During the initial period after disposal, in which no mass transfer occurs with environmental materials, spent fuel undergoes modifications mainly arising from the effects of α self-irradiation damage. To simulate this aging phenomenon and estimate the consequences on the matrix, an experimental study was carried out in which old UO2 samples doped with short-lived actinides were characterized and analyzed. Changes that occur rapidly at the beginning of the aging process were identified in two lots of actinide-doped samples. Volume and microscopic swelling increased, reaching a maximum relative level of 1.5%. At the same damage level, microscopic swelling remained below 0.9%, suggesting that helium formation and atomic defects have a role in this regard. Similarly, the sample microhardness rapidly increased by up to more than 12%. Characterizing 238PuO2 sources showed at very high integrated and instantaneous doses, irradiation damage and helium accumulation at the grain boundaries resulted in grain decohesion or even powder formation. Helium mobility appeared to be modified by the presence of defects, and local implantation experiments were devised in the CEA's Pierre Süe Laboratory at Saclay to quantify this thermally enhanced diffusion component. The physical condition of the sintered PuO2 pellets after 30 years of storage revealed a purely athermal diffusion mechanism, depending on the sample α activity.

The Swedish programme for disposal of high level nuclear waste includes the development of a container which comprises a cast iron load bearing canister contained in a 50 mm thick copper corrosionshield.

The temperature of the outside of the canister is likely be up to 100°C, and repository processes may lead to long term loads of up to 50MPa. Creep of the copper is therefore an issue. SKB (The Swedish Nuclear Fuel and Waste Management Co.) have been conducted tests on three materials, OF copper containing 10ppm sulphur (OF1), OF copper containing 6ppm sulphur (OF2) and OF copper containing 6ppm sulphur and 50ppm phosphorus (OFP). In order to improve confidence in extrapolation of test results to practical temperatures and stresses, it is desirable to support the extrapolation procedures with a physical model.

Life predictions made using the Frost and Ashby model [5] together with a simplification of the Cocksand Ashby model [6] provide good agreement with published experimental data for OF2 and OFPmaterials. Extrapolation of this data using the model leads to life predictions of 12,000 years for the OF2material and 120,000 years for OFP material. The prediction for OFP depends on an assumption that an observed strengthening mechanism conferred by phosphorus at high stresses and temperatures, is equally as effective under repository conditions.

OF1 material fails after a relatively short life and by a different mechanism to OF2 and OFP. The changein mechanism may be explained, using the model [6], for cases where segregating species reduce surface energy of grain boundary voids. It is suggested that in this case the segregating species is sulphur.

For the construction of the geological disposal system of radioactive wastes, a large amount of cement is used. Such a material alters the pH of groundwater to highly alkaline region [1-3]. The highly alkaline plume contains rich Ca ion compared to the surrounding environment, and the Ca ions reacts with soluble silicic acid. Its product would deposit on the surface of flow-paths in the natural barrier and decrease the permeability. In this study, the influence of Ca ion in highly alkaline plume on flow-paths has been examined by using packed bed column.

The column was packed with the amorphous silica particles of 75-150 μm in diameter. The Ca(OH)2 solution (0.78 mM, 4.37 mM, and 8.48 mM, pH:12.2∼12.3) was continuously injected into the column at constant flow rate (1.1 mm/s), and the change of permeability was continuously monitored. At the same time, [Ca]total and [Si] in the eluted solution was measured by the inductively coupled plasma atomic emission spectrometry (ICP-AES). The Ca(OH)2 solutions were prepared with CO2-free pure water, and filtrated through 0.22 μm and 0.45 μm filter. The permeability was normalized by the initial permeability value.

