Search

A wide range of metals have been studied as candidate container material, which is considered in the geological disposal concept of high-level radioactive waste and spent fuel in Belgium. More than 40 years ago, SCK CEN started studies on the corrosion of these metals. The HADES underground research laboratory (URL) played an important role in the corrosion research as it enabled in situ corrosion experiments that simulate realistic disposal conditions as closely as possible. These experiments consisted of placing metallic coupons on a steel support tube, heated from the inside, that was installed in the Boom Clay, thereby exposing the coupons to various conditions representative of the disposal concept that was considered at that time. Test durations lasted from 6 months to approximately 7.5 years. This paper summarises the results from the various corrosion studies and discusses their implications on the choice of disposal concept. One of the main outcomes of these experiments was a change of rationale regarding the choice of the container material from carbon steel (corrosion-allowance) to stainless steel (corrosion-resistant). The main arguments for this change were the need to avoid severe pitting corrosion during the aerobic period and to minimise the generation of hydrogen gas during the subsequent anaerobic period.

The Belgian reactor 1 fuel cladding is made of aluminium. If disposed of in a geological disposal repository, aluminium could come into contact with ordinary Portland cement and is going to corrode to form hydrogen gas and aluminium hydroxide. In this study, the long‐term corrosion was evaluated by electrochemical impedance spectroscopy on metal embedded in cement paste immersed in saturated calcium hydroxide solution under anaerobic conditions. Mass transfer effects were investigated by embedding aluminium in cement paste with two different porosities. LiNO3 was also added to the systems to study its corrosion inhibiting effect. After high initial values, the corrosion rate rapidly declined to reach a steady state after a few days. After ~100 days, the corrosion rate was the lowest when cement paste possessing a low porosity with the addition of LiNO3 was used and the highest when cement paste with high porosity without LiNO3 was used. Finally, scanning electron microscopy–energy X‐ray dispersive spectroscopy revealed that the corrosion product layer was thicker than expected by electrochemical impedance spectroscopy measurements, while it is only composed of aluminium and oxygen. Aluminium was found to diffuse into the cement paste close to this corrosion product layer.

In this work, the nature of the electric field (E) in the defective oxide barrier layer on iron has been explored theoretically, within the classical regime, with the goal of ascertaining the validity of the postulates underlying the Point Defect Model (PDM). The analysis shows that if the thickness of the barrier layer under steady-state condition, L, is sufficiently small, the electric field strength and concentrations of point defects are essentially constants in the main part of this layer (with the exceptions in the very thin region near the metal/barrier layer (m/bl) and barrier layer/solution (bl/s) interfaces, depending upon the defect). A quantitative criterion was developed to decide when the barrier layer of the passive film can be considered to be sufficiently thin under these conditions. The potential drops inside the barrier layer and across the m/bl and bl/s interfaces can be represented as linear functions of applied voltage, V. This is the consequence of the change in thickness with applied voltage, with E being approximately (but with high accuracy) independent on the position inside the barrier layer and applied voltage. All conclusions stated above were derived from the solution of the system of mass transport equations for point defects and Poisson equation for the electric potential without applying any additional assumption except the small thickness of the barrier layer and without taking into account quantum (band-to-band) tunneling.

The nature of the electric-field strength in the passive film on tungsten is explored using photo-electrochemical techniques. A theoretical expression for the photo-stimulated growth of the film has been derived. Rotating ring-disk electrode (RRDE) experiments indicate that photo-corrosion of a tungsten electrode is negligible under super-band gap light illumination. XPS results show that tungsten is in the maximum possible oxidation state (VI) in the film and hence no higher oxidation state is available. The photo-stimulated transient film growth of tungsten in pH 8.5 ± 0.1 boric-borax buffer was recorded as a function of film formation potential (1 VSCE, 2 VSCE, 3 VSCE, 4 VSCE, 5 VSCE, 6 VSCE) and light intensities (50, 200, 1000 mW/cm2). Steady state passive film thickness measurements of the passive film on tungsten indicate that super band gap, UV light suppresses the electric field in the barrier layer, and hence stimulates anodic oxide film growth. The data obtained in this study demonstrate that the electric field strength in the steady state is independent of the applied potential and film thickness, as postulated in the Point Defect Model.

The Supercontainer (SC) is the reference concept for the post-conditioning of vitrified high-level nuclear waste and spent fuel in Belgium. It comprises a prefabricated concrete buffer that completely surrounds a carbon steel overpack. Welding is being considered as a final closure technique of the carbon steel overpack in order to ensure its water tightness. Welding is known to induce residual stresses near the weld zone, which may lead to an increased susceptibility to stress corrosion cracking (SCC). In this study, slow strain rate tests were conducted to study the SCC behavior of plain and welded P355 QL2 grade carbon steel exposed to an artificial concrete pore water solution that is representative of the SC concrete buffer environment. The tests were performed at 140°C, a constant strain rate of 5 × 10-7 s-1 and at open circuit potential under anoxic conditions. The effect of thiosulfate on the SCC behavior was investigated up to levels of 600 mg/L S2O32-.

