Two semi-static dissolution tests using oxidized PWR spent fuel specimens are being conducted under ambient hot cell conditions in Nevada Test Site J-13 well water and unsealed fused silica vessels. The test specimens were oxidized at 250°C in air to bulk oxygen-to-metal (O/M) values of 2.21 and 2.33. Following an initial 191-day test cycle, the specimens were restarted in fresh J-13 water for a second long-term test cycle. Results through the first 40 months of Cycle 2 are compared with results from similar tests at 25°C and 85°C using unoxidized spent fuel specimens.

Increased concentrations of U, Am, Cm and Np were measured in 0.4- μm filtered samples from the oxidized fuel tests compared to the unoxidized fuel tested at 25°C; Pu concentrations were not affected by the fuel oxidation state. Most of the Am and Cm, and a portion of the Pu, measured in 0.4-μm filtered samples was removed by 2-nm filtration. Fission product release results were normalized to specimen inventories and reported as fractional release. No attempt was made to normalize the data to surface area. Initial 99Tc release was greatly increased, and prolonged increases in the fractional release rates of 99Tc and 129I occurred as a result of fuel oxidation. Fractional release rates for 137Cs and 90Sr from oxidized fuel eventually decreased to levels similar to those observed with unoxidized fuel after equivalent testing times, suggesting that matrix dissolution rates normalized to fuel mass were not increased as a result of oxidation.

Experimental methods have been developed for measuring the grain-boundary inventories of radionuclides and for determining whether the UO2 matrix of spent light-water reactor fuel dissolves congruently. Both methods depend upon first separating the fuel into individual grains. With the grain boundaries thus exposed, the associated inventories of radionuclides can be completely dissolved and measured. To determine whether the UO2 matrix of spent fuel dissolves congruently, the fuel grains were placed in a flow-through column and water was pumped through the column at a rate sufficient to maintain the concentration of U in the column effluent far below saturation. Once the grain-boundary material has dissolved, the forward dissolution rate of the UO2 matrix can be measured and, by measuring the concentrations of other radionuclides in the column effluent, the degree of congruency of the dissolution process can be determined. Data obtained to date indicate that the grain-boundary inventories of Cs, Tc, and Sr are approximately equal to gap inventories and that the fractional dissolution rate of Cs from the U02 matrix is approximately equal to that of U, i.e., the Cs and U dissolved nearly congruently. In addition, the data appear to show a gradient in the concentrations of Cs and Sr across the individual UO2 grains.

In order to be able to model the effect of water radiolysis on the dissolution of fuel in the disposal vault it is essential to study the oxidation mechanism. Radiolysis of water produces radicals and molecules, amongst them the oxidizing species H2O2 and OH radicals, and, in aerated solutions, O2- as well. These species are probably responsible for the oxidation and dissolution of UO2 observed in a number of experiments. In the present investigation the effect of the oxidizing species has been studied separately using suitable scavengers. A polished specimen of UO2 was immersed in water and irradiated at ambient temperature with Co-60 gamma radiation. After irradiation the amount of uranium which had dissolved, or deposited elsewhere than the sample surface, was measured, and the oxidation of the surface was determined, using ESCA and SIMS. In 50 mM H2O2 (without irradiation) the corrosion rate was 5±3 μg.cm−2d−1 after 6 d exposure at pH 8. Corresponding rates for oxidation by O2- and OH radicals were about 3 μg.cm−2d−1, (dose rate 600±90 Gy/h.)

The dissolution of a non-radioactive chemical analogue of spent nuclear fuel (SIMFUEL) has been studied as a function of two different synthetic groundwaters at 25°C. The cumulative release of U, Mo, Ba, Y and Sr is presented after 170 days of total leaching time.

The results obtained show the potential usefulness of SIMFUEL to ascertain the kinetics and mechanisms of dissolution of the minor components of spent fuel. In the case of Sr, a good correlation is found with the dissolution of this minor component measured in spent fuel leaching experiments.

