X-Ray Absorption Fine Structure (XAFS) spectroscopic studies were performed on a suite of compositionally identical Pu-doped simulated waste glasses prepared 15 years ago at different aactivities by varying the 239Pu/f238 Pu isotopic ratio. The resulting α-activities range from 1.9 × 107 to 4.2 × 109 Bq/g. These samples have a current, accumulated dose that ranges from 8.8 × 1018 to 1.9 × 1018 αdecays/g. A second suite of polycrystalline zircon samples that were synthesized 16 years ago with 10.0 wt.% Pu was also investigated. The 239Pu238Pu isotopic ratio in these samples resulted in α-activities of 2.5 × 108 and 5.6 × 1010 Bq/g and an accumulated dose of 1.2 × 1017 and 2.8 × 1019 α-decays/g. The multicomponent composition of the simulated waste glass permitted XAFS investigation at six elemental absorption edges. For both the glass and ceramic waste forms, initial analysis of Extended X-Ray Absorption Fine Structure (EXAFS) and X-Ray Absorption Near Edge Structure (XANES) indicate that the local environment around the cations exhibit different degrees of disorder as a result of the accumulated a-decay dose. In general cations with short cationoxygen bonds show little effect from self-radiation where as cations with long cation-oxygen bonds show a greater degree of disorder with accumulated a-decay dose.

The objective of this study is to evaluate the effects of water redox conditions and of the presence of iron corrosion products (magnetite) on the release of actinides from HLW glasses. Static corrosion tests were performed on a simulated HLW glass doped with Pu and Np in deionized water in the presence of magnetite under oxidizing and reducing conditions. The tests under oxidizing conditions were performed in air, while the tests under reducing conditions were performed in mixed gas (Ar+5%H2) atmosphere.

The following results were obtained: (1) The presence of magnetite enhances formation of colloids containing Pu and Np in the leachates under both redox conditions. (2) Redox conditions have no remarkable influence on the release of Pu. (3) The reducing conditions combined with the presence of magnetite cause a decrease in the concentrations of Np dissolved species, which is probably the result of the reduction of Np(V) to Np(IV) at the magnetite surface.

Research on synthetic zircons fabricated in system ZrO2-UO2-SiO2 at T = 400 - 750° C, PH20 = 100-200 MPa, and Ni-NiO oxygen buffer was made in order to estimate the limits of uranium incorporation into the phase's structure. It was shown that up to 27 wt.% of uranium may locally enter into the synthetic zircons. SEM/EDS study of the samples has showed that the element is uniformly distributed within the zircon crystals. Incorporation of U into zircon increases its unit-cell dimensions up to α = 6.70 A and c = 6.00 A relative to 6.60 and 5.98 A for a pure phase without impurities.

After thermal high temperature treatment of the samples under 1400° C, in air, for 6 hours inhomogeneous distribution of uranium within the zircon grains was revealed. Very small uranium rich spots probably corresponding to a newly formed phase were observed in the zircon crystals. Contents of the element throughout the grains decreased to about 10 wt.%. It is probably caused by destruction of zircon-coffinite solid solution formed at hydrothermal conditions in the course of its re-heating.

Study of interaction of natural zircons with alkaline carbonate solution at 100° C shows that U release from the phase can reach up to 6 % of total uranium in the samples even under relatively short term interaction. Loss of uranium rises with the element content (or amorphisation degree) and run duration, as well as with grain size reduction. Research on uranium distribution in natural zircons picked up from altered and weathered rocks is characterised by strong depletion of U at grain margins relative to the centre of the crystals. These results indicate that zircon may lose some uranium and, probably, the other actinides with time due to its lattice destruction under radioactive decay followed by hot water attack.

We have designed an apparatus which simulates the conditions of a deep geological repository in order to study the behaviour of spent fuel under leaching. A spent fuel and a Simfuel have thus been leached by synthesized clayey or granitic groundwater in sandy clay (90 % sand) or granite, in reducing conditions, at 90°C and 40 bars.