The experimental results showed that the permeability change and the chemical reactions in the packed bed depended on the concentration of Ca(OH)2 in the feed solution. The SEM images showed that the feed solution of 8.48 mM in [Ca]total apparently altered the particle surface with the products with Ca (deposition). Moreover, from the distribution of Ca on the cross section of packed particle (by EPMA), it was found that Ca deposited also in the inner pores of the particle. These results suggest that the deposition does not always change the permeability in the early stage of precipitation reaction. In the eluted solution, the concentrations of Ca and Si dramatically changed, depending on the concentration of Ca in the feed solution.

The vitrification of nuclear waste salts similar to the salts present in Eurobitum produced by the former Eurochemic reprocessing plant, Mol-Dessel, Belgium, is investigated. In order to approach the composition of the high-level waste reference glass SON68, mixtures with different ratios of glass frit to waste salts have been prepared and vitrified. The concentration of the waste salts varied between 17 and 25 % by weight.

To compare the short term corrosion behaviour with reference waste glass, leach tests have been performed. During 8 weeks, the glass samples were submitted to leach tests at 363 K using Milli Q water as leachant, at a glass exposed area over leachant volume ratio equal to 100 m−1.

During this short period, scanning electron microscopy with energy dispersive X-ray analysis (SEM-EDX) showed no gel layer formation. This technique did not detect changes in the chemical composition. However, X-ray photoelectron spectroscopy (XPS) showed that the binding energy of the Zr 3d and Si 2p electrons decreased as a function of leaching time, indicating an increasing covalent character of the glass surface. Inductively coupled plasma-mass spectrometry (ICP-MS) of the leachate as a function of time shows that ionic species such as Na and K leach faster than e.g. Si and B. These results together with the XPS data suggest the onset of a gel layer formation. The effect of waste content in the glass on the leaching rate is also studied. The results are compared with similar leaching tests on R7T7 type SON68 glass. Most elements are better retained in the simulated Eurobitum glass than in the SON68 glass.

Long-term field tests of cemented aqueous radioactive wastes in an experimental mound type surface repository were carried out at Moscow Scientific and Industrial Association “Radon” from 1965 to 2004. Aqueous radioactive wastes of different compositions containing short-lived radionuclides including 90Sr and 137Cs at concentrations from 0.34 to 1.8 MBq/L were immobilized using cementation technology. Water solution to cement ratio was 0.66, grout mixing time 10-15 minutes, and cement paste hardening time 7 days. 73 cement blocks with a volume of 0.027 m3 were disposed of for long-term tests in a simple mound type surface repository. The atmospheric precipitates, which contacted radioactive cement blocks, were collected and analyzed for the content of radionuclides. In August 2004 the experimental repository was opened, cemented blocks, underlying and covering materials were retrieved for analyses. XRD analyses showed that along with amorphous tobermorite gel the main crystal phases in cements are calcite and portlandite. Both visual inspection and radiometric analyses demonstrate that cemented blocks are in good condition and that the cement paste has retained radionuclides from the wastes. Thus after 39 years of storage in the mound type repository the cemented aqueous wastes are reliable immobilized.

Cementitious materials have been studied for their applicability as support and lining in the construction of shafts and tunnels of the underground facilities for HLW disposal system. However, in the case of Ordinary Portland Cement (OPC), the buffer materials and the host rock may be altered due to the high alkalinity of groundwater contacting the OPC in the long-term. Low alkaline cements with high content of pozzolanic material were developed in order to reduce such hyper alkaline alteration. This paper shows that the pH of the leaching solution of this cement is about 11, and that it can be applied for full scale lining and for shotcrete.

The immobilisation of strip product (TUE + REE), collected at Russian research centers and laboratories during R&D on partitioning technology for liquid high–level radioactive waste (HLW), was studied using simulated solutions.

Ultraporous glass crystalline material “Gubka”(G) and multiple processing stages were used to immobilize the simulated strip product. The processing consisted of the following stages: Gubka saturation by solution – drying into Gubka pores – compaction by cold uniaxial pressing followed by sintering (CUPS) or hot uniaxial pressing (HUP).

