1.1 Reactor materials
Current light water reactor (LWR) core internal component materials include several types of corrosion-resistant materials such as type 304 and 316 stainless steels (SS) and nickel-based Inconel 600 and 690 and alloy X-750. The cladding of the fuel in LWRs is commonly a zirconium-based alloy such as Zircaloy-2, Zircaloy-4, Zirlo, or M5. Table 1 shows the nominal compositions of the most common structural alloys inside the reactors. These alloys are designed to resist the high-temperature aqueous environments inside the reactors for several decades. The alloys could be austenitic SS or austenitic nickel-based alloys, and they generally contain chromium to provide protection against corrosion (Ford et al., 2006; Scott and Combrade, 2006).
The most common LWRs are boiling water reactors (BWR) and pressurized water reactors (PWR), and the environments are high-purity water near 300°C. The BWR environment is generally at 270–288°C, and its original water chemistry was known to be normal water chemistry (NWC) with high oxidants (e.g. oxygen and hydrogen peroxide O2 and H2O2), but later, hydrogen was injected – and it was called hydrogen water chemistry (HWC) to reduce the oxidizing power of the reactor water and thus decrease the corrosion potential of the alloys to mitigate the occurrence of stress corrosion cracking (SCC). The PWR primary water is generally at a higher temperature of ~315°C and may contain lithium and boron plus excess hydrogen gas to decrease the corrosion potential of the alloys for SCC mitigation purposes (Ford et al., 2006; Scott and Combrade, 2006).
1.2 Composition and microstructure of oxides formed on SS
The composition and microstructure of oxides formed on SS in reactor environments were studied extensively because of their importance in explaining the susceptibility of SS to SCC. In general, it is reported that oxide films consist of more than one layer both under oxidizing conditions (oxygen or hydrogen peroxide) and under reducing conditions (excess hydrogen gas). The inner layer generally consists of a finer grained oxide enriched in chromium. The outer coarse grained layer is enriched in iron. Kim performed detailed characterization of oxides formed on both type 304 and 316 SS in 288°C water containing either oxygen, hydrogen peroxide, or hydrogen (Kim, 1995, 1999). Type 316 SS coupons were exposed to 288°C water for 7 months changing the conditions from NWC (200 ppb oxygen) to HWC (20 ppb hydrogen) every 2 weeks (Kim, 1995). The corrosion potential of type 316SS changed from approximately −450 mV SHE under HWC to approximately −150 mV under NWC for the entire immersion time. After 1 month of exposure, the total oxide thickness remained constant between 500 and 600 nm (Kim, 1995). In general, Kim (1995) reported that the oxide consisted of two layers, an outer layer with larger oxide particles mainly of magnetite (Fe3O4) and a finer grain structured chromium enriched inner layer. It was reported that cycling the water chemistries between NWC and HWC altered the particle distribution in the outer oxide layer; the closely packed inner oxide layer was not affected by the cycling of water chemistries (Kim, 1995). Kim (1999) also studied the characteristics of the oxide layers formed on type 304 SS by performing immersion tests for 2 weeks at 288°C under (1) NWC (200 ppb oxygen), (2) 10–200 ppb of hydrogen peroxide, and (3) HWC or 150 ppb hydrogen. It was reported that a thicker oxide film formed under oxygen and hydrogen peroxide conditions (approximately 1-μm thick) than under hydrogen conditions (approximately 0.8-μm thick). The oxide film consisted of two layers, an external one based on the α and γ Fe3O4 and an inner fine-grained layer enriched in chromium (Kim, 1999).