The apparent leach rate for spent fuel in clayey water is 3.33 µg.m−2.dr−1 in the presence of clay and 3.37 µg.m-−2.d−1 in the presence of granite. The apparent leach rate in granitic water is slower, being 0.37 µg.m−2d−1 in the presence of clay and 0.74 µg.m−2d−1 in the presence of granite.

For Simfuel, the apparent leach rate in clayey water is 7.4 µg.m−2.d−1 in the presence of clay and 1. 1 µg.m−2.d−1 in the presence of granite, which is the same order of magnitude as that for spent fuel. In granitic water, the apparent leach rate is 20 to 40 times greater than that for spent fuel. It is 14.8 µg.m−2.d−1 in the presence of clay and 18.5 µg.m−2.d−1 in the presence of granite.

The chemical durability of aluminosilicate glasses was investigated experimentally between 90 and 200°C. In order to evaluate their potential for containment of minor actinides, these glasses were doped with Nd (to simulate the presence of trivalent actinides) or U. The proportions of the glass network formers, Si and Al, were near those found in basaltic glasses (tholeitic end members). The differences between the chemical compositions allow the clarification of the influence of glass network modifier elements on the chemical durability of silica glasses. More precisely, we tested the ability of Ti, Zr, Nd and N to potentially improve the chemical durability.

Two types of leach tests were conducted:

Dynamic leach tests to determine the initial dissolution rates at 90, 150 and 200°C and the activation energy (Ea) of the glass dissolution reaction;

Static leach tests at high SA/V (200 cm−1) and 90°C to determine the long-term alteration rates and the apparent silica solubility.

For all the glass compositions tested, the initial dissolution rates at nearly neutral or weakly basic pH are similar at the same temperature. Consequently all aluminosilicate glasses have the same activation energy, 60 kJ/mol. This suggests that the initial hydrolysis mechanism is controlled by the breakdown of Si-O and/or Al—O bonds, whatever the nature of glass network modifier elements.

On the other hand, modifier elements have major effects on the formation of protective films at the glass surface. The nature of this protective film depends on the chemical composition of the glass. A phenomenon similar to metal passivation, has been observed on a glass highly enriched with Nd (53.8 oxide wt%) which may lead to alteration rates in undersaturated media up to two orders of magnitude lower than those observed for basaltic glasses. This “passivation” effect disappears when a sodium sulfate solution (a Nd complexing agent) is used as a leachant. Nitrogen does not improve the chemical durability of aluminosilicate glasses. Finally, all these glasses have the same low dissolution rate (10−4 g/m2/day) under saturated conditions.

Preliminary analytical and transmission electron microscopy (AEM and TEM) results for a small suite of natural perovskites are reported in this paper and discussed in relation to previous work. We show that perovskite compositions in Synroc and tailored ceramics plot within the known fields of natural perovskite compositions. AEM analyses and electron diffraction work on selected samples indicate that they are predominantly stoichiometric variants of the cubic perovskite structure. Geochemical alteration was observed in one sample of loparite from Bratthagen, Norway. The primary result of this alteration was leaching of Na from the A-site. Although sufficient alpha-decay dose levels for complete amorphization are not realized in this suite of samples, the available data bracket the beginning of the crystalline-amorphous transformation at doses that are ∼ 2-4 times greater than those of zirconolite of similar age. These results may be due to fundamental differences in the damage annealing rates of perovskite and zirconolite.

Dose-age relationships have been determined by analytical and transmission electron microscopy (AEM and TEM) for the onset dose and critical amorphization dose of a suite of natural zirconolites. Together with a preliminary investigation of the thermal histories of the zirconolite-bearing rocks, the results indicate that valid estimates of D0 (intercept dose at t = 0) and K (annealing rate constant) are obtained and that the host rocks experienced temperatures on the order of 100-200 °C averaged over time. The natural samples are therefore best suited as radiation damage analogues for waste forms stored under deep borehole conditions where temperatures of 100-450 °C are expected. Our results indicate that the critical amorphization (or saturation) dose of zirconolite will increase by a factor of approximately 2 or more as a result of storage at elevated temperature. The effect of long-term annealing on the critical amorphization dose is only important for times in excess of 107-108 years due to the low annealing rate constant of 10−9/yr.

