Search

The long‐term disposal of nuclear waste in deep geological repositories requires safe containment for up to one million years. To understand how the waste will respond to the gradually evolving environment, researchers can perform sensitive experiments in the laboratory to replicate the repository conditions at specific time points. In Belgium, spent nuclear fuel will be housed within a carbon steel overpack, which is encased within a cement buffer, which may include a stainless steel liner (the "envelope"). The range of corrosion behaviors for carbon steel, and to a lesser extent, stainless steel, has been studied through hydrogen evolution, an end‐product of oxidation under repository conditions. Considerable time may be required for metallic surfaces to approach a steady hydrogen evolution and this is dependent upon the starting condition of the specimen surfaces. The use of nominally‐similar materials can result in significant variation in the reported corrosion rates. The presence of cements, which generate their own hydrogen and influence water chemistry as a function of time, add complexity to even simple experimental configurations that requires careful interpretation. [ABSTRACT FROM AUTHOR]

The long‐term disposal of low‐ and intermediate‐level radioactive waste is anticipated to involve the encapsulation of the material within a cementitious matrix before placement in a deep geological repository. After sealing, the waste repository will become anoxic and the corrosion of emplaced metals will result in the production of hydrogen gas. Knowledge of the rate at which this hydrogen is generated is valuable in that it permits the assessment of pressure build‐up within the repository. The corrosion behavior of mild steel has been monitored through the generation of hydrogen gas, under simulated repository conditions that mimic various stages of cement evolution. It was found that, when cement is used to encapsulate steel, the passage of several years is required for the steel and cement to reach a steady hydrogen production rate. Steel that is directly immersed in simulated cement pore water requires only months to stabilize. All measured steel uniform corrosion rates were significantly below 1 nm year−1 at 50°C, regardless of the test environment, after more than 4 years of exposure. [ABSTRACT FROM AUTHOR]

Microscopic and spectroscopic analyses were conducted on steel/bentonite interface samples removed from four in situ experiments that were carried out in three underground research laboratories at different temperatures and under different hydraulic and geochemical conditions. The results provide valuable information about the corrosion processes occurring in high‐level radioactive waste repositories. Systematic patterns can be deduced from the results, irrespective of carbon steel grade, type of bentonite and its degree of compaction, geochemical environment or experimental setup. Thus, a clear dependence of the corrosion rates on temperature and exposure period, as well as on the availability of H2O and O2 provided by the surrounding bentonite buffer, is observed. Furthermore, Fe(II) ions released by corrosion interact with the structural Fe in the clay. Recent developments highlight the usefulness of reactive transport modelling in understanding the coupled corrosion and Fe–clay interaction processes. [ABSTRACT FROM AUTHOR]

Deep geological repositories for radioactive waste contain metallic materials, either used to construct disposal canisters or as low‐/intermediate‐level waste (L/ILW). The safety relevance of corrosion is linked to canister lifetime in the former case and gas generation in the latter. More specifically, the Belgian "supercontainer" concept envisages mild steel for the used fuel disposal canister, and in the case of the Swiss L/ILW repository, mild steels are the largest metallic waste component due to the decommissioning of civilian power‐generating facilities. For these circumstances, the corrosion environment is dominated by the chemistry of cement, which is used as buffer or backfill material. The corrosion behaviour of mild steel in anoxic environments was studied through the analysis of the hydrogen end‐product. Hydrogen analysis was conducted by periodically purging the cell head‐space and analysing the gas using a solid‐state hydrogen sensor. While this method is limited to providing only uniform corrosion rates averaged over periods of time, ranging from weeks to months, it provides excellent resolution and sensitivity. The test cell environments were matched against the anticipated Belgian high‐level waste and Swiss L/ILW repository environments, and also against experiments that have been conducted by other researchers for comparative purposes. Samples were exposed to synthetic cement pore waters, representing fresh and degraded cement. In young cement waters, the formation of initial corrosion products resulted in steel wire corrosion rates of the order of µm/year, which, at 80°C rapidly declined to ∼10 nm/year. In contrast, SA516 grade 70 steel plate corroded much more slowly under similar conditions. In aged cement waters, initial corrosion rates were higher but declined faster towards a longer‐term rate of ∼10 nm/year. 316L stainless steel, embedded in cementitious material, corroded at a rate of <1 nm/year at 50°C. [ABSTRACT FROM AUTHOR]

The Swiss waste management programme foresees that low‐ and intermediate‐level radioactive waste will be disposed of in a deep geological repository constructed in Opalinus Clay. Gas generation is expected in the repository due to the decomposition of organic materials and the corrosion of metals, with carbon steel being the primary source. The corrosion behaviour of mild steel under anoxic conditions has been studied over the course of several years to better understand the long‐term hydrogen evolution profile under anticipated repository conditions. Steel, either bare or encased within mortar, was tested in water vapour or immersed in electrolytes representative of aged cement waters at 50°C. The corrosion rate was measured indirectly through the hydrogen analysis using a solid‐state probe. The hydrogen evolution behaviour of grout was also monitored to more accurately determine the hydrogen generating from the corrosion of the embedded steel. For steel in water vapour or in alkaline environments, embedded in cementitious material or immersed in simulated aged cement pore water, corrosion rates were invariably <1 nm/year after several years of analysis. [ABSTRACT FROM AUTHOR]

Carbon steel (C-steel; E24 and S235 grades) and stainless steel (316L) electrodes were corroded in situ in Opalinus clay under anoxic conditions in a vertical descending borehole. The electrodes were exposed at ambient temperature for two years, and then at 85°C for five years. In situ electrochemical impedance spectroscopy showed that the instantaneous corrosion rate of C-steel decreased over time down to 1 µm year−1, followed by steady state. Microbial and chemical investigations showed that sulphate and thiosulphate reducing prokaryotes were present in the porewater and at the metal surface. Post mortem characterisation revealed contrasting corrosion aspects. The E24 corrosion interface was thick and contained magnetite, mackinawite, hydroxychloride and siderite, together with more oxidised species (goethite, greigite, elemental sulphur). The S235 corrosion interface was thinner and contained siderite and mackinawite. Corrosion damage of the 316L electrode was negligible, and the surface was covered by a thin fringe of pyrite. [ABSTRACT FROM AUTHOR]

The anaerobic corrosion of steel was monitored under simulated Swiss low-/intermediate-level waste repository conditions. Steel was studied both as bare rods and cast in cement, in water vapour, simulating the unsaturated anoxic phase, and during immersion in various pore waters, representative of the saturated phase. All tests were performed at 50°C. Hydrogen evolution from grout, comprising commercial cement and sand, was also studied. Grouted steel specimens initially generated hydrogen that was comparable to grout alone, which was attributed to the presence of iron detritus from the ball-milling of cement clinker. The corrosion rate of immersed test specimens peaked during the first 100 days, likely due to the formation of a passive film. Corrosion rates generally declined to <3 nm/year, although there were several notable exceptions where hydrogen generation was erratic. This was attributed to localised corrosion and may be an important factor to understand when modelling hydrogen generation rates. This paper is part of a supplement on the 6th International Workshop on Long-Term Prediction of Corrosion Damage in Nuclear Waste Systems. [ABSTRACT FROM AUTHOR]