Kuang et al. (2010) also studied the characteristics of the oxide films formed on type 304 SS in oxygenated water (3 ppm oxygen) at 290°C. They reported a two-layer oxide, comprised of an external loosely packed or disorderly layer probably formed by precipitation and an internal layer containing nano crystalline spinel oxides enriched in chromium (Kuang et al., 2010). Ziemniak and Hanson (2002) studied the oxidation behavior of type 304 SS for up to 10,000 h in hydrogenated (45 scm³/kg) water at 260°C. They also reported a double layer oxide, an outer spinel ferrite based layer, and an inner spinel chromite-based layer (Ziemniak and Hanson, 2002). The outer layer could have crystals in the order of 5–10-μm size, while the inner layer was only 1–2-μm thick containing crystals in the order of 5–10 nm in size (Ziemniak and Hanson, 2002).
Terachi and Arioka (2006) studied the properties of oxide films formed on type 316 SS after 380-h immersion in water containing 500 ppm boron (B)+2 ppm lithium (Li) at 320°C both under oxygen (8 ppm dissolved oxygen) and hydrogen (2.75 ppm dissolved hydrogen) conditions. The reported oxide film thicknesses are between 0.2 and 0.4 μm, being slightly thicker under hydrogen conditions. Under hydrogen conditions, the film was a double layer with magnetite in the outer layer and fine-grained inner layer composed of chromite, which acted as a diffusion layer (Terachi and Arioka, 2006). In the oxygen environment, the oxide was also a double layer with hematite outer layer and spinel inner layer with chromium concentrated at the boundary between the metal and the oxide (Terachi and Arioka, 2006). These results agree well with previous findings by Da Cunha Belo et al. (1998), who performed immersion tests of type 316 SS for 2000 h in PWR water environment at 350°C containing 1000 ppm B and 2 ppm Li with 27 ppm dissolved hydrogen. The total oxide thickness was approximately 200 nm, consisting of an inner fine grained chromium oxide layer and an outer layer formed by iron and nickel spinel oxides with larger grains (Da Cunha Belo et al., 1998).
1.3 Composition and microstructure of oxides formed on nickel alloys
Morton et al. (2007) tested coupons of Alloy 600 and X-750 for up to 5000 h in high-purity water containing 18 cc hydrogen gas/kg water at 288°C and 60 cc hydrogen/kg water at 338°C. After the 5000-h immersion tests, they reported for Alloy 600 a total oxide thickness of 0.16 μm at 288°C and 1.48 μm at 338°C (Morton et al., 2007). The oxide was formed by two layers, a thicker ferrite rich outer layer and a protective thinner inner oxide layer comprised of fine-grained chromium-rich oxide (Morton et al., 2007). Terachi et al. (2003) studied the composition of the oxide film formed on Alloy 600 at 320°C after 1000-h immersion in PWR type primary water containing 500 ppm boron and 2 ppm Li. They varied the dissolved hydrogen content from nil to 4 ppm (Terachi et al., 2003). For a dissolved hydrogen content of 2.75 and 4 ppm, they reported a dual oxide layer, with an external spinel layer and a thinner chromium-rich internal layer (Terachi et al., 2003).
Huang et al. (2009, 2010) studied the characteristics of the oxide layer formed on Alloy 690 during electrochemical tests in 0.15 M sodium sulfate (Na2SO4) and 0.04 M sodium hydroxide (NaOH) solutions up to 300°C. They reported a dual layer spinel oxide film approximately 200-nm thick at 300°C (Huang et al., 2010). The outer layer was Ni-Fe spinel and Ni hydroxide, and the inner layer was fine-grained and chromium-rich (Huang et al., 2010).
Studies reported in the literature show that austenitic chromium containing alloys including SS and nickel based are corrosion resistant in both oxygenated and hydrogenated environments that could represent BWR and PWR water chemistries. These alloys form a double layer oxide, with a coarser outer layer formed mainly by iron and nickel oxides and a thinner finer grained compact layer, which is chromium-rich and which is the actual barrier for corrosion.