The Supercontainer is the reference concept for the post-conditioning of vitrified high-level radioactive waste and spent fuel in Belgium. It consists of a prefabricated concrete buffer that completely surrounds a carbon steel overpack. In this highly alkaline environment (pH ∼ 13.6) and under normal conditions (i.e. without the ingress of aggressive species), the carbon steel overpack will be protected by a passive oxide film, which is believed to result in very low uniform corrosion rates. This paper gives an overview of the status of the uniform corrosion, pitting corrosion and stress corrosion cracking behaviour of carbon steel expected during the waste disposal period.

A method for calculating porosity of the corrosion product on the canister and of the engineered barriers (concrete) in the disposal of HLNW in Belgium supercontainer has been developed. The porosity of the rust is predicted to decrease substantially, such that the growth of the outer layer ceases and further corrosion is prevented. The decrease in the porosity of concrete is less than the decrease in the porosity of the rust. This is the consequence of the different thicknesses of the rust and concrete layers and different mechanical properties of the materials.

A mixed potential model has been developed to deconvolve the negative total current density that is observed at potentials more negative than the open circuit potential, E corr , during potentiostatic polarization of carbon steel in the simulated concrete pore solution under anoxic conditions into its partial anodic and cathodic components as a function of potential across the passive range. Deconvolution was successfully achieved by optimizing a Mixed Potential Model (MPM) comprising the Point Defect Model (PDM) to describe the partial anodic process and the Generalized Butler-Volmer equation to describe the partial cathodic process of hydrogen evolution. In this manner, the corrosion rate can be determined across the entire passive range, including the range of potential ( E E corr ) within which the net observed current density is negative. It is found that the corrosion rate increases significantly when the potential is lowered below a certain value, possibly due to the change of the oxidation state of iron within the outer-layer of the film.

A bounding model has been developed to calculate the impact of the porous, outer layer of the passive film on the corrosion rate of the carbon steel overpack and on the structural integrity of the concrete in the annulus of a supercontainer proposed for storage of high level nuclear waste. This model allows us to estimate the corrosion potential, rates of general corrosion of the overpack, hydrogen pressure, water saturation, and the porosity of the rust as corrosion proceeds along the corrosion evolutionary path. It is shown that, at lower values of water saturation S < 0.2-0.3, complete exhaustion of water within the annulus in predicted, completely protecting the overpack from further general corrosion. At higher values of water saturation the overpack is also predicted to be protected by the compression of the outer layer against the concrete buffer and hence by the reduction in porosity of the rust to form an impervious barrier.

We report an investigation of artificial pit growth (APG) in P355QL2 Grade Carbon Steel in the highly alkaline, synthetic concrete pore solution [sat. Ca(OH)2/NaOH of pH of 13.5] + 0.5 M NaCl under both “oxic” and “anoxic” environmental conditions. The corrosion potential of the external surface, as well as the coupling currents flowing between the external surface and the pit tip and pit walls of the segmented carbon steel specimens under the selected conditions, were measured as a function of time and pit aspect ratio. The findings were interpreted in terms of the Differential Aeration Hypothesis (DAH).

In this paper, we report an EIS study of the evolution of the passive state of carbon steel in contact with simulated concrete pore water. By mining the impedance data, we have determined that, with time, the defect structure and the defect-dependent properties of the barrier layer of the passive film on iron remain invariant, but that the change in the properties of the passive film, as a whole, can be accounted for in terms of changes in the thickness of the precipitated outer layer. The importance of this finding is that we may apply, with added confidence, the time-independent Point Defect model for describing the barrier layer on carbon steel in contact with concrete pore water over periods of time that are well in excess of experimentally accessible times. Thus, as envisaged in the super-container concept in Belgium’s technology for the disposal of High Level Nuclear Waste (HLNW), the carbon steel over-pack will be exposed to concrete pore water for times exceeding 100,000 years, with the changes in the passive state of the metal being attributed to the increase in thickness of the precipitated, outer layer alone. The rate of formation of the outer layer film at 80 oC is calculated to be 4.6 10-12 cm/s or 1.45 10-4 cm/year = 1.45 ƒÝm/year, which appears to be an eminently reasonable value.