Sintered silicate glass waste forms with a continuous glass phase and embedded but not completely dissolved waste particles were prepared by powder technology. We used a typical defense waste compositions, containing 38 wt.% of ZrO2 and prepared sinter glass samples with 30, 35, and 40 wt.% waste loading at 830°C and 27.6 MPa, by hot isostatic pressing. The glass former was amorphous silica powder. Simulated waste was added as calcine. Samples were characterized using scanning and transmission electron microscopy. The results show that extensive sintering took place, particularly with the 40 wt.% samples. Waste components, such as Na2O, CaO, MnO2, La203, Fe2O3, NiO, Cr2O3, and P2O5, dissolved completely in the continuous glass phase. ZrO2 was partly dissolved and occurred as aggregates of tiny baddeleyite crystals. The porosity of the samples was 2-3%, indicating that the optimum sintering conditions have not been met to reach porosities <1%.

Chemical durability tests were carried out with sintered glass with 35 wt.% waste loading in deionized water at 90°C under static conditions, showing that the glass is at least as chemically durable as a borosilicate glass with 31 wt.% melted at 1350°C.

We model the dissolution of a soluble matrix protected by a diffusion barrier. The dissolution, driven by a first order kinetics law, allows the solute to diffuse in the barrier with sorption and precipitation interactions. The dissolved mass evolution and the lifetime of the wasteform can be expressed in a closed form and as a function of relevant length scales, reflecting the phenomenon controlling the dissolution. Application to a nuclear glass in a clay barrier shows that diffusion in the barrier is the main phenomenon unless precipitation of chalcedony occurs or silicon is very strongly sorbed in the barrier. This requires further experimental investigation of silicon/clay interaction.

ANSTO has studied material resulting from an In Situ Vitrification (ISV) trial on contaminated soil from the Maralinga nuclear weapons test range in Australia. The investigations included determining the distribution of Pu and α-activity in the intermediate scale melt, leach testing of sub-samples from the melt, and physical characterisation of the ISV material, prior to and after leaching. The results show that Pu is well mixed though the whole vitrified/ceramic block from the ISV trials at Maralinga and the durability of the vitrified/ceramic product ensures that the Pu is satisfactorily encapsulated.

Most of performance assessment models for spent fuel repository safety consider radiolysis-self-oxidation to describe fuel matrix release. Nevertheless, due to radioactive decay, the matrix dissolution process under reducing conditions would be controlled by the solubility limit of the steady U solid phase. In this work, leaching behaviour under anoxic and reducing conditions of spent fuel unirradiated chemical analogues (natural U02 and SIMFUEL) in simulated groundwater is studied.

The trial procedure was performed taking into account the possibility that the uranium oxide to be leachated had an initial outer layer with an oxidation state higher than the matrix. This oxidised layer would produce an overestimation on U concentration in solution for the solid studied. In order to avoid this effect, a complete replacement of the leaching solutions was carried out after several days of experimentation. After this initial experimental step, the steady state concentrations obtained in all tests were more than one order of magnitude lower than before. Uranium concentrations found in reducing and anoxic experiments for both U02 and SIMFUEL tests were very close. This fact is attributed to similarity in environmental conditions (pH, Eh, etc.). From that, it can be assured that steady state concentration obtained is independent of solid leached (UO2 or SIMFUEL). In order to assess which is the solid phase that could control the solubility of U, the experimental concentration obtained was compared with results from geochemical code EQ3/6. According with the theoretical calculations U4O9 would be the controlling pure phase formed in whole experimental tests described in this work. Comparisons with bibliography data from leaching experiments of spent nuclear fuel were made as well.

SUGAR (Simplified Understading of Glass And environment Reactions) is a one dimensional (lD) reactive transport code which sequentially couples dissolved chemical species transport and sorption with, chemical speciation, mineral precipitation and kinetic controlled glass dissolution. SUGAR also has the capability to model radioactive decay and steel corrosion. This paper describes the computer model and presents results of model testing against experimental data. The capabilities of the model are demonstrated by a preliminary investigation of the geochemical evolution of glass in crystalline rock formations.

