Contribution to a better understanding of the detrimental role of hydrogen on the corrosion rate of Zircaloy-4 cladding materials

The kinetics of waterside corrosion of Zircaloy-4 in PWR systems reveal accelerated reaction rates at high burnup levels. In such cases, a thick and dense hydride layer is observed beneath the metal/oxide interface. Laboratory corrosion rests on samples prehydrided using gaseous and cathodic charging techniques clearly indicate a deleterious effect of hydrogen on the corrosion rate. This effect increases with the hydrogen level, particularly in the case of dense hydride layers formed by cathodic charging. The purpose of the present work was to confirm the detrimental effect of hydrogen over a wider range of experimental conditions [steam, primary light water (2 ppm Li, 1000 ppm B), heavy water and air] to determine whether or not it is due to matrix damage during the charging process and to contribute to the understanding of the mechanisms involved. With respect to the effect of the hydrogen-charging technique, a detailed examination of representative specimens after cathodic charging was performed using SEM and TEM. The main conclusions of these observations are that no evolution or degradation of the alpha-Zr matrix or second-phase particles could be identified after hydrogen charging. Moreover, these results are supported by steam corrosion tests on materials with the same hydrogen level and different surface properties (cathodic charging with and without homogenizing heat treatment, pickled or untreated outer surface, gaseous charging). As a result of these investigations, the detrimental effect of hydrogen has been confirmed in different environments studied (primary light water, air, heavy water). Many different techniques were used to study the oxide and the effect of the hydrides. The periodicity of the stratification within the oxide was shorter on prehydrided specimens, and the density of cracks was higher in oxide grown above a dense hydride layer. TEM examination close to the metal/oxide interface showed no significant variation in the oxide morphology, with no change in either the defect population (shape, number, size, etc.) or the zirconia crystallites, which remain essentially columnar. These results are discussed in relation to the different working hypotheses concerning the mechanisms involved, i.e., mechanical degradation of the pride integrity or modification in the transport properties of the oxide grown on hydrided metal.