Over the last decades, several models describing glass dissolution have been published. Starting from the basic equations in their simplest form, the relationships between models are evaluated to address the following questions: 'What is the relationship between their basic assumptions?, What is the resulting long term leach rate? and, Which element profiles do they allow one to predict?' Although not part of it, this paper could complement the European project GLAMOR, where two models describing the dissolution of glass in a water solution, the r(t) and the GM models, are used by several groups to fit the same sets of experimental data. In this paper, other models are considered as well and all models are compared with each other.

From comparison with the Boksay model, which uses the same equations, a simplification is suggested for the GM model concerning the water diffusion in the glass. The use of the numerical code developed to solve part of it, can mostly be avoided by using the analytical solution of the Boksay model.

One of the waste forms resulting from the Belgian programme for nuclear energy is the R7T7 high level waste glass, simulated by the SON68 glass. When exposed to distilled water or synthetic interstitial clay water, SON68 glass dissolves incongruently, releasing some elements (B, Li, Na) faster to the solution than others (Si, Al, Ca). The objective of this study is to describe the composition of the leachate in contact with inactive SON68 glass by assuming congruent glass dissolution followed by the precipitation of a secondary solid phase represented as an ideal solid solution in equilibrium with the leachate. Experimental SON68 dissolution data in distilled water was available at three temperatures. The solubility of the different end members in the solid solution is optimized using the available data for each temperature. The temperature dependence of the different end members was then obtained by Van 't Hoff equation. The calibrated model can describe the composition of the leachate and the altered glass phase during glass dissolution at different temperatures. The model, calibrated for the distilled water system, could successfully describe the composition of the leachate during glass dissolution experiments in a synthetic clay water system.

European Commission supported a wide research project entitled “Spent Fuel Stability under repository conditions” (SFS) within the 5th FWP, the aim of which was to develop a common understanding of the radionuclides release from spent nuclear fuel in geological disposal and build a RN release model in order to assess the fuel performance. This project achieved by the end of 2004 focuses both on the Instant Release Fraction (IRF) model and the Matrix Alteration Model (MAM).

A new IRF model was developed based on the anticipated performances of the various fuel microstructures (gap, rim, grains boundaries) and the potential diffusion of RN before the canister breaching. However, this model lets the choice to the end-user about the degree of conservativeness to consider.

In addition, fuel alteration has been demonstrated to be linked to the production of radiolytic oxidants by water radiolysis at the fuel interface, the oxidation of the fuel interface by radiolytic oxidants and the subsequent release of uranium under the influence of aqueous ligands. A large set of experimental data was therefore acquired in order (i) to upgrade the current radiolytic kinetic scheme, (ii) to experimentally correlate the fuel alteration rate and the fuel specific alpha activity by performing experiments on alpha doped samples, (iii) to experimentally test the potential inhibitor effect of hydrogen on fuel dissolution. Based on these results, a new MAM was developed, which was also calibrated using the experiments on inactive UO2 samples. This model was finally applied to representative granitic, salt and clayey environment to predict spent fuel long-term fuel performance.

A methodology and computer software is described which can be used to track the inventory of radionuclides as they are affected by various nuclear, physical and chemical processes during reactor, storage, effluent and disposal phases of the nuclear fuel cycle. Such a model is required to provide an assessment of economic, environmental and societal performance indicators which underpin decisions regarding options for the use and management and nuclear materials. An example generic deep repository model is described which can be used to provide an indicator of environmental performance of vitrified high level waste and UO2 and mixed oxide (MOX) spent fuels. The assessment models highlight the significance of the I-129 fission product which necessitates the use of appropriate dose assessment models to be considered for each process step of the nuclear fuel cycle in order that a complete environmental assessment of process options can be determined.

