The international Phébus Fission Product Programme investigated key phenomena occurring in light water reactor core meltdown accidents in a series of five in-pile experiments. Four of these tests focused on the degradation of fuel rod bundles, containing a central control rod, and on the resulting release of fission products, structural materials and actinides from the fuel rod bundle, their transport in the reactor coolant system (RCS) and their subsequent behaviour in the containment vessel. Various steam contents were used in the RCS, from highly oxidising conditions (in FPT0 and FPT1) to more reducing ones (in FPT2 and FPT3).
The “early degradation phase” took place at the beginning of the Phébus driver core and fuel bundle heat-up phase, with a quasi-intact fuel bundle geometry. During this phase, the degradation of the control rod and the oxidation runaway due to the fast oxidation of the Zircaloy claddings of the fuel rods, were two major events which took place.
The oxidation runaway locally increased the temperatures much above the temperatures resulting from the Phébus driver core heat transfer to the bundle and yielded a large hydrogen release, which amounted to 70–80% of the whole hydrogen production during the tests. The maximum hydrogen flow rates increased with increasing steam flow rates injected at the fuel bundle inlet.
The failure mechanisms of silver–indium–cadmium (used in three tests) and boron carbide (used in one test) control rods involve eutectic interactions amongst the components of these control rods. Mechanical deformations of the control rod stainless steel cladding against the Zircaloy control rod guide tube are the main presumed mechanisms for the beginning of these eutectic formations. However, different post-failure scenarios can be postulated for the effect of control rod degradation on fuel bundle degradation for both types of control rods. The exothermic oxidation of the exposed boron carbide pellets led to the release of carbonaceous species (CO, CO2) as well as of additional hydrogen, but no significant methane release could be detected above the limits of detection.
Overall, the results confirmed existing knowledge concerning early phase degradation phenomenology found in previous integral experiments such as CORA and QUENCH (Karlsruhe Institute of Technology) and Phébus SFD (IRSN Cadarache), and formed a sound basis for analysis of the late phase degradation subsequently observed. Quantitative analysis of boron carbide control rod degradation in FPT3 pointed to a need for improved modelling of chemical reactions involving this material, particularly its oxidation in steam; this has been studied in the BECARRE experiments conducted by IRSN in the International Source Term Programme, leading to better quantitative understanding and improved modelling in codes such as the European reference severe accident analysis code ASTEC.