Since 1980, considerable research has been invested in the development of sphene-based glass-ceramics as hosts for possible future CANDU™ fuel recycle wastes. This paper reviews the fabrication of the glass-ceramics, their detailed microstructure, the partitioning of waste ions between the glassy and crystalline phases, and their resistance to leaching in deionized water and synthetic brines.

Study of radiation damage in natural titanite (CaTiSiO5) permits estimation of the long-term performance of titanite-bearing glass ceramics in the immobilization of radioactive waste. Single-crystal X-ray structure analyses have been made on two geologically old, U, Th-bearing titanites (sphenes), on an undamaged gem-quality titanite and on their annealed equivalents. The alpha-damaged titanite crystal structure is characterized by a marked increase in magnitude of the apparent thermal parameters of all structural positions, reflecting a general increase in positional disorder. The ratio of observed structure factors (F0-damaged/F0-annealed) may be used to estimate the overall structural damage.

238Pu-substituted zirconolite (CaPuTi2O7) was held at ≈350 K for a time sufficient to allow self-irradiation damage to convert the material to the metamict condition. Samples were then stored at 875 K, while dimensional and microstructural changes were monitored by dilatometry and transmission electron microscopy. Densification proceeded sporadically over a period of ≈400 days, attaining a final value of 4 vol% as a result of crystallization. The material was found to be friable at an intermediate stage of recovery, apparently as a result of internal strains. Evidence is presented that CaPuTi2O7 can exist in more than one metamict state, depending on extent of prior damage and thermal history.

Betafite, a radioactive mineral, is the Ti-rich member of the pyrochlore group, A1–2B2X6Y0–1, (Fd3m, Z = 8), where A = Ca, Na, U, Th, REE (rare earth elements), Y, Ba, Sr, Bi, Pb; B = Nb, Ti, Ta, Zr, Sn, Fe: X = oxygen; Y = 0, OH and F. The pyrochlore structure (or structural derivatives, e.g. zirconolite in SYNROC) is a common constituent of polyphase, crystalline radioactive waste forms. Naturally occurring minerals with this structure often occur in the metamict state. Therefore, a principal concern is the effect of alpha-recoil damage on the structure and dissolution behavior of synthetic, analogue, waste-form phases which contain actinides.

We describe the dissolution behavior of 238U, 234U, 232Th, 230Th and 228Th in betafite during leaching in a bicarbonate-carbonate solution. Results indicate enhanced leaching of recently produced alpha-recoil damage as evidenced by the more rapid dissolution rate of 228Th relative to 232Th and 230Th, and only slightly enhanced leaching of older damage as evidenced by a small increase in the dissolution rate of 234U relative to 238U. These experiments demonstrate that individual alpha-recoil tracks are preserved for some time as disordered regions of higher chemical reactivity in already fully-damaged, aperiodic structures. We estimate an annealing time for the alpha-recoil tracks of 2,000 ± 1,300 yr in the metastable, aperiodic structure. Similar alpha-recoil tracks should be formed and annealed in other aperiodic nuclear waste forms, such as borosilicate glass.

The uranium site in naturally occurring metamict minerals of the pyrochlore group (A1–2B2O6Y0–1) has been investigated using x-ray absorption spectroscopy (XAS). Pyrochlore structures are common phases in proposed polycrystalline waste forms. Betafite, a member of the pyrochlore group (B = 2Ti ≥ Nb+Ta), exhibits U-O bond lengths of 1.94 and 2.37 A for the metamict state, and 2.03 and 2.51 A for the crystalline (annealed) state. The U-O bond lengths decrease (∼0.1 A) and there is a disruption in the second nearest neighbor periodicidy as material is converted from the crystalline to the metamict state.

In order to provide basic data for the calculation of the dose burden from a repository for spent LWR fuel in a salt dome, a formula is given for determining activity release into the surrounding salt brine under accident conditions. The main characteristic of this source term is a time-independent, instantaneous release of part of the activity and mobilization of another part at a continuous and temperature dependant rate, using nuclide-specific leach rates. Changes in leach rates resulting from reduction of the waste surface during mobilization and precipitation are taken into account. The model also takes into account axial and radial activity distribution in the fuel rods.

The effects of the radiolysis of water exposed to mixed alpha- and beta-radiation originating from spent fuel have been calculated. The water is assumed to have penetrated the copper canister and fuel cladding, and then to exist as a 30 μm thin surface film on the fuel pellets.

The combined effects of alpha and beta-radiation, and the presence of iron in the water have been found to be important parameters. The beta-radiation will lower the yield of hydrogen, as will the presence of low concentrations of iron ions. The most likely conditions for water exposed fuel give rise to a total production of 1 mol H2 per m2 fuel surface after 1 million years. A stoichiometric amount of oxygen is also formed.