1.4 FeCrAl alloys cladding for UO2 fuel
After the tsunami disaster in northeast Japan in March 2011, the US Department of Energy and the commercial fuel vendors have teamed up to develop an accident-tolerant fuel, which will resist catastrophic events such as the ones in Fukushima (Carmack and Goldner, 2014). In the Fukushima accident, zirconium reacted rapidly with water and steam above 400°C forming combustible hydrogen gas and releasing a large amount of heat of oxidation. One of the concepts under development is to use iron-chromium-aluminum alloys such as APMT (Table 1) as cladding for the current uranium dioxide (UO2) fuel (Terrani et al., 2014; Rebak, 2015). APMT is an advanced powder metallurgy, dispersion strengthened, ferritic Fe-Cr-Al-Mo alloy, which is used at tube temperatures up to 1250°C. APMT forms an excellent, non-scaling surface oxide, which gives good protection in most furnace environments, i.e. oxidizing, sulfurizing, and carburizing gas, as well as against scaling deposits of carbon, ash, etc. When exposed to high-temperature (>1200°C) steam, APMT forms a compact protective layer of alumina on its surface (Rebak and Kim, 2016). The alumina layer forms initially underneath the chromia layer, and after the chromia layer volatilizes by reaction with steam, the alumina layer protects the metal from further attack by the steam. Therefore, APMT was selected as a good candidate for the nuclear fuel cladding. As APMT has never been used in LWR before, the objective of this work is to investigate the general corrosion and passive film formation of APMT and other candidate materials under BWR and PWR normal operation conditions water chemistries.
2 Materials and methods
Table 2 shows the alloys tested and the testing conditions. Newer materials were investigated in parallel with current zirconium-based alloy such as Zircaloy-2. Coupons of the materials were tested for 1 year to determine general corrosion under laboratory simulated normal operation conditions of commercial LWRs. Four sets of autoclaves were used (Table 2) to simulate BWR and PWR conditions. Only the results concerning APMT are reported here. The degradation of the immersion coupons was evaluated by weight (mass) change, standard metallographic procedures, and surface analysis techniques. In each of the autoclaves, all the alloys were exposed at the same time, side by side with each other. In each autoclave, the water was re-circulated at a flow rate of 100 cm³/min and reconditioned (filtered) before reentering the autoclaves, that is, the coupons were constantly exposed to refreshed water, there was no enrichment of the water in corrosion products. Once removed from the autoclaves, the coupon surface oxide morphology was examined using secondary electron imaging in a focused ion beam-scanning electron microscope workstation (FIB-SEM). The FIB-SEM also allowed examination of the oxide layer in cross-section. This was accomplished by sputtering a trench into the sample surface using the ion beam. The trench face reveals the oxide in cross-section and is imaged using backscattered electrons. The FIB-SEM is also able to produce cross-section samples for examination in a transmission electron microscope (TEM). Examination in a TEM requires the sample to be electron transparent, as the incident electrons must pass through the specimen. This is accomplished by using the FIB-SEM to sputter a double trench into the surface of the coupon. Between the two trenches, a thin lamella is left behind. This lamella is about 30-μm long, 1-μm thick, and 10-μm deep. The lamella is removed from the coupon using an omniprobe micromanipulator and then reattached to a post on a 3-mm-diameter copper half-grid. Once attached to the half grid, the lamella was further thinned to a thickness of about 100 nm. This was then placed into an FEI Tecnai Osiris TEM operated at 200-kV accelerating voltage for examination. The TEM can be operated using a stationary electron probe, or the electron probe can be reduced to a very small probe size, about 1 nm in diameter, and then scanned across the surface of the specimen. This scanning transmission electron (STEM) mode is used in conjunction with an integrated energy dispersive spectrometer to produce X-ray maps (more specifically spectral images). These maps were used to determine the elemental distribution in the near surface material and the surface oxide(s).
