Canada's container for permanent disposal of used nuclear fuel is a welded carbon-steel vessel coated with a millimetre-scale copper corrosion barrier (electrodeposition on the body, cold spray over the closure-weld zone), emplaced in a deep geological repository surrounded by highly compacted bentonite. The 2016 experiments of Standish et al. (Electrochimica Acta 211, 331-342) established that corrosion at a through-coating defect proceeds by galvanic coupling between anodic iron dissolution at the defect base and oxygen reduction on the surrounding copper, and that corrosion also propagates along the copper/steel interface, most likely along deposition-induced damage. The research program that study initiated left open a coherent set of coupled questions - stated by the same authors in follow-up work and consolidated in the wider program literature: how defect chemistry and redox gradients respond to ambient dissolved oxygen; how coating-process parameters and post-deposition treatments govern the interfacial damage along which corrosion propagates; how the full repository sequence - trapped oxygen, humid-air radiolysis during early resaturation, and eventual anoxic sulfide-dominated corrosion - accumulates damage toward a predictive failure model for defective containers; and how trace sulfide concentrations and their depletion kinetics control long-term corrosion in realistic bentonite porewaters. Subsequent work has quantified parts of this space (galvanic-current dependence on area ratio, chloride and oxygen in bulk solution and in bentonite slurries; trapped-oxygen and radiolysis allowances; sulfide film-growth kinetics and transport; annealing mitigation of coating heterogeneity; Monte-Carlo life-expectancy models with latent defects), but the NWMO-commissioned 2024 systematic review records that early unsaturated-phase corrosion and compounding effects remain understudied, and no experimentally validated, defect-inclusive predictive failure model spanning the repository lifecycle exists. The problem remains open.

Background

Permanent disposal of spent nuclear fuel uses deep geological repositories (DGRs): excavations hundreds of metres underground in which the waste, sealed inside corrosion-resistant containers, is isolated by engineered barriers and the surrounding rock. In the Canadian design, developed by the Nuclear Waste Management Organization (NWMO), the used fuel container (UFC) is a welded carbon-steel vessel that carries the fuel and provides mechanical strength, coated on its outside with a copper layer roughly 3 mm thick that serves as the corrosion barrier. This is far thinner than the 25-50 mm thick copper shells of the Swedish/Finnish KBS-3 design because the steel carries the structural load. Each container sits in a box of highly compacted bentonite, a swelling clay that restricts water flow and limits the transport of dissolved species to diffusion. The container is required to contain the waste for up to one million years.

Copper is chosen as the barrier because it is thermodynamically stable in oxygen-free water, so only a few species can oxidize it under repository conditions. The repository environment evolves through distinct phases. Immediately after closure the repository is warm, unsaturated, and contains trapped atmospheric oxygen, so early copper corrosion is oxic. Gamma radiation from the decaying fuel splits water and moist air (radiolysis), producing oxidants such as hydrogen peroxide and nitric acid that can accelerate corrosion while the repository is still only partially saturated. As the repository cools and groundwater resaturates the bentonite, oxygen is consumed and conditions become anoxic; from then on the significant oxidant for copper is sulfide (HS-), produced mainly by sulfate-reducing bacteria in the far-field rock and transported by diffusion through the bentonite, so long-term corrosion is limited by sulfide supply. Canadian crystalline-rock groundwaters are concentrated chloride solutions, and chloride strongly affects both copper and steel corrosion. Programs manage these processes through a corrosion allowance: the expected total corrosion depth over one million years (about 270 micrometres, with a conservative extreme bound near 1.2 mm, against the roughly 3 mm coating) computed from separate process models.

Two additive manufacturing routes produce the coating. In electrodeposition, copper is electroplated from aqueous solution onto the steel body. In cold spray, copper powder particles are accelerated to several hundred metres per second and bond to the substrate by plastic deformation without melting; cold spray is used over the closure-weld zone, the region where the steel vessel is sealed after fuel loading and which cannot be electroplated. Cold-sprayed copper retains a work-hardened microstructure and oxide inclusions inherited from the particle surfaces at particle-particle interfaces, which post-deposition heat treatment (annealing) can partly remove or redistribute.

