Development and application of an advanced fuel model for the safety analysis of the generation IV gas-cooled fast reactor

Until about the year 2030, current-day nuclear power plants (NPPs) will be replaced by so-called Gen-III or Gen-III+ units, which are mainly based on light water reactor technology. The principal new features are increased safety and improved economical effectiveness. However, these systems use the same fuel forms and are based on the same fuel cycle. Beyond 2030, the interest is likely to shift towards fourth generation NPPs, which offer the possibility of complete fuel cycle closure. Generation-IV reactor concepts include both thermal and fast systems, and involve a wide range of fuel forms and compositions. The present research has been focused on the development of a thermo-mechanical model for the innovative fuel design of the Generation-IV Gas-cooled Fast Reactor (GFR). The principal distinctive feature of the fuel is that the fuel pellets are arranged within plates which enclose an inner honeycomb structure. Apart from the geometry, the usage of new materials is foreseen. Thus, the fuel pellets are of mixed uranium-plutonium carbide, and the cladding is bulk or fiber-reinforced SiC. The setting up of an appropriate materials database was thus the very first task which had to be carried out in the current work. The main purpose of the currently developed model is to provide reliable data, in the context of transient analysis, for the calculation of the principal neutronic feedbacks in the GFR core, viz. the fuel temperature for the Doppler effect and the fuel plate deformation for the axial core expansion effect. None of the available fuel modeling codes is suitable for a realistic simulation of the GFR fuel, as the inner honeycomb structure cannot be explicitly taken into account. The development work has been carried out largely in the context of PSI's generic code system for fast reactor safety analysis, FAST. Thereby, it has mainly involved extension of the thermo-mechanical code FRED, developed originally for the modeling of traditional rodded fuel. Within the FAST system, FRED is coupled to the TRACE code for the thermal-hydraulic modeling, so that the present work has comprised not only the development of a 2D FRED model for the plate-type GFR fuel, but also the implementation of corresponding changes in TRACE for ensuring appropriate information exchange between the two codes. The 2D thermo-mechanical model has been developed with certain assumptions. Since no experimental data exist for this fuel type, benchmarking of the new simulation tool was carried out by building up a detailed 3D model using the finite-elements code ANSYS. The 3D model has, moreover, been employed for conducting certain supplementary studies to obtain an in-depth understanding of the thermal and mechanical behavior of the fuel. It was found how the complex, multi-dimensional, heat transfer in the plate-type fuel accounts for the discrepancies between results of 2D and 1D simulations. Furthermore, it was shown that, under certain conditions, the temperature field can be well predicted by the 1D model with slight modifications of the solution algorithm. Other insights have been obtained from the detailed mechanical analysis. Thus, it has been shown that, during operation, cusping occurs at the pellet periphery which results in an unfavorable concentration of stresses both in pellet and cladding. Several alternative ways to optimize the fuel design and to avoid, or at least minimize, this effect have been proposed. As mentioned, the new fuel model is intended for usage in GFR transient analysis. In order to quantify the impact of the current model developments, a range of hypothetical accident events have been analyzed using the FAST code system, with and without usage of the new fuel model. It has been shown that the pure geometry effects on the temperatures are quite significant. However, for the specific honeycomb structure geometry considered, these are somewhat mitigated by the fuel and cladding expansion and the corresponding decrease of the axial fuel-cladding gas gaps. Accounting for the deformations thus, in this case, brings the 2D model results closer to those of a 1D treatment. The evolution of reactivity feedbacks and the reactor power were evaluated using a point-kinetics approximation and hence were driven by the average temperature change rate (dT/dt) rather than by the absolute temperature values. Correspondingly, the results from the two models are not drastically different. The more significant discrepancies have been obtained for the transients with similar magnitudes of the Doppler and axial core expansion effects (core overcooling, loss of heat sink). For the transient-overpower transient, which is mainly determined by the fuel temperature, very similar results are obtained for total reactivity and reactor power. In brief, the present research has resulted in a flexible and easy-to-use simulation tool for carrying out reliable transient analysis for the Generation IV GFR with its innovative plate-type fuel, the implemented methodology combining an explicit accounting of the fuel inner structure with acceptable computing time.

Related material