Zirconolite-rich ceramics, containing strontium (∼1 wt.%) or uranium (5-20 wt.%) were prepared by cold pressing of mechanically activated oxide powders and sintering in air at 1200 TC for 2 hours. They were examined with XRD, optical and electron microscopy (SEM/TEM/EDS) methods. Zirconolite content in the samples was found to be as much as 85-90 vol.%, Baddeleyite, rutile, and perovskite (in the Sr-containing sample only) were present as minor phases. In the U-containing samples, two zirconolite varieties with different Zr/U ratio were found: zirconolite-1 and zirconolite-2 with average compositions Ca0.90Zr0.95U0.10Ti2.00O7.00, and Ca1.15Zr0.40U0.50Ti2.02O7.00 respectively. The amount of zirconolite-2 increases from ∼1 vol.% in the sample with the lowest uranium content to >50 vol.% in the sample containing ∼20 % U02. In the latter sample about 5 vol.% of cubic pyrochlore-type phase was also present.

Since the early of the 1990s the method of inductive melting in a cold crucible (IMCC) has been applied at SIA “Radon” for production of various wasteforms, including glasses and Synroc-type ceramics. Sphene-based glass-ceramics composed of glass and crystalline phases were considered as appropriate wasteform for High Level Waste immobilisation. Investigation of two glass-ceramic specimens prepared with the IMCC has been performed using optical microscopy, XRD, SEM/EDS, and TEM methods. The samples produced consist of vitreous and crystalline phases. The vitreous phase consists of two varieties of glass formed by the immiscibility of the initial melt onto two separate liquids. One of the glasses is observed as spherical microinclusions in the matrix glass. The glass of the microspheres are differed from the matrix glass composition by higher contents of Ca, Ti, Ce, Sr, Zr (or Cr), while the matrix glass contains higher amounts of Si, Al, and alkalies. The crystalline phases with sphene- and perrierite-like structures have been also occurred. Their total quantity reaches up to 50 vol.%. The synthetic perrierite has similar unit-cell parameters with its natural mineral analogs with the only exception in two-fold value of c dimension. Zr, Ce, and Sr are incorporated into synthetic sphene and perrierite, while Cs is hosted by the glass phases.

Calculation of used nuclear fuel dissolution rates in a geological disposal vault requires a knowledge of the redox conditions in the vault. For redox conditions less oxidizing than those causing UO2 oxidation to the U3O7 stage, a thermodynamically-based model is appropriate. For more oxidizing redox conditions a kinetic or an electrochemical model is needed to calculate these rates. The redox conditions in a disposal vault will be affected by the radiolysis of groundwater by the ionizing radiation associated with the fuel. Therefore, we have calculated the alpha-, beta- and gamma-dose rates in water in contact with the reference used fuel in the Canadian Nuclear Fuel Waste Management Program (CNFWMP) as a function of cooling time. Also, we have determined dissolution rates of UO2 fuel as a function of alpha and gamma dose rates from our electrochemical measurements. These room-temperature rates are used to calculate the dissolution rates of used fuel at 100°C, the highest temperature expected in a container in the CNFWMP, as a function of time since emplacement. It is shown that beta radiolysis of water will be the main cause of oxidation of used CANDU fuel in a failed container. The use of a kinetic or an electrochemical corrosion model, to calculate fuel dissolution rates, is required for a period of ∼ 1000 a following emplacement of copper containers in the geologic disposal vault envisaged in the CNFWMP. Beyond this time period a thermodynamically-based model adequately predicts the fuel dissolution rates. The results presented in this paper can be adopted to calculate used fuel dissolution rates for other used UO2 fuels in other waste management programs.