The critical scenario for this thinner coating is a through-coating defect: a flaw that was undetected by manufacturing quality control and exposes the underlying steel to the porewater. Because the copper surface is vastly larger than the exposed steel, the couple is galvanic - the copper coating acts as the cathode where dissolved oxygen is reduced, while the small exposed steel area at the defect base acts as the anode where iron dissolves - which can greatly accelerate attack at the defect. Inside a millimetre-deep defect this coupling depletes oxygen and produces dissolved Fe2+ (ferrous iron), creating internal chemistry and redox gradients whose evolution depends on the ambient dissolved oxygen concentration and on transport through the surrounding bentonite. Standish et al. (2016), the source study for this problem, characterized cold-sprayed and electrodeposited coatings, showed that their general surface corrosion behaves similarly to wrought copper, demonstrated that corrosion at simulated millimetre-deep defects in cold-sprayed coatings is driven by galvanic coupling to oxygen reduction on the coating, and found that corrosion also propagates laterally along the copper/steel interface, which they attributed to damage inflicted during deposition. Later work frames lifetime prediction probabilistically: Monte-Carlo models over the container population (Järvine et al. 2020) treat latent coating defects and temperature, but currently only through the assumed life-limiting mechanism, sulfide-induced corrosion. The NWMO-commissioned systematic review of Harper, Brown and Amos (2024) consolidates the state of knowledge and identifies early unsaturated-phase corrosion and compounding effects of successive corrosion processes as the main remaining gaps.

Problem Statement

Establish a quantitatively validated predictive description of corrosion and failure for copper-coated steel used-fuel containers that contain through-coating defects, spanning the evolution of a deep geological repository from early oxic, partially saturated conditions - including radiolytically produced oxidants during early resaturation - to long-term anoxic, sulfide-dominated corrosion. The source study (Standish et al. 2016) established the two mechanism findings on which this program builds - galvanic coupling to oxygen reduction at a through-coating defect, and corrosion propagation along the copper/steel interface attributed to deposition-induced damage; the open questions of this research thread, stated by the same authors in follow-up work (Standish et al. 2018) and consolidated in the wider program literature, require four linked components: (i) quantification of how the chemistry and redox gradients inside a millimetre-deep coating defect - internal oxygen depletion, Fe2+ production and transport, and defect solution chemistry - evolve as the ambient dissolved oxygen concentration varies, including under the mass-transport constraints imposed by the surrounding compacted bentonite; (ii) determination of how the distribution of interfacial damage and the propensity for corrosion to propagate along the copper-steel interface depend on coating process parameters (notably cold spray particle-impact conditions) and post-deposition treatments such as annealing, and determination of whether and how process conditions and post-deposition treatments mitigate or prevent the deleterious interfacial damage; (iii) extension of the defect studies to longer times and the complete sequence of repository-relevant exposure regimes, integrated into a predictive failure model for defective containers; and (iv) systematic assessment of how trace sulfide concentrations and their depletion kinetics control long-term corrosion rates and corrosion-product formation in laboratory exposure media versus realistic bentonite porewaters. The components share a single objective: the predictive failure model of (iii), for which (i), (ii) and (iv) supply the defect-chemistry, damage-state, and long-term-rate inputs. Each of (i), (ii) and (iv) must be resolved by new work - the cited literature may inform these components but cannot substitute for them - and is independently assessable and refutable; resolving any one of them, or any proper subset, does not by itself resolve the task, which is resolved only by the integrated, experimentally validated model of (iii) that they feed. This defect-inclusive predictive model is validated against experimental data at the experimentally accessible stages of the repository sequence - the early oxic, partially saturated phase, including radiolytically produced oxidants during early resaturation, and the transition to anoxic, sulfide-dominated corrosion - and demonstrably predicts the accumulation of corrosion damage - and the failure, defined here as corrosion penetration of the copper barrier and the underlying steel wall (loss of containment) - of defective containers across the remainder of the repository lifecycle, extrapolated from those validated stages. Validation means that the model reproduces the measured defect-corrosion behaviour and damage accumulation over the accessible stages within the uncertainty of the source data; the full repository lifetime, which cannot be reproduced experimentally, is consequently predicted by extrapolation under explicitly quantified uncertainty rather than measured directly.

