Material release from the bundle in Phébus FP

Abstract The experiments Phebus FP FPT-0 to FPT-3 tests conducted by IRSN have produced a large quantity of data on fission product, structural material and actinide release under prototypic nuclear reactor accident conditions that form a consistent set of data suitable for assessment of release models implemented in severe accident codes. Additionally, FPT4 provides valuable complementary data in a late phase debris bed geometry. Based on on-line and post-test measurements, the release of the main fission products could be assessed as well as their release kinetics in relation with bundle degradation progresses. Noble gases (Xe, Kr) are considered as very highly volatile; the other fission products can be classified as high-volatile (Cs, I, Te, Mo), semi-volatile (Rb) and low-volatile (Ru, Ba, La). Release of fission products is observed mainly in the oxidation phases, with the proportions depending on the elements concerned. Molten pool formation corresponds to a decrease of the release process. Post-test gamma-measurements show that deposition of fission products over the upper parts of the fuel rods is important with lower steam flow rates as given in FPT2 and FPT3. The fuel material fractional release (mainly U) is very low but could represent a significant fraction of the transported aerosol mass (FPT0 and FPT1). Release from the silver–indium–cadmium control rod is characterised by the initial Cd burst release on failure of the control rod, followed by a steady release of Ag, In and Cd from the molten absorber pool remaining in the stub. These elements are important as they can affect the physical and chemical forms of fission products such as iodine, hence their transmission to the containment. A large boron release, linked to an extended degradation of the boron carbide control rod, was observed during FPT3. The established link between bundle degradation and release mechanisms is clearly confirmed for both fuel and control rod materials, with the release being dependent on temperature, burnup, interactions between fuel and structural materials (mainly Zircaloy cladding, i.e. for Ba) and atmosphere (oxidising/reducing conditions, as can be noticed for Mo, Sn and Te), and mass flow rate. Comparison with the results of separate-effects test series (VERCORS, QUENCH) performed under similar temperature and atmosphere conditions will also be presented.