The solubility of a crystalline unirradiated UO2 (s) has been studied under reducing conditions at 25°C in three different ionic media: 0.008 mol dm–3 NaCIO4, 1 and 5 mol dm-3 NaCl+. The species responsible for the solubility in perchlorate medium are U(OH)3 and U(OH)4, with the stability constants log β13 = -0.4(±0.2) and log β14 = -5.7(±0.1), respectively. The solubility in both1! and 5 M chloride media is explained at pH values higher than 6 by the species U(OH)4 with log β14 = -5.5(±0.2) while at lower pH values an abnormal behaviour is observed, with higher uranium concentrations in solution and a proton independent solubility for pH values lower than 4. The XPS observations of the reacted solids do not show the presence of solid surface phases other than uranium dioxide, with an upper oxidation state of UO2.1.

Four powdered samples of natural UO2+x (uraninite) were annealed in a reducing atmosphere up to 1200°C. The initial unit cell parameters ranged from ao=0.5463 to 0.5385 nm. Small amounts of UO2.25 occur in all samples after annealing. Annealing curves show the effects of recovery of point defects in the oxygen sublattice, ordering of U4+ and U6+, vacancy migration in the cation sublattice, and second order phase transformations. The difference in the annealing behavior of UO2+x with x < 0.15 as compared to x = 0.25 between 400 and 700°C is due to ordering of U4+ and U6+. Density increased after annealing except for one sample in which a large number of cavities (1–2 μm in size) formed. Oxidation and chemical composition have a more dramatic effect on the structural state of natural UO2+x than self-irradiation caused by a-decay damage.

Uranyl oxide hydrates, formed by the alteration of uraninite, are natural analogues for the long-term corrosion products of spent fuel in a geologic repository under oxidizing conditions. The uranyl oxide hydrates may be represented by the general formula:

Pb-bearing hydrates require the addition of a neutral uranyl group into the structural sheet (UO2(OH)2) for each interlayer Pb ion. Distortion of the structure associated with the additional uranyl groups is reduced by replacing two structural hydroxyls with a structural oxygen and a molecular water. The general formula for the Pb-uranyl oxide hydrates is:

This hypothesis explains the paragenetic sequences:

1) schoepite ➛ billietite ➛ protasite ➛ bauranoite

2) schoepite ➛ vandendriesscheite ➛ fourmarierite ➛ masuyite ➛ wölsendorfite

3) schoepite ➛ vandendriesscheite ➛ fourmarierite ➛ ± masuyite ➛ sayrite ➛ curite, and indicates that, under relatively high pH conditions, schoepite will not be the long-term solubility-controlling phase for uranium in uranium-rich groundwaters.

Candidate container materials for high-level nuclear waste packages to be emplaced at the proposed Yucca Mountain repository site are being considered for their long-term resistance to corrosion, oxidation, embrittlement and other kinds of degradation. Selection criteria have been established, and a method has been developed for recommending a material for advanced container design activities. An extensive compilation of the degradation phenomena for six candidate materials is complete, and further studies have begun on the degradation modes affecting additional candidate materials. Phenomenological models for predicting container degradation rates are being advanced for environmental conditions applicable to Yucca Mountain. An experimental program is underway to evaluate the susceptibility of container materials to localized corrosion, stress corrosion cracking, and enhancement of corrosion and oxidation attack by gamma radiation. Initial evaluations of container fabrication and welding processes have identified some processes that appear to alleviate some long-term corrosion susceptibility concerns.

Long term corrosion resistance is evaluated for G-2 and G-12 titanium as a candidate of the high-level nuclear waste packages. The repassivation potential for crevice corrosion, ER, in comparison with the well known spontaneous potential, ESP, of a metal passivated in given environment, allows conclusions to be drawn whether it performs “permanent” or does not. This repassivation method is extended to determine critical conditions in terms of NaCl concentration and temperature for specimens kept at an electrode potential which is more noble than ER and is included in ESP ranges. Thus obtained NaCl concentration - temperature - crevice corrosion map could predict critical conditions for the titanium used in geologic formations.