References

  1. Thalia E. Standish, Jian Chen, Rebecca Jacklin, Pellump Jakupi, Sridhar Ramamurthy, Dmitrij Zagidulin, Peter G. Keech, David W. Shoesmith (2016). Corrosion of copper-coated steel high level nuclear waste containers under permanent disposal conditions. Electrochimica Acta 211, 331-342. DOI: 10.1016/j.electacta.2016.05.135. https://doi.org/10.1016/j.electacta.2016.05.135
  2. Thalia E. Standish, Dmitrij Zagidulin, Sridhar Ramamurthy, Peter G. Keech, David W. Shoesmith, James J. Noël (2018). Synchrotron-based micro-CT investigation of oxic corrosion of copper-coated carbon steel for potential use in a deep geological repository for used nuclear fuel. Geosciences 8, 360. DOI: 10.3390/geosciences8100360. https://doi.org/10.3390/geosciences8100360
  3. Thalia E. Standish, Lindsay J. Braithwaite, David W. Shoesmith, James J. Noël (2019). Influence of area ratio and chloride concentration on the galvanic coupling of copper and carbon steel. Journal of The Electrochemical Society 166, C3448-C3455. DOI: 10.1149/2.0521911JES. https://doi.org/10.1149/2.0521911JES
  4. Lindsay Braithwaite, Katarina Albrechtas, Dmitrij Zagidulin, Mehran Behazin, David Shoesmith, James J. Noël (2022). Galvanic coupling of copper and carbon steel in the presence of bentonite clay and chloride. Journal of The Electrochemical Society 169, 051502. DOI: 10.1149/1945-7111/ac5ff2. https://doi.org/10.1149/1945-7111/ac5ff2
  5. Tadeja Kosec, Miha Hren, Klara Prijatelj, Bojan Zajec, Nina Gartner, Andraž Legat (2023). Monitoring the galvanic corrosion of copper–steel coupling in bentonite slurry during the early oxic phase using coupled multielectrode arrays. Materials and Corrosion 74, 1656-1664. DOI: 10.1002/maco.202213689. https://doi.org/10.1002/maco.202213689
  6. Ryan P. Morco, Jiju M. Joseph, David S. Hall, Chantal Medri, David W. Shoesmith, J. Clara Wren (2017). Modelling of radiolytic production of HNO3 relevant to corrosion of a used fuel container in deep geologic repository environments. Corrosion Engineering, Science and Technology 52, 141-147. DOI: 10.1080/1478422X.2017.1340227. https://doi.org/10.1080/1478422X.2017.1340227
  7. David S. Hall, Thalia E. Standish, Mehran Behazin, Peter G. Keech (2018). Corrosion of copper-coated used nuclear fuel containers due to oxygen trapped in a Canadian deep geological repository. Corrosion Engineering, Science and Technology 53, 309-315. DOI: 10.1080/1478422X.2018.1463009. https://doi.org/10.1080/1478422X.2018.1463009
  8. Elham Salehi Alaei, Mengnan Guo, Jian Chen, Mehran Behazin, Erik Bergendal, Christina Lilja, David W. Shoesmith, James J. Noël (2023). The transition from used fuel container corrosion under oxic conditions to corrosion in an anoxic environment. Materials and Corrosion 74, 1690-1706. DOI: 10.1002/maco.202313757. https://doi.org/10.1002/maco.202313757
  9. David S. Hall, Mehran Behazin, W. Jeffrey Binns, Peter G. Keech (2021). An evaluation of corrosion processes affecting copper-coated nuclear waste containers in a deep geological repository. Progress in Materials Science 118, 100766. DOI: 10.1016/j.pmatsci.2020.100766. https://doi.org/10.1016/j.pmatsci.2020.100766
  10. Wilfred Jeffrey Binns, Mehran Behazin, Scott Briggs, Peter G. Keech (2023). An overview of the Canadian nuclear waste corrosion program. Materials and Corrosion 74, 1580-1586. DOI: 10.1002/maco.202313763. https://doi.org/10.1002/maco.202313763
