Containers for the deep geological disposal of high-level radioactive waste (HLW) and spent fuel (SF) are the only engineered barrier expected to provide complete radionuclide containment, yet their lifetimes — tens of thousands to one million years — must be predicted far beyond any direct observation. A 2014 review by Fraser King (JOM 66(3):526–537) identified as its most important unresolved issue the absence of a fully satisfactory general lifetime model that couples mechanical and corrosion degradation for containers whose material properties evolve through wall loss, defect growth, or hydrogen uptake, together with the unresolved problem of demonstrating that any long-term prediction is robust over repository timescales. Subsequent work developed fracture-mechanics-based coupled assessments (failure-assessment-diagram trajectories and a probabilistic canister breaching model) for specific carbon steel designs, but the later literature (King et al. 2024 two-part review; the 2024 EURAD ConCorD state-of-the-art report; King 2026) states that validating performance-assessment models remains difficult or impossible, and that the component prediction problems — localized-corrosion extrapolation with stifling and environmental evolution, hydrogen pickup, microbiologically influenced corrosion under biofilms, and the definition and spread of failure times in probabilistic safety assessment — remain open. The problem posed is to develop and defend a general coupled mechanical–corrosion lifetime-prediction framework whose extrapolations over repository timescales can be validated.

Background

In deep geological disposal, high-level radioactive waste (HLW) and spent nuclear fuel (SF) are sealed in metallic containers (also called canisters or overpacks) and emplaced in a repository several hundred metres underground, typically surrounded by a compacted clay (bentonite) buffer. The container is the only component of the engineered and natural barrier system that provides complete containment, so repository safety cases lean heavily on a defensible prediction of the container lifetime, which for some materials is specified as up to 10610^{6} years.

Candidate container materials divide into 'active' materials that corrode more or less uniformly — copper and carbon steel — and 'passive' alloys (titanium and Ni–Cr–Mo alloys such as Alloy 22) that resist corrosion but are susceptible to localized attack. Traditionally, designers treated mechanical and corrosion failure separately: the wall thickness was fixed so that the container withstands external loads (hydrostatic and lithostatic pressure, buffer swelling, seismic shear, accidental drops) with the remainder serving as a corrosion allowance, a sacrificial thickness consumed by general corrosion. This separation is untenable when degradation is time-dependent: corrosion thinning reduces load-bearing capacity, absorbed hydrogen (produced by anaerobic corrosion) lowers fracture toughness, and flaws — especially in the closure weld — can grow by stress-corrosion cracking (SCC, cracking driven by the combined action of tensile stress and a corrosive environment) or hydrogen-induced cracking (HIC).

The degradation modes that a lifetime model must cover include: uniform corrosion, often predicted with reactive-transport models that couple electrochemistry to mass transport in the evolving near field; localized corrosion (pitting and crevice corrosion), whose initiation criteria and pit-depth statistics are hard to extrapolate in time and whose cathode-to-anode surface area ratio is ill-defined; environmentally assisted cracking (EAC, the family comprising SCC and hydrogen-related cracking), which requires stress distributions, time dependence, and hydrogen-pickup predictions; microbiologically influenced corrosion (MIC), where biofilms and microbial metabolites such as sulfide can attack the container if microbial activity in the buffer is not fully suppressed; and radiation-induced corrosion. A further structural question is how to define container 'failure' (a through-wall pinhole? extensive breach? plastic collapse?) for the different stages of disposal, and how to represent the resulting spread of failure times in probabilistic safety assessment (PSA, the repository-scale analysis that converts container failures into dose consequences to humans).

Fraser King's 2014 review in JOM surveyed these prediction methods across the international programs (the Swedish/Finnish KBS-3 concept of copper canisters with cast-iron inserts, carbon steel concepts in Japan, Belgium and Switzerland, titanium and Alloy 22 in Canada and the US) and concluded that, although mechanistic and empirical approaches existed for each mode, the interactions of mechanical and corrosion degradation had until recently not been considered in detail, and that 'the most important aspect... is to be able to justify the long-term predictions by demonstrating a mechanistic understanding of the various degradation modes.' One coupling route uses fracture mechanics: engineering critical assessment against a failure-assessment diagram (FAD), a plot whose axes measure proximity to plastic collapse (load ratio LrL_r) and to brittle fracture (stress-intensity ratio KrK_r), with the assessment point required to stay inside a material-dependent safety envelope. Time-dependent FAD 'trajectories' can then represent a container whose wall thins by corrosion, whose toughness falls with hydrogen uptake, and whose flaws grow or become surface-breaking — the approach developed for carbon steel containers and extended into probabilistic Monte Carlo breaching models.

The residual difficulty is epistemic: repository timescales exceed any laboratory experiment by four to six orders of magnitude. Validation must therefore lean on short-term tests under accelerated conditions, natural and archaeological analogues, and full-scale in-situ experiments, none of which precisely matches repository materials and environments. Whether such evidence can make a long-term extrapolation trustworthy — and how to formalize that trust — is itself part of the open problem.

