Book Description

Zirconium-based alloys have been used in nuclear reactors as fuel cladding and structural materials since the development of nuclear energy. Zircaloy-2, a Sn-Fe-Cr-Ni alloy was widely in service for years, and still is today in boiling water reactors (BWR). Among the many challenges the materials face during operation in the nuclear reactor, hydrogen pickup during corrosion of the metal components is of great concern due to the embrittlement properties of the zirconium hydrides. Zircaloy-2 materials show great corrosion resistance in the boiling environment but many in-pile fuel cladding and structural components, such as water rods and channel boxes, revealed accelerated ingress of hydrogen at high burnup when exposed for additional cycles in the reactor, while Zircaloy-4 components did not. The industry is driven toward increasing the fuel burnup in the reactors, as it reduces operation costs, and therefore it is necessary to prevent this effect from happening in modern alloys. Because the main difference between Zircaloy-2 and Zircaloy-4 is the removal of nickel replaced by additional iron in Zircaloy-4 - nickel was linked to increased hydrogen pickup as early as the 1960's - nickel was thought responsible for this acceleration of hydrogen pickup during the additional cycles in the reactors. In a previous study, metallic nickel was measured in the oxide layer near the metal interface of high hydrogen pickup Zry-2 water rods. In this work, additional materials were selected at low and high elevations in the Zircaloy-2 water rods corroded for 3 and 4 cycles in a BWR (Limerick-1) with low and high hydrogen pickup respectively; and were examined by microbeam X-ray absorption near-edge spectroscopy (XANES), microbeam X-ray diffraction (XRD), and scanning electron microscopy (SEM) in an effort to verify and understand further this observation. Cross-sectional samples were prepared from the two water rods and investigated at the Advance Photon Source (APS) at Argonne National Laboratory (ANL). In each material, the oxidation state of nickel atoms in the thick oxide layers was measured as a function of distance from the metal interface by XANES. The results confirm the presence of metallic nickel in the oxide layer of the high elevation material/high hydrogen pickup material (4 cycles) where 30-35% metallic nickel was seen in the near oxide (up to 10-12 [mu]m from the metal interface), as previously observed in two other high hydrogen pickup materials. At low elevation in the high hydrogen pickup water rod, the correlation was not directly verified (nickel atoms were fully oxidized in the oxide layer past 3-4 [mu]m from the metal/oxide interface) but we argue that the high hydrogen content observed at that location results from the diffusion down the water rod of hydrogen absorbed at higher elevation, driven by the concentration and temperature axial gradients. A detailed analysis of the XANES signal from the metallic nickel atoms in the oxide layer of the high hydrogen pickup material suggest that these nickel atoms are no longer bonded to zirconium atoms, which shows that the metallic nickel which can affect hydrogen pickup consists of atoms in solid solution or in small clusters in the oxide layer, rather than in second phase precipitates. This is in agreement with recent APT examinations of high burnup Zry-2 materials with high hydrogen pickup in which the nickel atoms were seen uniformly distributed in the oxide layer and only small clusters were observed. Additionally, metallic nickel in the outer oxide region close to the water interface was observed in most materials, with the highest metallic fraction (up to 75%) in the low hydrogen content samples. Nickel and iron high fluorescence counts near the oxide/water interface confirmed that the nickel atoms at that location corresponded to deposits from the corrosion of other reactor components on the water rod oxide surfaces. However, these metallic nickel atoms near the water interface of the thick oxide layers (>25 [mu]m) do not seem to affect the hydrogen uptake in the Zry-2 materials as they were mostly observed in the low hydrogen pickup samples. Many cracks (lateral and through thickness) were seen in the oxide layers of the materials with SEM imaging of the prepared samples, especially in the high hydrogen pickup water rod at high elevation. In all four materials investigated, the oxide layers were rather uniform, but extensive circumferential oxide thickness variations could be observed between different regions of the water rods. An increase in oxidation kinetics during the 4th cycle was seen at mid/high elevation, where the irradiation flux is the most intense, by comparing the oxide thicknesses of the 3-cycles to 4-cycles GNF water rods and was correlated to the presence of nickel in the oxide layer. As such, irradiation seems to play an important role in accelerating corrosion (as previously reported) and in stabilizing metallic nickel in the oxide layer (and in turn enhancing hydrogen pickup). Concurrently to the XANES examinations, X-ray diffraction patterns were collected in the oxide layers of the cross-section samples as a function of distance from the bulk metal in order to investigate the oxide microstructure (phase content, grain size, texture) of in-reactor Zry-2 materials at high burnup with low and high hydrogen pickup fraction. The oxide layers formed on the BWR Zry-2 water rod consisted of small and highly oriented monoclinic oxide grains, with a small fraction of tetragonal grains, maximum near the metal interface (3-6%). Grain growth was observed in all materials as the oxide thickens, especially at high elevation, with grain sizes at 17-20 nm near the bulk metal and 33-38 nm in the outer region. Additionally, small grains compose the oxide region near the metal/oxide interface of the high elevation/high hydrogen pickup material which is coherent with accelerated corrosion taking place during the 4th cycle. In all materials, an orientation relationship was apparent between the (111) m-ZrO_2 and the (101 ̅0) [alpha]-Zr crystal planes, and for a significant fraction of the oxide grains throughout the whole oxide layer, the (200) m-ZrO_2 direction is close to the oxide growth direction. This is coherent with previous XRD examinations of autoclave and in-reactor corroded Zr-alloys. After a thorough review of the presented results and of the literature available, the author proposed a mechanism for the enhancement of hydrogen uptake in Zry-2 materials in BWR at high burnup. A combination of a thick, porous oxide layer, of high fluence, of high irradiation flux and of low linear power -- especially for fuel rods -- are thought to be necessary conditions for the stability of metallic nickel in the near oxide layer of Zry-2 materials during additional cycles at high burnup. These metallic nickel atoms then catalyze the hydrogen absorption surface reaction at cracks and pores surfaces near the metal interface, as previously suggested, resulting in increased hydrogen pickup by the material. In turn, the results presented in this study support that the acceleration of hydrogen pickup observed in Zry-2 materials at high burnup in BWR is not likely to occur in the modern Ni-free Zr alloys.