Within the scope of the 5 th EURATOM Framework Programme (FP) for the HTR-F and HTR-F1 projects, a new 4-year integrated project on very high temperature reactors (RAPHAEL: ReActor for Process Heat And Electricity) was started in April 2006 as part of the 6 th Framework Programme. The Sub-Project on Fuel Technology (SP-FT) is one of eight sub-projects constituting the RAPHAEL project. R&D conducted in this sub-project focuses on understanding fuel behaviour, determining the limits of state-of-the-art fuel, and developing potential performance improvements. Fabrication processes were worked out for alternative fuel kernel composition (UCO instead of UO 2 ) and coating (ZrC instead of SiC): i) UCO microstructure reduces fission product migration and is thus considered superior to UO 2 under high burn-ups and high temperature gradients. For this reason, the manufacturing feasibility of UCO kernels using modified external sol-gel routes was addressed. The calcining and sintering steps were particularly studied. ii) For its better high temperature performance, ZrC is a candidate coating material for replacing SiC in TRISO (TRistructural ISOtropic) particles. One of the objectives was therefore to deposit a stoichiometric ZrC layer without impurities. An “analytical irradiation” experiment currently performed in the HFR — named PYCASSO for PYrocarbon irradiation for Creep And Swelling/Shrinkage of Objects — was set up to measure the changes in coating material properties as a function of neutron fluence, with samples coming from the new fabrication process. This experiment was started in April 2008 and will provide data on particle component behaviour under irradiation. This data is required to upgrade material models implemented in the ATLAS fuel simulation code. The PYCASSO irradiation experiment is a true Generation IV VHTR effort, with Korean and Japanese samples included in the irradiation. Further RAPHAEL results will be made available to the GIF VHTR Fuel and Fuel Cycle project partners in the future. Post-irradiation examinations and heat-up tests performed on fuel irradiated in an earlier project are being performed to investigate the behaviour of state-of-the-art fuel in VHTR normal and accident conditions. Very interesting results from destructive examinations performed on the HFR-EU1bis pebbles were obtained, showing a clear temperature (and high burn-up) influence on both kernel changes (including fission product behaviour) and the coating layers. Based on fuel particle models established earlier, the fuel modelling capabilities could be further improved: i) Modelling of fuel elements containing thousands of particles is expected to enable a statistical approach to mechanical particle behaviour and fission product release. ii) A database on historical and new fuel properties was built to enable validation of models. This paper reports on recent progress and main results of the RAPHAEL sub-project on fuel technology.

The irradiation experiment HFR-EU1bis, coordinated by the European Joint Research Centre – Institute for Energy, was performed in the High Flux Reator (HFR) at Petten to test five spherical HTR fuel pebbles of former German production with TRISO coated particles in conditions beyond the specifications of current HTR reactor designs (central temperature of 1250°C). In this paper, the behaviour of the fission products (FPs) and kernel micro-structure evolution during the test are investigated. While FP behaviour is a key issue for potential source term evaluation it also determines the evolution of the oxygen potential in the oxide kernel which in turn is important for formation of carbon oxides (amoeba effect and pressurization). Fission-gas release from the kernel can induce additional mechanical loading and finally some FPs (Ag, Cs, Sr) might alter the mechanical integrity of the coatings. This study is based on postirradiation examinations (ceramography + EPMA) performed both on UO 2 kernels and on coatings. Significant evolutions of the kernel as a function of temperature are shown (grain structure, porosity, size of metallic inclusions). The quality of the ceramography results allows characteristics of the intergranular bubbles in the kernel (and estimation of swelling) to be determined. Remarkable results considering FP release from the kernel have been observed and will be presented. Examples are the significant release of Cs out of the kernel as well as Pd, whereas Zr remains trapped. Mo and Ru are mainly incorporated in metallic precipitates. These observations are interpreted and mechanisms for FP and micro-structural evolutions are proposed. These results are coupled to the results of calculations performed with the mechanistic code MFPR (Module for Fission Product Release) and the thermodynamic database MEPHISTA (Multiphase Equilibria in Fuels via Standard Thermodynamic Analysis). The effect of high flux rate and high temperature on fission gas behaviour, grain size evolution and kernel swelling are discussed. In addition, solid-FP behaviour (Cs, Mo, Zr, Ba, Sr) is discussed in connection with the evolution of kernel oxygen potential and evolution of the pressure of carbon oxides. The paper intends to be exemplary on how the combination of post-irradiation examination results and fuel modelling increases fundamentally the understanding of HTR fuel behaviour.