Problem Statement

Develop and defend a general lifetime-prediction model for high-level waste and spent fuel disposal containers that couples mechanical and corrosion degradation in a single framework — accounting for containers whose material properties evolve through wall thinning, defect or flaw growth, and hydrogen uptake — and demonstrate that its predictions are robust over repository timescales (tens of thousands to one million years), where direct observation is impossible. Within this single objective, the framework must address the component prediction problems documented as open across the 2014–2026 review lineage (King's 2014 JOM review of container-lifetime prediction, as updated and restated by the King et al. 2024 two-part review, Muñoz et al. 2024, and King 2026): (i) a defensible treatment of localized corrosion, including the cathode-to-anode surface area ratio, the time-extrapolation of initiation criteria and pit-depth statistics, and the stifling of propagating attack as the near-field environment evolves; (ii) reliable stress distributions, their time dependence, and predictions of hydrogen pickup for environmentally assisted cracking; (iii) prediction of corrosion damage under biofilms when microbial activity cannot be excluded; and (iv) an operational definition of container failure appropriate to the assessment context and a representation of the resulting spread of failure times in probabilistic safety assessment. The scientific target is the coupled model together with the validation logic by which its multi-millennial extrapolations are justified and the standard against which they are judged robust; the framework is general across container materials and design concepts rather than restricted to a single program or material, and it addresses the component problems (i)–(iv) jointly.

References

  1. Fraser King, 'Predicting the Lifetimes of Nuclear Waste Containers', JOM 66(3):526–537, 2014. DOI: 10.1007/s11837-014-0869-3, https://doi.org/10.1007/s11837-014-0869-3
  2. F. King, D. Burt, J. Ganeshalingam, P. Gardner, D. Sanderson, S. Watson, C. Padovani, 'Coupled analysis of mechanical- and corrosion-related degradation of carbon steel spent fuel container', Corrosion Engineering, Science and Technology 49(6):442–449, 2014. DOI: 10.1179/1743278214Y.0000000189, https://doi.org/10.1179/1743278214Y.0000000189
  3. Fraser King, Miroslav Kolář, 'Lifetime Predictions for Nuclear Waste Disposal Containers', Corrosion 75(3):309–323, 2018. DOI: 10.5006/2994, https://doi.org/10.5006/2994
  4. Fraser King, Miroslav Kolàř, Scott Briggs, Mehran Behazin, Peter Keech, Nikitas Diomidis, 'Review of the Modelling of Corrosion Processes and Lifetime Prediction for HLW/SF Containers—Part 1: Process Models', Corrosion and Materials Degradation 5(2):124–199, 2024. DOI: 10.3390/cmd5020007, https://doi.org/10.3390/cmd5020007
  5. Fraser King, Miroslav Kolàř, Scott Briggs, Mehran Behazin, Peter Keech, Nikitas Diomidis, 'Review of the Modelling of Corrosion Processes and Lifetime Prediction for HLW/SF Containers—Part 2: Performance Assessment Models', Corrosion and Materials Degradation 5(2):289–339, 2024. DOI: 10.3390/cmd5020013, https://doi.org/10.3390/cmd5020013
  6. Andrés G. Muñoz, Abdesselam Abdelouas, Ursula Alonso, Ana María Fernández, Rizlan Bernier-Latmani, Andrea Cherkouk, Roberto Gaggiano, James Hesketh, Nick Smart, Cristiano Padovani, Kristel Mijnendonckx, Vanessa Montoya, Andrés Idiart, Arnau Pont, Olga Riba, Nicolas Finck, Ashutosh R. Singh, Fraser King, Nikitas Diomidis, 'WP15 ConCorD state-of-the-art report (container corrosion under disposal conditions)', Frontiers in Nuclear Engineering 3:1404739, 2024. DOI: 10.3389/fnuen.2024.1404739, https://doi.org/10.3389/fnuen.2024.1404739
  7. Fraser King, 'The Importance of Considering the Service Environment When Studying and Predicting the Performance of Corrodible Structures', Corrosion and Materials Degradation 7(1):8, 2026. DOI: 10.3390/cmd7010008, https://doi.org/10.3390/cmd7010008
  8. Qichao Zhang, Yishan Jiang, Binbin Zhang, Jizhou Duan, Xin Zhao, Juna Chen, 'Research progress on the hydrogen permeation behavior of the nuclear waste container – A mini review', Frontiers in Materials 9:1058394, 2022. DOI: 10.3389/fmats.2022.1058394, https://doi.org/10.3389/fmats.2022.1058394
  9. H. Asano, A. Nakamura, M. Kobayashi, 'Long term integrity of overpack closure weld for HLW geological disposal. Part 1 – Prediction and evaluation method for structural integrity of weld joint', Corrosion Engineering, Science and Technology 46(2):165–170, 2011. DOI: 10.1179/1743278210Y.0000000016, https://doi.org/10.1179/1743278210Y.0000000016