  11. Colleen O. Harper, Julie L. Brown, Richard T. Amos (2024). Corrosion processes affecting copper-coated used fuel containers for the disposal of spent nuclear fuel: critical review of the state-of-knowledge. npj Materials Degradation 8, 124. DOI: 10.1038/s41529-024-00540-z. https://doi.org/10.1038/s41529-024-00540-z
  12. Colleen O. Harper, Julie L. Brown, Richard T. Amos (2026). The impact of bentonite on copper corrosion early in deep geological repository lifecycles. Journal of Nuclear Materials 628, 156651. DOI: 10.1016/j.jnucmat.2026.156651. https://doi.org/10.1016/j.jnucmat.2026.156651
  13. Allan K. Järvine, Alan G. Murchison, Peter G. Keech, Mahesh D. Pandey (2019). A probabilistic model for estimating the life expectancy of used nuclear fuel containers in a Canadian geological repository: Baseline model. Nuclear Engineering and Design 352, 110202. DOI: 10.1016/j.nucengdes.2019.110202. https://doi.org/10.1016/j.nucengdes.2019.110202
  14. Allan K. Järvine, Alan G. Murchison, Peter G. Keech, Mahesh D. Pandey (2020). A probabilistic model for estimating the life expectancy of used nuclear fuel containers in a Canadian geological repository: Effects of latent defects and repository temperature. Nuclear Technology 206, 1036-1058. DOI: 10.1080/00295450.2019.1700730. https://doi.org/10.1080/00295450.2019.1700730
  15. Erik P. Kremer (2017). Durability of the Canadian used fuel container. Corrosion Engineering, Science and Technology 52, 173-177. DOI: 10.1080/1478422X.2017.1330024. https://doi.org/10.1080/1478422X.2017.1330024
  16. Mengnan Guo, Jian Chen, Taylor Martino, Christina Lilja, Johannes A. Johansson, Mehran Behazin, Wilfred J. Binns, Peter G. Keech, James J. Noël, David W. Shoesmith (2021). The nature of the copper sulfide film grown on copper in aqueous sulfide and chloride solutions. Materials and Corrosion 72, 300-307. DOI: 10.1002/maco.202011710. https://doi.org/10.1002/maco.202011710
  17. Jian Chen, Xinran Pan, Taylor Martino, Christina Lilja, Mehran Behazin, Wilfred J. Binns, Peter G. Keech, James J. Noël, David W. Shoesmith (2023). The effects of chloride and sulphate on the growth of sulphide films on copper in anoxic sulphide solutions. Materials and Corrosion 74, 1665-1676. DOI: 10.1002/maco.202313766. https://doi.org/10.1002/maco.202313766
  18. S. Ramamurthy, J. Chen, D. Zagidulin, J.D. Henderson, C. Lilja, E. Bergendal, M. Behazin, N. Diomidis, P.G. Keech, J.J. Noël, D.W. Shoesmith (2024). The influences of traces of oxygen and sulfide on the corrosion of copper in concentrated chloride solutions. Corrosion Science 233, 112047. DOI: 10.1016/j.corsci.2024.112047. https://doi.org/10.1016/j.corsci.2024.112047
  19. Nicholas A. Senior, Taylor Martino, Jeff Binns, Peter Keech (2021). The anoxic corrosion behaviour of copper in the presence of chloride and sulphide. Materials and Corrosion 72, 282-292. DOI: 10.1002/maco.202011783. https://doi.org/10.1002/maco.202011783
  20. Scott Briggs, Jennifer McKelvie, Brent Sleep, Magdalena Krol (2017). Multi-dimensional transport modelling of corrosive agents through a bentonite buffer in a Canadian deep geological repository. Science of the Total Environment 599-600, 348-354. DOI: 10.1016/j.scitotenv.2017.04.242. https://doi.org/10.1016/j.scitotenv.2017.04.242