For Very High Temperature Reactors (V-HTR), the study of the Uranium-Carbon-Oxygen system is of major importance to predict the high temperature behaviour of the TRISO fuel particle. Firstly, the high level operating temperature of the fuel materials in normal and accidental conditions requires studying the possible chemical interaction between the UO 2 fuel kernel and the surrounding structural materials (C, SiC) that could damage the particle. The formation of the gaseous carbon oxides at the fuel (UO 2 )-buffer (C) interface that leads to the build up of the internal pressure in the particle has to be predicted. Secondly, the U-C-O ternary system is also involved in the fabrication process of “UCO” kernels made of a mixture of UO 2 and UC 2 . For the fabrication of such mixture of uranium oxide and carbide, the phase diagram and thermodynamic properties of the U-C-O system are necessary to investigate in order to perform adequate heat treatments. For both reasons, a new study of the U-C-O ternary system has been undertaken. Firstly, some thermodynamic calculations (equilibrium CO (g) and CO 2(g) pressures, phase diagrams) were performed using the thermodynamic FUELBASE database dedicated to generation IV fuels [1]. The results allow representing the different phase equilibria involving carbide and oxicarbide phases at high temperature. They also show the high level of CO (g) and CO 2(g) equilibrium pressures above the UO 2±x fuel in equilibrium with carbon which could lead to the failure of the particle for high oxygen stoichiometry of the uranium dioxide. In a second step, the partial pressures of CO (g) and CO 2(g) resulting from the UO 2 /C interaction have been measured by high temperature mass spectrometry. Two types of samples were used (i) pellets made of a mixture of UO 2 and C powders or (ii) UO 2 kernels disseminated in a carbon bed. The kinetic measurements of the release of CO (g) and CO 2(g) lead to measured pressures that are lower than the equilibrium pressures predicted from thermodynamic calculations. This discrepancy can be explained by limitations due to distinct kinetic mechanisms. Rates of CO (g) formation have been established taking into account the oxygen stoichiometry of uranium oxide and temperature. The major gaseous product is always CO (g) which release significantly starts at 1473 K. The influence of the different geometries is shown. The limitative kinetic step can be an interface or a diffusion process as a function of the type of sample. These results underline the up most importance of kinetic factors for studying the UO 2 / C interaction to determine realistic CO (g) pressure levels inside a TRISO particle or to improve the fabrication process of the “UCO” kernels.

Various countries engaged in the development and fabrication of modern fuel for the High Temperature Gas-Cooled Reactor (HTGR) have initiated activities of modeling the fuel and fission product release behavior with the aim of predicting the fuel performance under operating and accidental conditions of future HTGRs. Within the IAEA directed Coordinated Research Project CRP6 on “Advances in HTGR Fuel Technology Development” active since 2002, the 13 participating Member States have agreed upon benchmark studies on fuel performance during normal operation and under accident conditions. While the former has been completed in the meantime, the focus is now on the extension of the national code developments to become applicable to core heatup accident conditions. These activities are supported by the fact that core heatup simulation experiments have been resumed recently providing new, highly valuable data. Work on accident performance will be — similar to the normal operation benchmark — consisting of three essential parts comprising both code verification that establishes the correspondence of code work with the underlying physical, chemical and mathematical laws, and code validation that establishes reasonable agreement with the existing experimental data base, but including also predictive calculations for future heating tests and/or reactor concepts. The paper will describe the cases to be studied and the calculational results obtained with the German computer model FRESCO. Among the benchmark cases in consideration are tests which were most recently conducted in the new heating facility KUEFA. Therefore this study will also re-open the discussion and analysis of both the validity of diffusion models and the transport data of the principal fission product species in the HTGR fuel materials as essential input data for the codes.