Calculations have also been carried out in connection with experiments on alpha-radiolysis from Am-241 or from spent fuel pellets. In the case of alpha-radiolysis from Am-241 it was calculated that a 20 cm2 thin layer of 4 mCi Am-241 produced 1.4 × 10−3 cm3 hydrogen per day.

Calculations simulating the conditions of experiments on the leaching of spent fuel have shown that both alpha- and gamma-radiation accelerate the dissolution. The rate is strongly dependent on the dose rate and on the area of the water film covering the fuel surface and fuel fissures.

Iron canisters for high level nuclear waste embedded in compacted bentonite in deep geologic repositories will corrode forming hydrogen gas. The equilibrium pressure (when corrosion would stop) has been estimated to be between 500 and 1000 atm. under repository conditions. As this is much higher than the lithostatic pressure (weight of rock overburden) the gas must be allowed to escape before it disrupts the repository. Escape by diffusion alone is not sufficient but recent experiments have demonstrated that the larger pores in the bentonite are blown free of water and let the gas escape before excessive pressures build up.

The potential effect of a capillary breaking layer (CBL) has been explored. A fine layer nearest the canister (e.q. quartz sand) would have much lower capillary suction pressures than the bentonite clay and would keep the water out as long as there is sufficient overpressure. As long as the CBL is void of liquid water no radionuclides can escape, even if the canister is penetrated.

The application of carbon steel corrosion allowance containers for the long term encapsulation of HLW or spent fuel requires data on the likely rates of corrosion attack so that the metal thickness needed to prevent penetration can be estimated. This paper describes a joint mathematical modelling and experimental approach to the evaluation of the rate of localised corrosion.

Measurements of the corrosion rate of unalloyed steel have been made under conditions representative for repositories in the granitic bedrock of Northern Switzerland using two independent methods: (1) Immersion tests in two representative groundwaters, with and without bentonite; (2) Hydrogen evolution measurements in these groundwaters. The immersion tests were carried out at 80°C and 140°C. In both cases the corrosion rate was higher than 50 μm/a as determined at the end of the first observation period of about 500 h. Corrosion rates of under 10 μm/a were estimated after the first 500 h. The corrosion rates were similar in both groundwaters with mineralisations of approx. 3 g/11 and 14 g/11 respectively, and were generally higher at 80°C than at 140°C. The hydrogen evolution measurements allow an hourly determination of the corrosion rate, with a sensitivity expressed as a corrosion rate of better than 0.1 μm/a. The results generally confirm the observation made in the immersion testing; high corrosion rates were observed over the first few days but then decreased to values well below 10 μm/a. The steady state corrosion rates measured were 1.1 μm/a, 6.5 μm/a, and 2.5 μm/a at 25°C, 50°C, and 80°C respectively in the water with the higher mineralisation. The inverse temperature effect on corrosion rate above 50°C is attributed to a change in the nature of the passive film at higher temperatures.

In applying new conditioning methods dehydrated wastes (e.g. concentrates, ion exchange resins) from operation of nuclear power plants and wastes from decommissioning (e.g. highly-activated core components) are conditioned using high integrity nodular cast iron casks without further solidification. Therefore more stringent requirements have to be made for the packaging material. The objective of the investigations was to evaluate the corrosion resistance of the cask material. For this purpose chemical corrosion tests were made applying corrosion media actually considered or postulated in accident scenarios during interim storage or after disposal in the abandoned iron ore mine Konrad or in a salt formation. The conclusion from the corrosion experiments, which have been performed up to now for the basic packaging material, indicates that nodular cast iron casks with wall thicknesses about 200 mm seem to ensure good protection for wastes especially from nuclear power plants with relatively short-lived radionuclides.

In order to prevent a possible return of activity to the biosphere, the current concept for High Level Waste disposal is to build a multi barrier system consisting of a slow leaching waste form, the waste container, backfilling materials and the geological formation. The main threat to the integrity of the container is aqueous corrosion and two concepts are currently accepted, one based on a thin “corrosion resistant” overpack, the other on a thick “corrosion allowance” canister.

Light optical and electron microscopy have been used to study both laboratory-scale and full scale hot isostatically pressed (HIPed) copper as employed for encapsulation of spent nuclear fuel elements. Particular attention has been paid to the degree of densification achieved, as well as to the status of the joins between the sleeve, lid and the HIPed powder. The work includes the development of computing models for predicting the optimal conditions for HIPing copper and other possible encapsulation materials such as lead and alumina. Mechanical testing of HIPed samples has also been carried out.

The current method of using hydrogen to reduce oxides at powder surfaces prior to HIPing is questioned on the basis that metallography reveals definite signs of hydrogen embrittlement. An alternative method of reduction is to use CO/CO2 mixtures instead of hydrogen, and it is suggested that this approach could be strongly advantageous in future.