  21. Tarek L. Rashwan, Md. Abdullah Asad, Ian L. Molnar, Mehran Behazin, Peter G. Keech, Magdalena M. Krol (2022). Exploring the governing transport mechanisms of corrosive agents in a Canadian deep geological repository. Science of the Total Environment 828, 153944. DOI: 10.1016/j.scitotenv.2022.153944. https://doi.org/10.1016/j.scitotenv.2022.153944
  22. Sifat Azad Papry, Tarek L. Rashwan, Pulin K. Mondal, Mehran Behazin, Peter G. Keech, Magdalena M. Krol (2023). Investigating bisulfide sorption onto bentonite through laboratory batch experiments. Applied Geochemistry 152, 105626. DOI: 10.1016/j.apgeochem.2023.105626. https://doi.org/10.1016/j.apgeochem.2023.105626
  23. Jason Tam, Weiwei Li, Bosco Yu, Dominique Poirier, Jean-Gabriel Legoux, Peter Lin, Gino Palumbo, Jason D. Giallonardo, Uwe Erb (2020). Reducing complex microstructural heterogeneity in electrodeposited and cold sprayed copper coating junctions. Surface and Coatings Technology 404, 126479. DOI: 10.1016/j.surfcoat.2020.126479. https://doi.org/10.1016/j.surfcoat.2020.126479
  24. Jason Tam, Bosco Yu, Weiwei Li, Dominique Poirier, Jean-Gabriel Legoux, Jason D. Giallonardo, Jane Howe, Uwe Erb (2022). The effect of annealing on trapped copper oxides in particle-particle interfaces of cold-sprayed Cu coatings. Scripta Materialia 208, 114333. DOI: 10.1016/j.scriptamat.2021.114333. https://doi.org/10.1016/j.scriptamat.2021.114333
  25. Liyang Zheng, Dominique Poirier, Jean-Gabriel Legoux, Jason D. Giallonardo, Jane Y. Howe, Uwe Erb (2024). Microstructural characterization of interfaces between cold sprayed copper and electrodeposited copper corrosion barrier coatings for used nuclear fuel containers. Surface and Coatings Technology 494, 131396. DOI: 10.1016/j.surfcoat.2024.131396. https://doi.org/10.1016/j.surfcoat.2024.131396
  26. Liyang Zheng, Jason D. Giallonardo, Mehran Behazin, Jane Y. Howe, Uwe Erb (2025). Microstructure and initial corrosion response of electrodeposited/cold sprayed copper coating interface regions on used nuclear fuel containers. Corrosion Science 255, 113103. DOI: 10.1016/j.corsci.2025.113103. https://doi.org/10.1016/j.corsci.2025.113103
  27. Satoru Suzuki, Yusuke Ogawa, Jason Giallonardo, Peter G. Keech (2021). The design of copper-coating overpack for the high-level radioactive waste disposal concept in Japan. Materials and Corrosion 72, 94-106. DOI: 10.1002/maco.202011855. https://doi.org/10.1002/maco.202011855
  28. Xiangju Liu, Nazhen Liu, James J. Noël, David W. Shoesmith, Jian Chen, Baorong Hou (2023). The influence of hydrogen permeation on the protection performance of the Cu coating of nuclear waste containers. Corrosion Science 221, 111314. DOI: 10.1016/j.corsci.2023.111314. https://doi.org/10.1016/j.corsci.2023.111314
  29. Fraser King, Mehran Behazin, Peter Keech (2023). Natural and archaeological analogues for corrosion prediction in nuclear waste systems. Materials and Corrosion 74, 1811-1822. DOI: 10.1002/maco.202313764. https://doi.org/10.1002/maco.202313764
  30. Yanxin Qiao, Tianyu Wang, Zhilin Chen, Jun Wang, Chengtao Li, Jian Chen (2024). Corrosion techniques and strategies for used fuel containers with copper corrosion barriers under deep geological disposal conditions: a review. International Journal of Minerals, Metallurgy and Materials 31, 2582-2606. DOI: 10.1007/s12613-024-2949-x. https://doi.org/10.1007/s12613-024-2949-x