The effect of environmental variables on the localized corrosion behavior of alloy 825 is examined in this paper. Cyclic, potentiodynamic polarization tests based on a two-level, full factorial experimental design were conducted. An index incorporating both the visual and scanning electron microscope examinations of localized corrosion and the electrochemical parameters was used for the statistical analysis. The analysis showed that chloride is the single most important promoter of localized corrosion, while nitrate was the single most important inhibitor. Fluoride was a weak inhibitor, especially at low chloride levels. Sulfate was a weak promoter, especially at high chloride levels. Temperature did not have a significant effect within the chloride levels examined. Separate experiments indicated that silicon, added as metasilicate, did not have any significant effect on localized corrosion. The adverse effect of chloride was observed at concentrations as low as 100 ppm. Addition of H2O2 increased the corrosion potential of alloy 825 to a value above the repassivation potential observed in the 300 ppm chloride solution.

The general corrosion rate of mild steel is so small in neutral water environments that adequately provided corrosion allowance can ensure the requirement of one thousand years’ integrity of geological disposal package of high-level nuclear wastes. In alkaline water environments, however, mild steels can passivate themselves and often undergo localized corrosion in much the same manner stainless steels do in neutral water environments. This paper describes a study of localized corrosion behavior of the mild steel conducted to assess the long-term performance of the disposal packages. The critical potentials for pitting and crevice corrosion and critical pH for general corrosion-to-passivity transition were determined for neutral and alkaline water environments. Effects of temperature, pH, and chloride and other anion concentrations on the critical potentials and the critical pH were discussed. The initiation and propagation behavior of crevice corrosion was also analyzed under potentiostatic conditions.

The corrosion rate of carbon steel in compacted bentonite was evaluated with regard to the test period length, temperature, chemicals of solution and bulk density of compacted bentonite.

The average corrosion rate decreased gradually with increasing test period up to 180 days in immersion tests. The corrosion rate of carbon steel in compacted bentonite at a dry density of 1.32g/cm3 was estimated to be about 0.01 mm/y which was one order of magnitude lower than that in bentonite slurry. No significant influence of temperature on corrosion rates was observed in compacted bentonite in the range of 50∼180 °C. Variation of kinds and concentration of anion(chloride, floride, sulfate, and carbonate)in aqueous solution did not have much influence on the corrosion rate of carbon steel.

Immersion tests of carbon steel in compacted bentonite at a dry density of 0.69 ∼ 1.32 g/cm3, which was mixed with an aqueous solution(synthetic sea water and distilled water), were carried out. The corrosion rate in compacted bentonite decreased from 0.04 to 0.005mm/y as the density of bentonite increased.

This result suggests that the corrosion rate of carbon steel in compacted bentonite is governed by the diffusivity of corrosive materials. In general, oxygen is the dominate factor affecting corrosion rate, therefore prediction of the average corrosion rate of carbon steel was carried out on the basis of the diffusion behavior of dissolved oxygen in compacted bentonite. The prediction agreed with experimental results.

The production of hydrogen from the anaerobic corrosion of iron or steel is an important issue in low/intermediate level nuclear waste repositories where large quantities of iron and steel (e.g. as drums and reinforcing steel) accompany the waste.

Most of the iron in intermediate level repositories is in a cemen-titious environment. A review of the literature on the corrosion of iron and steel at high pH values, in particular in cementitious environments, points to hydrogen evolution rates between 22 and 220 mmol(H2)m−2a−1. There is some indication that the rates might be lower but for normal engineering applications there has been no practical need to demonstrate this, and hence a lower rate cannot be assumed on current evidence.

In the present work a volumetric method was used to measure hydrogen evolution rates over several thousand hours under conditions relevant to intermediate level waste repositories. The sensitivity of this method (0.4 mmol(H2)m−2a−1) is sufficient to detect hydrogen evolution rates lower than those predicted for iron and steel in concrete.