Several kinetic models have been proposed to evaluate the aqueous dissolution/alteration rate of nuclear waste glass for long-term. However, reaction processes controlling the long-term rate are much more subjected to controversy. Temperature dependence of the long-term alteration rate is an essential issue to understand the rate controlling processes. In the present study, the static aqueous alteration tests were performed with a Japanese simulated waste glass P0798 as a function of temperature from 60°C to120°C, and the temperature dependence of the long-term alteration rate was evaluated to understand the rate controlling processes. The tests were performed in 0.001M NaOH solution to maintain a constant solution pH of around 10 during the test period and to provide smectite forming conditions where smectite forms as the major secondary phase without zeolite formation. From the test results on dissolution of boron, the alteration rate at each temperature was analyzed by use of a water-diffusion model. The water-diffusion model used is based on a simple assumption; the glass alteration is controlled by water diffusion with ion-exchange between water (hydronium ion: H3O+) and soluble elements (B, Na, Li, etc) at the glass surface layer with the apparent diffusion coefficient Di. A good agreement was observed between the model analysis and the test results, and the value of Di was evaluated to be 1.2 × 10−22 m2/s at 60°C to 1.8 × 10−21 m2/s at 120°C. The Arrhenius plot of Di showed a good linearity to give the activation energy of 49 kJ/mol, which value is close to that for the residual dissolution rate of French waste glass (53 kJ/mol) by Gin [1], and is very close to that for ion-exchange in sodium aluminosilicate glass (49 kJ/mol) by McGrail [2]. These results suggest that water diffusion with ion-exchange can be the dominant process controlling the alteration rate under smectite forming conditions. At elevated temperatures (100°C and 120°C), however, the model-predicted boron releases deviated from the experimental data at the later stage beyond 50-80 days, which suggests that the alteration layer developing at the glass surface may evolve to be protective against the water diffusion to depress the alteration rate as the alteration proceeds.

Since reprocessing is no longer the reference policy in Belgium, studies on the direct disposal of spent fuel in a clay formation have gained increased interest in the last years. In order to determine to what extent the clay properties and the α-activity influence the dissolution kinetics of spent fuel for the long term disposal, static dissolution tests have been performed on 5 different types of α-doped UO2, representing a PWR fuel with a burn-up of 45 or 55 GWd · tHM−1 and fuel ages ranging between 150 and 90,000 years, in different Boom Clay (BC) media at room temperature and in an anoxic atmosphere for 90 to 720 days. The uranium activity in the clay fraction over time was found to be much higher than the U-activity in the leachates, which has been mainly ascribed to the high retention capacity of the BC. The average dissolution rate between 0 and 90 days obtained for the 5 types of α-doped UO2 were all found to be high and quite similar at ~263 µg · m−2 · d−1and a “longer-term” rate (181 to 720 days) ranging between zero and 15 µg · m−2· d−1. These results suggest that the activity of the fuels does not seem to have an effect on the UO2 dissolution rates under the considered test conditions. In order to study the partition/redistribution of U during UO2dissolution, sequential extraction experiments were performed. Results of the latter have provided a better mechanistic understanding of BC/spent fuel interaction processes as well as information on the role of the different minerals controlling the U-retention/immobilization.

In the ongoing Swedish site investigations it has been found that the rock matrix adjacent to many open fractures has been altered. The alteration can extend from mm to several cm. The altered rock can have different sorption and diffusion properties compared to the undisturbed rock and this may influence the retardation of the nuclides. The paper presents how the Channel Network model has been adapted to handle diffusion into a matrix composed of several layers with different properties in addition to the infinite undisturbed matrix. For one channel, the solution for the Residence Time Distribution, RTD, may be found in the Laplace-plane. For the transport in the Channel Network, a particle tracking technique is used to determine the paths followed by solute particles. The RTD for this path is obtained using convolution, which in the Laplace-plane means multiplication of the transfer functions for each channel. The inversion to the time-plane is made by numerical inversion of the Laplace transforms for each path. The method has been tested with data from the TRUE (Tracer Retention Understanding Experiments) project, Task 6F, fluid flow and solute transport in two features in 100 m scale where a complex matrix was modelled. The model was used to predict the transport of the tracers (I-129), Cs-137, and Am-241) over some 20 m. The paper also addresses how the RTD is influenced by the different retardation mechanisms under Site Characterisation (SC) as well as Performance Assessment (PA) conditions.