A model has been developed to calculate the impact of the porous, outer layer of the passive film on the corrosion rate of the carbon steel overpack and on the structural integrity of the concrete in the annulus of the super container of the Belgian High Level Nuclear Waste (HLNW) disposal concept. Because the corrosion product occupies 2 – 6.5 times the volume of the iron from which it forms the development of the outer layer in the confined space between the carbon steel overpack and the concrete will result in the compression of the outer layer and a concomitant decrease in the porosity. If the porosity decreases to a sufficiently low value, the growth of the outer layer ceases and further corrosion is prevented. If the thickness of the corroded layer exceeds a critical value, the concrete will crack, because of the poor strength of cementations materials under tensile loading.

The prediction of corrosion damage to times that are experimentally inaccessible by a large factor (e.g., to over 1000 years) is vitally important in assessing various concepts for the disposal of High Level Nuclear Waste. Such prediction can only be made using deterministic models, whose predictions are constrained to being “physically real” by the natural laws [conservation of mass, energy, charge and mass-charge equivalence (Faraday’s Law)]. In this paper, we describe the measurement of experimental data that will allow the deterministic prediction of damage to the carbon steel overpack of the super container in Belgium’s proposed Boom Clay repository by using the Point Defect Model to extrapolate damage to future times. We also describe an experimental program that is designed to generate values for various PDM parameters for the corrosion of carbon steel and stainless steel that will be required in making the damage predictions.

The prediction of corrosion damage to times that are experimentally inaccessible by a large factor (e.g., to over 1000 to 1,000,000 years) is vitally important in assessing various concepts for the disposal of High Level Nuclear Waste (HLNW). Such prediction can only be made using deterministic models, whose outputs are constrained to being “physically real” by the time- and space-invariant natural laws [conservation of mass, energy, charge and mass-charge equivalence (Faraday’s Law)]. In this paper, we explore the long-term passivity behavior of carbon steel in contact with concrete pore water solution at 80 ◦ C [sat. Ca(OH)2 + sufficient NaOH to yield pH25C = 13.5] by modeling the electrochemical impedance spectroscopic behavior using the Point Defect Model. As a result, we developed a single set of kinetic parameters that might be used, in order to predict the accumulation of general corrosion damage to the supercontainer in Belgium’s HLNW clay repository. The results of the optimization indicate that iron interstitials are the dominant defects in the defective barrier oxide layer and that the diffusivity of the defect calculated to be ≈10 −15 cm 2 s −1 .

The corrosion behaviour in repository conditions is an important factor in the selection of a container material for the disposal of conditioned high-level radioactive waste. To assess the corosion resistance of candidate container materials, electrochemical experiments are performed at SCK.CEN. More specifically, polarisation curves are recorded for eight selected alloys: carbon steel TStE 355, AISI 316L, AISI 316L hMo, UHB 904L, Cronifer 1925 hMo, Hastelloy C4, Ti/0.2Pd, and AISI 316Ti. From these curves, the characteristic pitting potentials E np and E pp were derived. The investigated parameters were: chloride concentration, sulphate concentration, temperature, and alloy composition. From the results can be concluded that an increasing chloride content increases the pitting susceptibility of the investigated metals in clay water. A higher sulphate content decreases the pitting susceptibility in clay water. At higher temperatures, the alloys are less resistant to pitting.

Passivity breakdown and pitting on carbon steel, the material for the overpack holding the waste canister inside high-level nuclear waste supercontainer in Belgium, has been studied in simulated concrete pore solution (SCPS) with a pH of 13.5 and the data are interpreted in terms of the Point Defect Model (PDM). The near normal distribution in the critical breakdown potential (V c) measured in deaerated SCPS with different chloride concentrations is in satisfactory agreement with the quantitative prediction of the PDM. The critical areal concentration of condensed cation vacancies at the metal/film interface that leads to passivity breakdown, as predicted by the PDM from the experimental dependence of the V c on the square root of the voltage sweep rate, ξ 16 cm−2, is in excellent agreement with that calculated from the unit cell dimensions of the substrate Fe (ξ ≈ 1.53 × 1016 cm−2) and the barrier layer oxide (Fe2O3) (ξ ≈ 6.89 × 1015 cm−2) for vacancy condensation upon the metal lattice or upon the cation sublattice of the film, respectively. This observation provides convincing evidence for the validity of the PDM for modeling passivity breakdown.