Hydrogen evolution rates in highly alkaline cement pore water were below 0.4 mmol m−2a−1 and almost 1 mmol m−2a−1 for a pore water representative of an aged cement; no decrease was observed even after 12000 h. In general hydrogen evolution rates in alkaline media were observed to take several thousand hours before approaching a constant rate.

This study is directed at investigating the rate of corrosion and its spatial distribution that develop under conditions where transport of oxygen and conductivity of the environment are controlled over a wide range, including low conductivity as expected in the Yucca Mountain environment. The results indicate that the corrosion rate of steel is directly related to the rate of oxygen transport over several orders of magnitude, and increasing conductivity by one order of magnitude increases corrosion rate by a factor of two or three. Of greater significance is the result that indicates that as conductivity of the environment decreases, and corrosion rate decreases, the degree of localized attack increases.

Copper and copper alloys are frequently used in environments where short-term laboratory testing indicates that they should be resistant to corrosion. However, in long-term service, copper and its alloys show degradation mechanisms which have not been replicated in the laboratory. In this paper, corrosion mechanisms for copper alloys will be examined from the standpoint of corrosion product thermodynamics and corrosion kinetics, with particular reference to microbiologica11y influenced corrosion (MIC). Comparison will be made to the corrosion behavior of high-iron and high-nickel alloys.

The atmospheric corrosion of oxygen-free copper (CDA-102), 70/30 copper-nickel (CDA-715), and 7% aluminum bronze (CDA-613) in an irradiated moist air environment was investigated. Experiments were performed in both dry and 40% RH (@90°C) air at temperatures of 90 and 150°C. Initial corrosion rates were determined based on a combination of weight gain and weight loss measurements. Corrosion products observed were identified. These experiments support efforts by the Yucca Mountain Project (YMP) to evaluate possible metallic barrier materials for nuclear waste containers.

This paper describes galvanic corrosion studies of lead, a candidate matrix metal, coupled vith ASTM Grade-2 titanium, a candidate container-shell material, to determine the barrier-lifetime for a lead matrix container - one container concept considered in the Canadian Nuclear Fuel Waste Management Program. In these studies the fractional factorial statistical design of Box-Behnken was used to develop mathematical relationships between the independent variables and the corrosion rate of lead.

Using the mathematical expressions derived in this study, the additional barrier-lifetime that a 25-mm-thick lead layer could provide to a titanium container, which eventually perforates by corrosion, was calculated to be at least 3500 years.

Stochastic variations in the local chemical environment of a geologic waste repository can cause corresponding variations in container corrosion rates and failure times, and thus in radionuclide release rates. This paper addresses how well the future variations in repository chemistries must be known in order to predict container failure times that are bounded by a finite time period within the repository lifetime. Preliminary results indicate that a 5000 year scatter in predicted container failure times requires that repository chemistries be known to within ±10% over the repository lifetime. These are small uncertainties compared to current estimates.

A near-field performance evaluation of an advanced cold process canister for spent nuclear fuel has been performed jointly by TVO, Finland and SKB, Sweden. The canister consists of a carbon steel canister as a load bearing element, with an outer corrosion shield of copper. The canister design was originally designed by TVO. In the analysis, internal (ie corrosion processes from the inside of the canister) as well as external processes (mechanical and chemical) have been considered both prior to and after canister breach. Throughout the analysis, present day underground conditions have been assumed to persist during the service life of the canister.

An advanced cold process canister design, ACPC-design, is introduced in this paper. The cold process canister consists of a steel canister as a load bearing element, with an outer corrosion shield of copper, Fig. 1 - Fig. 4. The canister will be rilled with granular material, such as lead shots, glass beads or quartz sand.

The conditions for the canister design are presented: dimensioning against radiation, allowed heat loads, strains, corrosion and leak tightness etc. The mechanical dimensioning of the copper/steel canister is made according to the Finnish standards (SFS) for pressure vessels.