To determine the maximum attainable solution concentrations of U, Th and fission products during long-term glass corrosion, co-precipitation studies were performed. The HLW-glass GPWAK 1 was dissolved in highly acid and basic media (100 g glass/L) at 80°C and by preparing acid solutions containing the various elements in soluble form. After dissolution the pH of the solution was slowly lowered/increased and the high concentrated solutions become super-saturated and strong precipitation takes place. The found pH-dependent concentrations of the various elements reflect their different solubility, lowest concentration for Th, Zr and Fe and highest for B, alkalis and alkaline earths. To find out the solid phases controlling the solution concentration, the run of the concentrations of some elements is compared with the solubility data of their pure solid phases (e.g. hydroxides) from literature. It was found that the concentrations of Nd, Th and U in dependence on pH agree quite well with solubility data of AmOHCO3, ThO2/Th(OH)4 and Schoepite UO2·(OH)2) ·H2O. Therefore, it can be assumed that the maximum attainable concentrations of many elements are controlled by such pure solid phases.

This paper focuses on how to extrapolate current knowledge of spent fuel matrix alteration processes from laboratory to repository conditions, i.e., the influence of changes in both the environmental conditions and the range of time scale considered. Therefore, a spent fuel matrix alteration model allowing the alteration rate evolution to be predicted as a function of both the host rock considered and evaluation time scale of interest is described.

At present, the model assumes that alteration of the spent fuel will start when the groundwater reaches the solid surface and that only the radiolytic species of the groundwater (oxidants generated by α-radiation of spent fuel) will produce the surface oxidation process and subsequent matrix dissolution; O2, H2O2 and OH· are the species that react with UO2(s) for oxidation of the pellet surface. The dissolution process of the surface sites that are oxidized is modelled in two steps: first, a surface co-ordination of the oxidized layer with aqueous ligands and, second, detachment (dissolution) of the product species. Taking this mechanism into account, the model gives the evolution of the spent fuel matrix alteration rate over periods as long as 1,000,000 years.

In this work the matrix alteration rate results obtained for two repository environments, granitic and argillaceous, will be presented. Furthermore, a sensitivity analysis study has been performed on the influence of the following variables: type of spent fuel considered, α-dose rate evolution, α-range in groundwater, carbonate and iron concentration in groundwater, H2 partial pressure, container time failure and specific surface area of the pellet.

Naturally occurring radionuclides can also end up in soils and groundwater due to human practices, such as application of certain fertilizers in agriculture. Many mineral fertilizers, particularly (super)phosphates, contain small amounts of 238U and 230Th which eventually may be leached from agricultural soils to underlying water resources. Field soils that receive P-fertilizers accumulate U and Th and their daughter nuclides, which eventually may leach to groundwater. Our objective was to numerically assess U migration in soils. Calculations were based on a new reactive transport model, HP1, which accounts for interactions between U and organic matter, phosphate, and carbonate. Solid phase interactions were simulated using a surface complexation module. Furthermore, all geochemical processes were coupled with a model accounting for dynamic changes in the soil water content and the water flux. The capabilities of the code in calculating natural U fluxes to groundwater were illustrated using a semi-synthetic 200-year long time series of climatological data for Belgium. Based on an average fertilizer application, the input of phosphate and uranium in the soil was defined. This paper discusses calculated U distributions in the soil profile as well as calculated U fluxes leached from a 100-cm deep soil profile. The calculated long-term leaching rates originating from fertilization are significantly higher after 200 years than estimated release rates from lowlevel nuclear waste repositories.