A fuel performance analysis code for a very high temperature gas-cooled reactor (VHTR) COPA ( Co ated Pa rticle) is being developed at the Korea Atomic Energy Research Institute (KAERI). The COPA code consists of nine modules: BURN, TEMTR, TEMPEB, TEMBL, MECH, FAIL, FPREL, ABAQ, and MPRO. The BURN determines neutron flux and fluence at a location of a reactor core, and then calculates a fuel burnup, a fission rate per volume and a fission product inventory throughout a fuel particle and a fuel element. The TEMTR, TEMPEB and TEMBL calculate the temperature distributions in a coated fuel particle, a pebble and a fuel block by using a one-dimensional finite difference method, respectively. The MECH performs mechanical calculations on a coated fuel particle by using a finite element method. The FAIL performs probabilistic calculations to estimate the failure probabilities of the coating layers during an experiment or a reactor operation. The FPREL estimates the migrations of gaseous and metallic fission products through a fuel particle and a fuel element by using a one-dimensional finite difference method. The ABAQ performs the analysis of the crack and debonding in a coated fuel particle. The MPRO calculates the material properties of the kernel, low-density pyrocarbon, high-density pyrocarbon, silicon carbide, matrix graphite, and structural graphite. Each module is used to produce input data for other modules or is inserted into other modules. The COPA code is one of the computer codes taking part in the IAEA-CRP-6 benchmarking program. The stresses and failure fractions calculated by the COPA-MECH and COPA-FAIL showed good agreements with the results by the other countries’ codes. In order to establish a good database of the related material properties, KAERI is participating in an international irradiation experiment, is planning its own irradiation and post-irradiation experiments, and will perform ab-initio calculations on the fuel materials.

The strong correlation between the density and the physical and, mechanical properties of graphite suggests that the method of nondestructive density evaluation could be developed into a characterization technique of great value for the overall improvement of safety of graphite moderator reactors. In this study, the oxidation-induced density changes in nuclear graphite for VHTR were determined by a conventional destructive bulk density measurement method (BM), and by a new non-destructive method based on acoustic microscopy and image processing (AM). The results were compared in order to validate the applicability of the latter method. For a direct comparison of the results from both measurements, two specimens were prepared from a cylindrical graphite sample (1 inch diameter and 1 inch height, oxidized to 10% weight loss at 973 K in air for 5 hours). The specimens were used for characterization by BM and AM methods, respectively. The results show that, even with a large standard deviation of the AM, the density changing trend from both methods appeared the same. This observation may be attributed to the fact that AM images reflect characteristic density changes of the graphite sample through the acoustic impedance changes. This study demonstrates the possibility of using AM as a nondestructive technique for the evaluation of density changes in graphite when a database is prepared through a systematic series of experiments.

The very high temperature reactor (VHTR) materials R&D started as an objective of the development of key technologies for a nuclear hydrogen production funded by the Korea government in 2006. We are performing materials R&D for 5 components: a reactor pressure vessel (RPV), a reactor/process intermediate heat exchanger (IHX) and hot gas duct, a control rods component, a reflector and support structures in the core region and reactor materials for the sulfur-iodine (SI) process. The scopes of our works are focused on a material screening/selection and qualification, codifications of a high temperature structural design to a very high temperature region, and to support the licensing of a system design, material characterizations and database establishments. Current target materials are modified 9Cr-1Mo, alloy 617, graphite, ceramic fiber reinforced composite, Fe-Si and SiC.