Prior to being placed in the SEM, some samples were partially coated by a sputtered platinum (Pt) layer approximately 50-nm thick to reduce charging and protect the sample surface during ion beam exposure. The sample was then placed in the FIB-SEM instrument. Areas not coated by Pt were examined by the electron beam to document the oxide morphology. For other samples, a microscopic layer of Pt-based material was deposited in situ (electron beam induced decomposition of Pt-based precursor gas) on the sample surface to protect the sample surface during subsequent ion beam exposure. A thicker, microscopic layer of Pt-based material was then deposited on all the samples (ion beam induced decomposition of Pt-based precursor gas). After surface protection, 90° cross-sections approximately 20–30 μm wide were prepared using the ion beam. The resulting cross-section was examined using the electron beam, and secondary electron images were recorded using the immersion lens mode.
3.1 Oxides formed on APMT under BWR normal conditions with 1000–2000 ppb oxygen at 288°C
Table 3 shows thickness data obtained from FIB-SEM cross-sections from coupons of eight alloys exposed to simulated BWR, normal water chemistry (NWC) (2000 ppb O2) at 288°C for 1 year in autoclave S-14 (Table 2) (Ellis and Rebak, 2016). The range listed in Table 3 is from five measurements along 20–30-μm-wide cross-sections. Zircaloy-2 had the thickest oxide, followed by T91, a ferritic steel with only 9% Cr. The thinnest oxides corresponded to HT9 and NFA, a nano ferritic steel. APMT had the third thinnest oxide in the aerated high-temperature water. The remaining parts of the manuscript will be dedicated only to the oxides formed on APMT for the four testing conditions listed in Table 2.
Figure 1 shows the surface morphology and oxide thickness for APMT coupon exposed to autoclave S-14 in Table 2 for 1 year. The surface of the APMT coupon shows the presence of crystals approximately 100 nm in size. Figure 2 shows the cross-sectional spectral images of a double layer oxide comprising external oxide particulates (mostly iron oxide crystals approximately 200–300-nm thick) and a continuous chromium rich oxide layer (approximately 20-nm thick) between the thicker external oxide particulates and the metal substrate. It is also interesting to note large oxide particulates enriched in chromium (possibly Fe-chromite spinel or FeCr2O4). This observation is in a good agreement with the previous literature (Kim, 1995, 1999). Oxide analysis showed that the oxide layer did not contain either aluminum or molybdenum (Figure 2), which is present in APMT (Table 1).
Figure 3 shows a tilted surface view and a cross-section of the oxides formed on APMT coupon exposed in autoclave S-6 (Table 2) to high-purity water+1000 ppb oxygen at 288°C for 1 year. The APMT coupon had a compact continuous uniform oxide thickness of approximately 150–180 nm. Figure 4 shows a higher magnification view of the oxide, showing that it is uniform, compact, and continuous. The inner oxide layer is fine grained (<10 nm), and the outer layer has coarser grains of approximately 25 to 50 nm in size. Figure 5 shows the spectral images of the oxide with its double layer comprising a thin (10–15-nm thick) inner oxide layer enriched in chromium.
3.2 Oxides formed on APMT under BWR and PWR hydrogenated conditions at 288°C and 330°C
Figure 6 shows a tilted surface view and a cross-section of the oxides formed on APMT coupon exposed in autoclave S-5 (Table 2) to high-purity water+300 ppb hydrogen at 288°C for 1 year. Under hydrogen conditions, very few surface crystals are observed as compared to oxygen conditions (Figure 3). The APMT coupon had a compact continuous uniform oxide thickness of approximately 100–120 nm. Figure 7 shows spectral images of the surface films, in which appeared single-layered oxide rich in chromium. Iron does not seem to be present on the oxide film.
Figure 8 shows a tilted surface view and a cross-section of the oxides formed on APMT coupon exposed in autoclave S-2 (Table 2) to high-purity water+3.75 ppm hydrogen at 330°C for 1 year. Under hydrogen conditions, fewer surface crystals are observed as compared to oxygen conditions (Figure 3). There are more crystals at 330°C (Figure 8) than at 288°C (Figure 6). The APMT coupon had a compact continuous uniform oxide thickness of approximately 10 to 100 nm. Figure 9 shows spectral images of the surface films, which appeared single layered comprising oxide rich in chromium. Similarly to the hydrogenated water at 288°C (Figure 7), the oxide layer that formed at 330°C appears to contain little or no Fe (Figure 9).
