRIGSIM involved the improvement of the temperature modeling capability. The new TRIGSIM was upgraded to be able to model various temperatures across different materials of the fuel element in the reactor core. The analysis of the temperature modeling capability demonstrated expected reactivity loss as a function of temperature. Secondly, the depletion capability of TRIGSIM was tremendously improved. The upgrade of the depletion capability involved the replacement of the simple predictor depletion algorithm used in the original TRIGSIM with the more advanced predictor-corrector depletion algorithm. Moreover, the methodology of combining the online burnup cross section generation from the Monte Carlo for “important” isotopes and the use of pre-generated TRIGA burnup cross section library for “non-important” isotopes was implemented in the new TRIGSIM as well. For the last part of the improvements of depletion capability, TRIGSIM was modified to be able to perform depletion calculations in several axial nodes which reflects better burnup gradient along the axial direction. Thirdly, the possibility to speed up the Monte Carlo calculation was studied and implemented. In this research, the speedup of the Monte Carlo calculation was performed by utilizing the fast nodal diffusion calculation to provide initial source distribution for the Monte Carlo method. The results showed that the some computational time was saved by eliminating the typical guess of large number of inactive cycles. Along with this speedup methodology, the algorithm to generate the consistent diffusion cross section from the Monte Carlo was also developed. In addition, the speed of the new fuel management system was also possible by the utilization of parallel computing as it was illustrated that parallel computing had a great potential for reducing clock time. Finally, the upgraded TRIGSIM was renamed as TRIGSIMS to reflect these major improvements. Subsequently, the new TRIGSIMS was validated by performing several core loading configurations starting from the first operating core loading (core loading 2). It was found that the new TRIGSIMS produced good agreements with the measured data when taking into account all uncertainties. With all improved features and advanced methodologies, TRIGSIMS is currently a state-of-art and powerful tool for PSBR fuel management analysis.