Passivity breakdown on carbon steel, which is the reference candidate material for fabricating high-level nuclear waste containers in Belgium, has been studied and the data are interpreted in terms of the point defect model (PDM). The near normal distribution in breakdown potential, measured in deaerated simulated concrete pore water solution (pH 12.5) at different chloride concentrations and temperatures are in satisfactory agreement with prediction of the PDM. An increase in temperature from 25°C to 85°C results in a decrease in localized corrosion resistance according to a linear relationship between the critical breakdown voltage, Eb , and log(chloride activity) for all chloride concentrations. The linear dependence of the passivity breakdown potential on the square root of potential scan rate (ν1/2), as predicted by the PDM, yields an estimate of the critical areal concentration of condensed cation vacancies at the metal/film interface (ξ14cm-2 at 85oC and 14cm-2 at 55oC ) that leads to the passivity breakdown. The experimental distributions in the breakdown potential for carbon steel at 55oC and 85oC are in satisfactory agreement with the relevant calculated distributions. Acknowledgment: The authors gratefully acknowledge the support of this work by ONDRAF-NIRAS of Belgium.

Stainless steel is being envisaged as the primary candidate container material for the final disposal of vitrified HLW in deep geological argillaceous formations in Belgium. The impact of an evolving underground repository environment, i.e. a progressive change from oxic to anoxic conditions (due to the consumption of entrapped oxygen), on the pitting behaviour of austenitic stainless steels AISI 316L hMo and UHB 904L was studied. CPP-experiments were performed in synthetic solutions, which are representative for the near-field chemistry of an underground repository. The solutions contained various amounts of Cl- (100–50,000 mg/L) at near-neutral pH. Experiments were conducted at 16 and 90°C.AISI 316L hMo and UHB 904L will not be subjected to immediate pitting problems neither under oxic, nor under anoxic conditions. However, AISI 316L hMo could present long-term pitting problems under oxic conditions. Pits are much easier initiated on AISI 316L hMo, for both oxic and anoxic conditions. The pits propagate in a rather similar manner under oxic conditions for both alloys, whereas under anoxic conditions the pits formed on AISI 316L hMo are much deeper. AISI 316L hMo is more susceptible to crevice attack.

The current worldwide trend for the final disposal of conditioned high-level, medium-level and long-lived alpha-bearing radioactive waste focuses on deep geological disposal. During the geological disposal, the isolation between the radioactive waste and the environment (biosphere) is realised by the multibarrier principle, which is based on the complementary nature of the various natural and engineered barriers. One of the main engineered barriers is the metallic container (overpack) that encloses the conditioned waste. In Belgium, the Boom Clay sediment is being studied as a potential host rock formation for the final disposal of conditioned high-level radioactive waste (HLW) and spent fuel. Since the mid 1980’s, SCK•CEN has developed an extensive research programme aimed at evaluating the suitability of a wide variety of metallic materials as candidate overpack material for the disposal of HLW. A multiple experimental approach is applied consisting of i) in situ corrosion experiments, ii) electrochemical experiments (cyclic potentiodynamic polarisation measurements and monitoring the evolution of ECORR as a function of time), and iii) immersion experiments. The in situ corrosion experiments were performed in the underground research facility, the High Activity Disposal Experimental Site, or HADES, located in the Boom clay layer at a depth of 225 metres below ground level. These experiments aimed at predicting the long-term corrosion behaviour of various candidate container materials. It was believed that this could be realised by investigating the medium-term interactions between the container materials and the host formation. These experiments resulted in a change of reasoning at the national authorities concerning the choice of over-pack material from the corrosion-allowance material carbon steel towards corrosion-resistant materials such as stainless steels. The main arguments being the severe pitting corrosion during the aerobic period and the large amount of hydrogen gas generated during the subsequent anaerobic period. The in situ corrosion experiments however, did not allow to unequivocally quantify the corrosion of the various investigated candidate overpack materials. The main shortcoming was that they did not allow to experimentally separate the aerobic and anaerobic phase. This resulted in the elaboration of a new laboratory programme. Electrochemical corrosion experiments were designed to investigate the effect of a wide variety of parameters on the localised corrosion behaviour of candidate overpack materials: temperature, SO42−, Cl−, S2O32−, oxygen content (aerobic - anaerobic),… Three characteristic potentials can be derived from the cyclic potentiodynamic polarisation (CPP) curves: i) the open circuit potential, OCP, ii) the critical potential for pit nucleation, ENP, and iii) the protection potential, EPP. Monitoring the open circuit potential as a function of time in clay slurries, representative for the underground environment, provides us with a more reliable value for the corrosion potential, ECORR, under disposal conditions. The long-term corrosion behaviour of the candidate overpack materials can be established by comparing the value of ECORR relative to ENP and EPP (determined from the CPP-curves). The immersion tests were developed to complement the in situ experiments. These experiments aimed at determining the corrosion rate and to identify the corrosion processes that can occur during the aerobic and anaerobic period of the geological disposal. Also, some experiments were elaborated to study the effect of graphite on the corrosion behaviour of the candidate overpack materials.

not Available.