The changes in microstructure and content of water phases during hydration of a 3:1 BFS:OPC blend are investigated by Mercury Intrusion Porosimetry (MIP), freeze-drying, Thermal Gravimetric Analysis (TGA) and 1H Nuclear Magnetic Resonance (NMR) relaxometry. MIP indicates that during the blend hydration, a reduction in the population of capillary pores (larger than about 100 nm) occurs while the population of gel pores (smaller than few tens of nanometres) increases. Between 3 and 90 days, the porosity estimated by MIP decreases from about 36% down to 18% while the median pore size decreases from about 140 nm down to 6 nm.

1H NMR relaxometry shows that after 1 day of hydration, nearly 70% of the evaporable water is held in capillary pores while about 30% is present in gel pores. After two weeks, most of the evaporable water (90%) is found in pores smaller than few tens of nanometres.

The amount of evaporable water detected by freeze drying decreases from less than 20 wt.% after one week of hydration down to about 16.3 wt.% after 90 days while the amount of chemically bound water related to the degree of advancement of the cement hydration and detected by TGA increases from 8 wt.% to 10.3 wt.%.

During hydration the BFS:OPC blend matrix evolves from an open microporous network to one of a poorly connected network of water rich nanopores with increasing amounts of chemically bound water.

For representing the source-term of a proposed repository at Yucca Mountain, NV, the performance assessment (PA) approach evaluates the disequilibrium degradation of the waste forms to capture a bounding rate for radionuclide source-term availability and use solubility constraints that are more representative of longer-term, equilibrium processes to limit radionuclide mass transport from the source-term. These solubility limits capture precipitation processes occurring either as the waste forms alter, or in the near-field environment as chemical conditions evolve. A number of alternative models for solubility controls on dissolved neptunium concentrations have been evaluated. These alternatives include idealized models based on precipitation of simple, discrete neptunium phases and more complex considerations of trace amounts of neptunium being incorporated into secondary uranyl phases that form during waste form alteration. Thermodynamic constraints for neptunium under oxidizing conditions indicate that tetravalent neptunium solids (e.g., NpO2) are more stable relative to pentavalent phases (e.g., Np2O5), and therefore have solubilities that set lower dissolved concentrations of neptunyl species. Within solution, the pentavalent neptunyl ion (NpO2+) and its complexes dominate. Data on solids and solutions from slow flow-through (dripping) tests on spent nuclear fuel (SNF) grains indicate that neptunium is tetravalent in the SNF and that, over the ∼9 years duration of the tests, the dissolved neptunium concentrations are near to, or below, calculated NpO2 solubility. Currently, the NpO2 solubility model is applied at the waste forms within waste packages because the degradation of SNF and corrosion of alloys ensures active reduction reactions. The more conservative Np2O5 solubility model is applied in environments distal to the waste forms (i.e., the invert below waste packages) where reduction of dissolved pentavalent neptunium is less certain. Consideration of Np incorporation into secondary uranyl phases suggests that both of these idealized models will provide conservative estimates of neptunium release to the geosphere.

The presence of organic complexants, such as gluconic acid, in an intermediate-level radioactive-waste (ILW) repository may have a detrimental effect on the sorption of radionuclides, by forming organic complexes in solution. In order to assess this, stability constants are required for the complexes formed with radionuclides at high pH. This study reports the stability constants for the reactions of some divalent metals with gluconic acid (Gl). The metals studied were Cd, Co, Fe(II), and U(VI) at pH 13.3; and Co and U(VI) at pH 7. The constants were measured by the Schubert (ion-exchange) or solubility product methods. Stoichiometries of the complexes were also determined. At pH 7 each complex was of the form M1Gl1, with log β values suggestive of salt formation. At high pH, log β values were between 13 and 20. The constants have enabled speciation calculations to be performed showing the effect of gluconic acid on the metal's solubility, with data for Ni included. Solubility is predicted to increase in the presence of Gl from pH 9 to 13.5, suggesting that it may have an impact on radionuclide behaviour. The largest solubility increases are for Cd and Co, the smallest, U(VI).