Thermal hydraulics analysis of ITU TRIGA MARK II Research Reactor with 3D computational fluid dynamics simulation

Abstract

Research reactors on the contrary to power reactors are not used for energy production. However, they play a key role for development of nuclear science and technology and contribute the development of power reactors. In this regard, the primary use of research reactors is to provide a neutron and gamma source through in core and out core positions for research or sometimes for commercial purposes such as radioisotope production. The other important contribution of them is on code development and validation. Indeed, they provide a connection between computational tools and experimental data therefore help the development of codes that are used for design and analysis of nuclear power plants. From the technical point of view, research reactors are smaller and simpler than power reactors. They usually operate at low temperature and low pressure conditions. In addition, they may contain highly enriched uranium. The energy from fission must be transferred from the fuel to a coolant. The cooling is generally based on natural convection for the low power research reactors whereas the high power research reactors need forced cooling. Developed by General Atomics, ITU TRIGA Mark II research reactor reached first criticality on March 11, 1979 and serves for training, education, neutron activation, gammagraphy, neutrongraphy, and irradiation. ITU TRIGA Mark II is a 250 kW open pool type research reactor housed in a hexagonal structure providing both structural integrity and biological shield. It can pulse up to 1200 MW for short periods of time. The fuel is specially designed to provide inherent safety to the reactor. Uranium Zirconium Hydride (UZrH) fuel material is in stainless steel cladding. Throughout the core graphite is heavily used for several purposes. There are three beam ports (out-core), central thimble, and pneumatic system (in-core) for irradiation. The reactor control is accomplished by three control rods and reactivity feedbacks. The general objective of nuclear reactor safety is to protect the safety barriers especially fuel clad integrity which is called the second barrier of defense in depth concept (first barrier being fuel material itself). Several criteria introduced to protect the system. Safety analysis are performed to make sure that these criteria met for normal operation and accidents. Therefore, the safety analysis of nuclear reactors is the most important aspects in the design and safe operation. The main scope of this thesis is to provide highly accurate 3D solution for neutronics and thermal hydraulic phenomena in ITU TRIGA Mark II research reactor. Since the governing physics of these phenomena are coupled, two solution models based on the governing physics principle should be performed integrally. In this way, the nuclear data and core parameters such as material temperatures, especially the fuel temperature, and coolant density are calculated as precisely as possible thanks to 3D conjugate heat transfer modelling. The first analysis performed in this thesis is neutronic analysis. It was performed using Monte Carlo code MCNP 6.2. In this regard, the MCNP 6.2 neutronic model with full core structure (fuel elements, graphite dummy elements, irradiation channels, neutron source element, control rods, thermal column, beam tubes, graphite reflector, and concrete shielding) generated by Asst. Prof. Dr. Senem Şentürk Lüle was modified according to the needs of this thesis. The neutronic calculations were performed with ENDF/B-VII data libraries for continuous energy interactions and S(α,β) kernel scattering tables to treat low energy (< 4 eV) thermal scattering contribution for Hydrogen in Zirconium Hydride (ZrH) and H2O moderation materials. In order to perform thermal hydraulic simulations, the axial distribution of volumetric heat generation at 250 kW power in each fuel element is necessary. In this regard, cell averaged flux (F4) and superimposed mesh tally (FMESH) features of MCNP code were employed. A second order polynomial was obtained by curve fitting to acquire the axial variation of volumetric heat generation in fuel elements and heat flux at the surface of each fuel elements. These polynomial functions were then inserted as a thermal boundary condition by using UDF feature of FLUENT commercial computational fluid dynamics code for thermal hydraulic calculations. The second analysis in this study is thermal-hydraulic analysis. The detailed geometry that included pool and core structure was generated to simulate fluid dynamics / heat transfer. However, preliminary simulations showed that this representation was time demanding due to large computational domain. Therefore, the thermal hydraulic investigation had been performed in two separate stages to reduce the computational cost. At first stage of thermal-hydraulic analysis, the TRIGA pool was modelled by FLUENT version 18.2 to analyze natural convection circulation under steady-state full power operating condition and to predict velocity field and pressure distribution in the core which will be used in the second stage of the thesis. All the components in the core (fuel elements, graphite reflector, thermal column and top and bottom grid plates) were modeled in detail. The calculated coolant temperatures were compared with experimental data from the literature. The results are in good agreement. Furthermore, the effect of grid plates on cooling performance and velocity streamlines in the pool tank was investigated by creating another pool model without grid plates. The grid plate sensitivity analysis showed that grid plates do not have significant influence on temperature distribution. Whereas, the velocity field of pool is reasonably affected from top grid plate. The existence of grid plates reduces coolant velocity at the core exit. As a result, it can be said that the grid plates play role in reduction of dose at top of the pool since they increase the rise time of activation product Nitrogen-16. At second stage of thermal-hydraulic analysis, only the part inside the reflector was modelled to perform conjugate heat transfer. Therefore, this stage is called as the core model. In this model, heat conduction in fuel elements and natural convection was performed by FLUENT code. The core model was validated and verified with fuel temperature results from instrumented fuel elements at 250 kW power recorded in the logbook of ITU TRIGA Mark II research reactor. The benchmarking showed that, the percent error between simulation and experimental results are below 1 % indicating excellent agreement. Furthermore, the pool boiling phenomenon had been numerically investigated in the core. According to temperature distribution in the core, the pool boiling curve indicates that the overall flow regime in the core is in single phase or at convective stage. However, the bubble formation occurs locally at some locations on the central fuel elements. The subcooled boiling regime arises at these points. Finally, the effect of thermal hydraulic parameters on neutronic behavior had been investigated by upgrading the density and temperate of coolant and temperature of fuel elements in MCNP neutronic model according to results of thermal-hydraulic analysis. In this regard, it can be said that the decrease in density of coolant and increase in fuel temperature inserts negative reactivity in the core due to reduction in moderation and Doppler Broadening, respectively. Unlike previous thermal hydraulic studies that had been performed with major simplifications such as having only 1D, no flow restriction namely no form losses, and no crossflow effects, this thesis offers 3D, fully detailed, validated, and verified neutronic and thermal hydraulic solution. The radial and axial temperature distributions in all 69 fuel elements were provided together with coolant temperature distribution in the tank. Furthermore, modelling of grid plates is out of ordinary since it is usually not performed to provide simplicity.