Zircon, ZrSiO4, is a prospective durable host material for the immobilization of excess weapons plutonium. Using cerium as a chemical analogue for plutonium, the experiments on the synthesis of Ce-doped zircon were conducted by sintering of sol-gel precursors in air and vacuum. The results showed that cerium substantially promotes zircon formation from sol-gel precursors and sintering in air is preferable for cerium incorporation in zircon structure. Based on measured lattice constants, solid solution compositions (Zr0.96Ce0.04)SiO4 and (Zr0.98Ce0.02)SiO4 were formed in samples sintered in air at 1400°C and 1600°C, respectively. The solubility limits of cerium and actinides in zircon and mechanism of zircon formation are discussed.

Synroc/glass composites were designed for simulated Hanford HLW sludges containing U (the current “All-Blend” formulation). The composite contained ∼ 50 wt% of simulated HLW (oxide equivalent), to which ∼ 6, 10, 10, and 24 wt% of CaO, Al2O3, TiO2 and SiO2 were added, and melted under an argon atmosphere at 1350°C. The phase assemblage consisted of zirconolite, perovskite, spinel, nepheline, whitlockite and glass as major phases. Seven-day PCT tests yielded values of < 0. 1 g/m2 for all elements studied. The PCT results were tolerant to changes of ∼ 20% of the inventories of the additives, and to variations in redox conditions.

Technetium separated out during decontamination of liquid Hanford wastes can be incorporated as metal in hot-pressed Synroc prepared under reducing conditions, and its leach resistance is good (∼ 10∼3 g/m2/day at 70°C), and can be improved by alloying with iron group metals. With a choice of “neutral” (P(O2) ∼ 10−4 atm., near the Ni/NiO buffer) hot-pressing conditions, Tc can also be incorporated as Tc4+, substituting for Ti4+ in the ceramic phases, and in this form, it should be highly leach resistant also.

Sorption experiments were performed with iodide, cesium and strontium ions on MX-80 Wyoming-bentonite treated with hexadecylpyridinium (HDPy+) in amounts equivalent to 0.2 - 4.0 times the cation exchange capacity (CEC) using 125I- 134Cs+ and 85Sr2+ as tracers. In HDPy-bentonite, iodide exhibited increasing adsorption, while cesium and strontium ions showed decreasing adsorption with increasing organophilicity. It was also found that the Cs+affinity to original and HDPy-bentonite was considerably higher than that of Sr2+ ions. HDPy+ uptake in increasing concentrations resulted in a pronounced expansion of the basal spacings (d002 reflex at 2.78 nm) and in a change of the negative particle surface charge to positive values.

The hydraulic conductivities in water saturated bentonites at different densities were measured within temperature range of 20 to 80 °C. The results show that the hydraulic conductivities increase with increasing temperature. The hydraulic conductivities of bentonites at the temperature of 80 °C increase up to about 3 times as high as those at 20 °C. The measured values are in good agreement with those predicted. The change in viscosity of water with temperature contributes greatly to increase of hydraulic conductivity.

Production of carbonate minerals will be a dominant alteration reaction during cement degradation in the near-field of the L/ILW repository planned at Wellenberg (Switzerland). Very large amounts of calcite are predicted to precipitate as a result of (a) the interaction between bicarbonate-rich ground water and cement phases and (b) the in-situ degradation of organic waste components in the repository.

A preliminary model was developed to estimate to which extent coprecipitation with calcite could reduce radionuclide concentrations in solution. Since the main objective was only to determine the order of magnitude of this effect, a very simple, closed-system model was considered to be sufficient. The model assumes instantaneous dissolution of the entire radionuclide inventory in the water-saturated pore space of the repository caverns and simultaneous reversible sorption on the cement. Coprecipitation of radionuclides with calcite and residual solution concentrations are calculated assuming in-situ precipitation of a given amount of calcite and using nuclide-specific partition coefficients.

The model calculations show that the solution concentration of several radionuclides (tetravalent and trivalent actinides, Sr and Ni) is likely to decrease by orders of magnitude as a consequence of coprecipitation with large amounts of secondary calcite. Further, radionuclide coprecipitation is found to depend on the extent to which the same radionuclide is sorbed on cement phases. The results confirm that coprecipitation with calcite will be an important process in determining concentration limits for dissolved radionuclides in the Wellenberg repository.