Investigation of very high burnup UO₂ fuels in Light Water Reactors

Historically, the average discharge burnup of Light Water Reactor (LWR) fuel has increased almost continuously. On one side, increase in the average discharge burnup is attractive because it contributes to decrease part of the fuel cycle costs. On the other side, it raises the practical problem of predicting the performance, longevity and properties of reactor fuel elements upon accumulation of irradiation damage and fission products both during in-reactor operation and after discharge. Performance of the fuel and structural components of the core is one of the critical areas on which the economic viability and public acceptance of nuclear energy production hinges. Along the pellet radius, the fuel matrix is subjected to extremely heterogeneous alteration and damage, as a result of temperature and burnup gradients. In particular, in the peripheral region of LWR UO₂ fuel pellets, when the local burnup exceeds 50-70 GWd/tHM, a microstructural transformation starts to take place, as a consequence of enhanced accumulation of radiation damage, fission products and limited thermal recovery. The newly formed structure is commonly named High Burnup Structure (HBS). The HBS is characterised by three main features: (a) formation of submicrometric grains from the original grains, (b) depletion of fission gas from the fuel matrix, (c) steep increase in the porosity, which retains most of the gas depleted from the fuel matrix. The last two aspects rose significant attention because of the important impact of the fission gas behaviour on integral fuel performance. The porosity increase controls the gas-driven swelling, worsening the cladding loading once the fuel-cladding gap is closed. Another concern is that the large retention of fission gas within the HBS could lead to significant release at high burnups through the degradation of thermal conductivity or contribute to fuel pulverisation during accidental conditions. Need of more experimental investigations about the fuel mechanical properties and their relationship with the local microstructure at high burnup has been recognised, being one of the factors influencing Pellet-Cladding Mechanical Interaction (PCMI). The knowledge of the fuel mechanical properties has also fundamental importance to assess the mechanical integrity of the spent fuel during the back end of the fuel cycle. In this context, the scope of this work was twofold. The first task was the experimental study of the fuel microhardness and Young's modulus in high burnup UO₂ fuels and their relationship with the local porosity, which has a major impact on their variation. Moreover, assessment of the accumulation of the decay damage during storage and its influence on the fuel microhardness has been carried out, in the framework of safety studies on the back end of the fuel cycle at high burnup. The second task consisted in the evaluation of the porosity and pore size distribution evolution in high burnup fuel, with particular focus on the HBS porosity. The experimental relationship between the high burnup fuel Young's modulus and local porosity obtained through combination of acoustic microscopy and microindentation measurements has been compared to the material property correlations commonly used in fuel performance codes, which are based on data from characterization of unirradiated UO₂. The investigation has revealed that the relationship is similar for non-irradiated and irradiated material, but in the latter case an additional factor that takes into account the Young's modulus decrease due to burnup accumulation has to be included in the correlation to match the experimental values. First analysis of the fuel microhardness as a function of the accumulated decay damage has shown that fuel microhardness does not significantly increase when the dose due to the additional decay damage accumulated during storage reaches ∼ 0.1 dpa, in agreement with what observed in unirradiated ²³⁸Pu -doped UO₂. Concerning the second part of the research, a new methodology, based on the adaptive kernel estimator, has been introduced for the experimental determination of the three- dimensional pore size distribution and pore density in the fuel, in order to introduce a best-practice approach in this type of experimental analyses. The study revealed that pore density and pore sizes comparable to what found in the HBS are present.