Immersion tests of coupons of APMT in environments simulating BWR and PWR water chemistries containing excess hydrogen or oxygen showed that the material always formed chromium-rich protective oxide layer on its surface (Table 4). In oxygen-rich environments, two layers are discernible, an external iron-rich oxide layer and a thinner internal oxide layer, which is rich in chromium. In hydrogen-rich environments, only one chromium-rich oxide layer is present on the surface (Table 4). These finding show that cladding candidate APMT, which has approximately 20% Cr in solution, behaved in a similar way as current nuclear materials such as SS type 304 and 316 and nickel-based alloys such as Alloy 600 (Table 1).
Previous results by Terrani et al. (2016) using experimental FeCrAl alloys showed similar results of the currently reported for APMT despite the different Cr and Al concentration. It was shown that in oxygenated environments, the oxide contained both Cr and Fe and, in hydrogenated environments, the oxide contained little or no Fe but was rich in Cr and may have contained some aluminum (Terrani et al., 2016).
SS and nickel alloys containing chromium have been used in reactors for over 60 years, and they perform well in resisting corrosion in the near 300°C water, with excess oxygen or excess hydrogen. Their resistance to corrosion can be attributed to a protective chromium oxide layer that develops on the surface and serves as a barrier for further oxidation. As in the current results, a similar chromium oxide barrier was found to form on APMT; it is expected that APMT will also resist corrosion in BWR and PWR environments for the required time for fuel rods application (approximately 10 years).
The accident-tolerant fuel program aims at developing a fuel for current LWRs that would be more resistant to accident scenarios (such as the one in Fukushima) than the traditional system of zirconium alloys with uranium dioxide pellet (Bragg-Sitton, 2014; Carmack and Goldner, 2014; Terrani et al., 2014; Rebak, 2015; Rebak et al., 2016). The results presented here are only a small portion (characterization of the oxide formed by general corrosion under normal operation conditions) of the entire project for nuclear fuel characterization. Other areas of current studies include the behavior of materials under accident conditions, neutronics, thermal hydraulics, behavior of materials under neutron irradiation conditions, FeCrAl tube fabrication, etc.
5 Summary and conclusions
Coupons of Alloy APMT (Fe+22Cr+5Al+3Mo) were immersed in BWR and PWR simulated conditions for 1 year. Environments tested included pure water at 330°C with excess hydrogen and pure water at 288°C with excess hydrogen or excess oxygen.
Surface analysis showed that APMT always developed a protective chromium-rich oxide film on the surface, which was a barrier to corrosion.
Under oxygenated conditions, the oxide formed on APMT included two layers, an external layer rich in iron (and chromium) and an internal thinner layer enriched in chromium.
Under hydrogenated conditions, APMT developed a single oxide layer rich in chromium.
Alloy APMT had a similar corrosion protection mechanism in high-temperature reactor environments than the well-documented cases of SS type 304 and 316 and nickel-based alloys such as Alloy 600.
The funding support from Kelly Fletcher and Steven J. Duclos from GE Global Research is gratefully acknowledged. The technical expertise of R. J. Blair, Frank Wagenbaugh, Dustin D. Ellis, and Timothy B. Jurewicz is gratefully acknowledged.
