Zircaloy-4 cladding has been oxidized in autoclave at 633 K with various chemistry conditions ([Li+] = 0, 10, 70, 700 ppm and [B] = 0, 650 ppm as boric acid) to quantify the effect of lithium hydroxide and boric acid on the oxidation kinetics and to provide oxide films that have been analyzed by SIMS, SEM, and RAMAN spectroscopy to improve our knowledge of the oxidation process in a lithiated environment. Additional specific tests consisting of isotopic exchanges (6Li+/7Li+) have also been conducted to study the incorporation of lithium in the oxide films. The main results are that: • Depending on the water chemistry, the lithium hydroxide has two effects on the oxidation kinetics: a decrease of the time to transition and a strong enhancement of the post-transition oxidation rate. • The boric acid strongly reduces these two effects of lithium hydroxide. • In the oxide film, the lithium is located mainly in the pores (or adsorbed on their walls). • The lithium has a quick access to the bulk of the oxide film, even in the inner barrier layer (considered as not completely impervious). • The enhancement of the oxidation rate due to lithium hydroxide is linked to the degradual alteration of the inner barrier layer, up to its quasi-disappearance. These experimental results are described extensively, and several factors involved in the alteration of the inner barrier layer, such as lithium pickup in the ZrO2 film, evolution of the morphology of the oxide grains, tetragonal zirconia phase transformation, and hydriding, are then discussed.

Waterside corrosion of Zircaloy cladding in pressurized water reactors (PWRs) is largely dependent upon the operating parameters and microstructure of the zirconium alloys. The impact of these parameters on the corrosion kinetics of Zircaloys is investigated on the basis of empirical data and experiences that can be interpreted using existing corrosion models. The influence of thermo-hydraulic data, heat flux, local boiling conditions, and of the growing oxide films has been studied from corrosion tests performed in static autoclaves or in out-of-pile loops. These parametric investigations are described as well as the models that were developed. The impact of microstructure is studied from the comparison of the corrosion behavior of different Zircaloy-4 specimens corroded in out-of-pile tests. In particular, a poor corrosion resistance of an experimental Zircaloy-4 material is analyzed as a function of the microstructure close to the metal/oxide interface. The impact of the alloy composition and primary coolant chemistry on the corrosion kinetics of Zircaloy-4 is modeled empirically or uses a mechanistic approach that proposes a series of chemical equations with a mathematical representation of the kinetics. These proposed models are then used to investigate the corrosion behavior of Zircaloy-4 cladding in 17 by 17 plants for rods irradiated at high burnups. Higher PWR operating cycles, core average coolant temperature, power, and elevated primary coolant lithium concentrations (3.5 to 4 ppm) are then simulated and discussed in terms of Zircaloy corrosion resistance considerations.

International audienc

No abstract

We report here the investigations on the corrosion of Zircaloy-4 cladding in out-of-pile and in-pile loops. Tests were conducted on the cladding in the fresh, preoxidized, and preirradiated conditions. The exposures were in low and high lithium coolant, with and without boron, and under single-phase and two-phase heat transfer conditions. In the out-of-pile loop tests, under single-phase heat transfer condition, acceleration was observed when lithium was at 70 ppm and higher. Boron abated the lithium effect. Under two-phase heat transfer conditions, increased corrosion was seen with 10 ppm lithium and 100 W/cm2 only at void fractions >30%. Boron, even at 100 ppm, had an ameliorating effect. In the n-pile tests, with 4 and 10 ppm Li and 1000 ppm B for a total exposure of 240d, the exit void fraction was low (∼5%), compared to the out-of-pile tests (34%). The prefilmed oxide thickness varied from 5 to 50 μm, compared to 5 μm in the out-of-pile tests. However, an enhanced corrosion due to lithium was not observed. Thin pretransition films, grown at locations corresponding to a zero void fraction, showed hundreds of ppm lithium in the case of oxides grown in high lithium coolant and tens of ppm lithium in the case of oxides grown in low lithium coolant. Addition of boron to the high lithium coolant reduced the lithium pickup by nearly an order of magnitude. In the case of thick post-transition oxides, the lithium depth profiles showed marked differences for films grown under single-phase (with high lithium in the coolant) and two-phase (with low lithium in the coolant) heat transfer conditions. In the former case, the lithium concentration was highand the profiles were flat. In the latter case, the lithium concentration was high only in the near surface layers and was one to two orders of magnitude less in the bulk of the oxide. It appears that lithium became concentrated in the liquid layer on the surface of the cladding, under conditions of boiling and high void fraction (>30%), and resulted in rapid corrosion; subsequently, further corrosion had occurred in steam containing volatilized lithium hydroxide. While some uncertainties regarding the lithium effect in-reactor remain from a mechanistic point of view, we have concluded that a lithium effect increasing rapidly the corrosion rate even towards the end of a fuel cycle, when the boron level is expected to be low, is highly unlikely. The heat flux and the void fraction, even under nucleate boiling conditions, would be quite low. In addition, with the lithium at 2 ppm and still some boron left in the coolant, the thermal hydraulics-water chemistry combination is far removed from the conditions of 10 ppm lithium, 100 W/cm2, and >30% void fraction required to bring about an enhanced corrosion in ∼26d.