This paper describes the work on cement paste development and short-term leaching tests in Standard Canadian Shield Saline Solution (SCSSS) in the presence of bentonite at 150°C. It has been found that:

-supplementary cementing materials such as silica fume or fly ash could significantly improve the properties of sulphate resistant portland cement (SRPC), in particular, permeability to water and pore size distribution.

-the addition of bentonite suppressed the normal tendency of the pH of groundwater to increase rapidly in the presence of cement.

-the presence of bentonite increased the release of potassium ions from the cements.

-– SRPC blended with 20% silica fume resulted in a groundwater pH lower than that of SRPC, with and without bentonite. Moreover, its cumulative fraction of release of potassium was significantly lower than that of SRPC.

In this paper an optimization of concrete container composition, used for storing low and intermediate level radioactive waste from nuclear power plants in Yugoslavia, is presented.

Mechanical properties 37−52 MPa, permeability 1.07. 10−13 - 1.50. 10−11cm2 and leakage rate 3.66. 10−6 - 1.77. 10−4 cm/d for concrete made of commercial materials, were tested.

An experimental investigation has been made of the corrosion behaviour of steel and Zircalloy 2 in deaerated synthetic pore solution and of steel in deaerated mortar. Zircalloy 2 was found to be passivated at all values of pH and electrochemical potentials investigated. Its passive corrosion rate was of the order of <0.01 μm/year. The steel exhibited active corrosion rates of a few μm/year with hydrogen evolution as the corresponding cathodic reaction. The measured free potentials indicated, however, that the hydrogen partial pressures were less than 1 atm.

Titanium is one of the materials proposed in the Swedish programme for the final containment of spent nuclear fuel. In the present investigation, the final repository environment was simulated on the laboratory scale by embedding titanium and a Ti-Pd alloy in dense, water-saturated bentonite clay. The temperature was 95°C and the exposures lasted between 4 months and 2 years. Analysis was performed using ESCA combined with ion etching.

The reaction products formed on the surface consists of TiO2 Montmorillonite - the main constituent of bentonite - is incorporated in the oxide. Suboxides exist near the oxide/metal interface.

The oxide thickness is in the range 70–100 Å. The oxide growth between 4 months and 2 years is small. No significant influence of Pd could be noted. If it is assumed, that the oxide growth follows a logarithmic law, the expression giving the thickest oxide is y = 5.5 In t (where y is the oxide thickness (Å) and t is the exposure time (s)).

Two series of experiments have been carried out In an attempt to derive a source term for the release of some radioisotopes from vitrified high-level waste after disposal In hard rock. In the first type of experiment, the repository was modelled In miniature and the release of the isotopes was measured after they had leached from the glass and diffused through the backfill. In the second type. the “solubilities” of the isotopes were measured In the presence of the constituents of the repository.

We call attention to the possible importance of pH-dependent speciation in determining the distribution coefficient of americium in a granitic ground water. We obtained experimental distribution coefficients for americium. We used the geochemical code PHREEQE to calculate the amount present of americium complexes with hydroxides and carbonates over a range of pH, using formation constants taken from the literature and one estimated for AmOHCO3, a neutral complex suspected of playing an important role in the aqueous chemistry of americium, although not experimentally evaluated. Combining the above inputs, we calculated individual distribution coefficients for americium complexes. We are able to show that the batch distribution coefficient for americium can be seen as the sum of individual distribution coefficients of various complexes.

In the safety analysis recently reported for a potential Swiss high-level waste repository, radionuclide speciation and solubility limits are calculated for expected granitic groundwater conditions. With the objective of deriving a more realistic description of radionuclide release from the near-field, an investigation has been initiated to quantitatively specify the chemistry of the near-field. In the Swiss case, the main components of the near-field are the glass waste-matrix, a thick cast steel canister horizontally stored in a drift, and a backfill of highly compacted bentonite.

Based on available experimental data, an ion-exchange model for sodium, potassium, magnesium, and calcium has been developed, in order to simulate the reaction of sodium bentonite backfill with groundwater. The model assumes equilibrium with calcite as long as sufficient carbonates remain in the bentonite, as well as quartz saturation. The application of this model to the reference groundwater used in ‘Project Gewaehr 85’ results in a significant rise in pH (by up to 3 units) as well as a marked increase in the carbonate concentration.

Neptunium and plutonium speciation and solubility limits are calculated for the reference groundwater chemistry gradually altered to that of saturated bentonite water and back again by a water exchange cycle model. The solubility limits estimated in this way generally turn out to be higher for the bentonite water than for the reference groundwater, mainly due to carbonate complexation of the actinide components AnO2+ and AnO22+. Uncertainties are particularly large for neptunium solubility due to its strong Eh dependence in bentonite water.