Bragg-Sitton SM. Development of advanced accident-tolerant fuels for commercial LWRs. Nuclear News March 2014: 83–91. Google Scholar
Da Cunha Belo M, Walls M, Hakiki NE, Corset J, Picquenard E, Sagon G, Noel D. Composition, structure and properties of the oxide films formed on the stainless steel 316L in a primary type PWR environment. Corros Sci 1998; 40: 447–463. CrossrefGoogle Scholar
Ellis DD, Rebak RB. Passivation characteristics of ferritic stainless materials in simulated reactor environments. Houston, TX: Paper C2016-7452, Corrosion/2016, NACE International, 2016. Google Scholar
Ford FP, Gordon BM, Horn RM. Corrosion in boiling water reactors. ASM Handbook, Vol. 13C, Cramer SD, Covino BS Jr., editors, Corrosion: Environments and Industries, 2006: p. 341–361, DOI: 10.1361/asmhba0004145. Google Scholar
Huang J., Wu X, Han EH. Influence of pH on electrochemical properties of passive films formed on Alloy 690 in high temperature aqueous environments. Corros Sci 2009; 51: 2976–2982. Web of ScienceCrossrefGoogle Scholar
Huang J, Wu X, Han EH. Electrochemical properties and growth mechanism of passive films on Alloy 690 in high-temperature alkaline environments. Corros Sci 2010; 52: 3444–3452. CrossrefWeb of ScienceGoogle Scholar
Kuang W, Han EH, Wu X, Rao J. Microstructural characteristics of the oxide scale formed on 304 stainless steel in oxygenated high temperature water. Corros Sci 2010; 52: 3654–3660. CrossrefWeb of ScienceGoogle Scholar
Morton D, Lewis N, Hanson M, Rice S, Sander P. Nickel alloy primary water bulk surface and SCC corrosion film analytical characterization and SCC mechanistic implications, In: 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, British Columbia, 19–23 August 2007. Google Scholar
Rebak RB, Kim YJ. Hydrogen diffusion in FeCrAl alloys for light water reactors cladding applications. Paper PVP2016-63164, 2016 ASME PVP Conference, MF-7 Materials and Technologies for Nuclear Power Plants, 17–21 July 2016, Vancouver, BC. Google Scholar
Rebak RB, Terrani KA, Gassmann WP, Williams JB, Ledford KL. Improving nuclear power plant safety with FeCrAl alloy fuel cladding. MRS Fall 2016 Meeting, Symposium ES5: Materials Research and Design for A Nuclear Renaissance, Boston 28 Nov to 02 Dec 2016. Google Scholar
Scott PM, Combrade P. Corrosion in pressurized water reactors. In: Cramer SD, Covino BS Jr., editors, ASM Handbook, Vol. 13C, Corrosion: Environments and Industries, Metals Park, Ohio, 2006: 362–385, DOI: 10.1361/asmhba0004146. CrossrefGoogle Scholar
Terachi T, Totsuka N, Yamada T, Nakagawa T, Deguchi H, Horiuchi M, Oshitani M. Influence of dissolved hydrogen on structure of oxide film on alloy 600 formed in primary water of pressurized water reactors. J Nuclear Sci Technol 2003; 40: 509–516. CrossrefGoogle Scholar
Terachi T, Arioka K. Characterization of oxide film behaviors on 316 stainless steels in high-temperature water – influence of hydrogen and oxygen considerations for initiation of SCC. Houston, TX: Paper 06608, Corrosion/2006 Conference and Exposition, NACE International, 2006. Google Scholar
Terrani KA, Pint BA, Kim YJ, Unocic KA, Yang Y, Silva CM, Meyer HM III, Rebak RB. Uniform corrosion of FeCrAl alloys in LWR coolant environments. J Nuclear Mater 2016; 479: 36–47. Web of ScienceCrossrefGoogle Scholar
About the article
Published Online: 2017-06-28
Published in Print: 2017-08-28
Citation Information: Corrosion Reviews, Volume 35, Issue 3, Pages 177–188, ISSN (Online) 2191-0316, ISSN (Print) 0334-6005, DOI: https://doi.org/10.1515/corrrev-2017-0011.
©2017 Raul B. Rebak et al., published by De Gruyter, Berlin/Boston. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. BY-NC-ND 4.0