High-FeO olivine-rock from Lovasjäarvi intrusion (SE-Finland) has been suggested as redox-active backfill-additive in deep high level nuclear waste repositories. In this w'ork, surface characterization of olivine-rock by X-Ray Photoelectron Spectroscopy (XPS) has been used in order to assess the behavior of this rock during leaching experiments and sorption properties of redox sensitive radionuclides. Uranium (VI) sorption has also been performed. Olivine leaching experiments at basic pH have shown the abundance of Fe and 0 relative to Si and Mg clearly indicating the precipitation of a Fe(III)-oxide secondary phase, probably amorphous FeOOH. At acid pH, the decrease of Mg/Si and Fe/Si ratios can be attributed to the formation of silicon dioxide on the mineral surface. The deconvolution of the 4f7 2 Uranium peak has shown the reduction of U(VI) to U(IV) onto the olivine surface.

This paper reports the coupled thermo-hydro-mechanical (THM) analyses of the “in situ” test carried out during the pre-operational stage of the FEBEX project. The determination of parameters has been performed using the results of the characterization stage and other available information. The results of the analyses provide a good understanding of expected test behaviour. This understanding is enhanced by the performance of a quite extensive programme of 1-D sensitivity analyses in which the values of critical parameters are varied. Two dimensional effects are discussed on the basis of the results of 2-D axisymmetric THM analyses carried out using a longitudinal section that provides a better representation of real test geometry.

Understanding the near-field evolution of a deep geological disposal system for HLW is crucial for assessing its long term performance and, therefore, safety. For that reason ENRESA devised the FEBEX Project, a Full Scale Engineered Barriers Experiment in crystalline rock. The project consists of a full-scale “in-situ” heating test, a large-scale laboratory mock-up and supporting materials tests, and modelling.

Even though the object of the project is to contribute to the search for methods of behaviour and of safety analyses for a repository, other subinvestigations have been/are being included. The stated objectives are to demonstrate the procedures of constructing an engineered barrier system (EBS), especially the fabrication, handling, and installation of bentonite blocks (buffer) at an almost industrial scale, to improve and validate the numerical models for thermo-hydro-mechanical behaviour, and to investigate the geochemical processes that are produced in the buffer including canister corrosion, as well as the generation and transport of gas. Since early 1997, with the commencement of the heating phase, both large-scale tests have been fully operative. At this point it can be said that the demonstration objective of constructing the EBS has been successfully achieved. The measured thermal response of the buffer follows the pattern predicted in the preliminary modelling. The saturation rate of the buffer and associated mechanical processes are being continuously monitored.

In high-level waste repositories, a carbon steel overpack will be corroded by consuming oxygen trapped in the repository after closure. Iron corrosion products are expected to interfere with migration of radionuclides by filling the pore in bentonite and sorbing radionuclides. In this study the apparent diffusion coefficients of cesium and strontium were measured in compacted Na-bentonites (Kunigel VI® and Kunipia F®, JAPAN) contacted with carbon steel and its corrosion products under reducing conditions or without carbon steel under oxidizing conditions for comparison. The apparent diffusion coefficients of cesium with and without corrosion product were 2.2 to 13 × 10−12 m2/s. The apparent diffusion coefficients of strontium with and without corrosion product were 3.1 to 25 × 10−12 m2/s. There were significant effects of dry density (0.8 to 2.0 g/cm3) and montmorillonite contents (50% for Kunigel V1 or 100% for Kunipia F). The presence of corrosion product decreased the apparent diffusion coefficients of Cs in both bentonites and that of Sr in Kunigel V1, especially at low dry density. This may be due to corrosion product filling the pore in the bentonite, decreasing the free pore size and density for diffusion.

For predicting the long-term behavior of bentonite, we present a new and unified simulation procedure of Molecular Dynamics Method (MD) and Homogenization Analysis (HA). The MD is applied to establish molecular-scale bentonite properties and the HA is introduced to extrapolate the molecular model to the bulk-scale continuum model.

This study focuses on demonstrating the feasibility of constructing a large reinforced concrete silo for the disposal of low- and intermediate-level radioactive waste. In a large-scale model test, the results of temperature (from heat of hydration) and thermal stress analyses were confirmed, as was the practicality of countermeasures to prevent thermal cracking. After one year, the average permeability of the model silo was estimated to be below 1×10−11m/s.

This paper presents a fully coupled model for unsaturated deformable materials like swelling clays of engineered barriers. A thermodynamic framework is adopted which allows to take into account the influence of the mechanical variables (stress and strain tensors) upon the hydraulic part of the state equations (suction - water content relation). So, a fully coupled behaviour formulation is stated which, when combined with « classical » phenomenological relations, makes it possible to study the influence of the mechanical state of the barrier on the kinetics of its resaturation.

An attempt is made to decompose some quantities, most commonly integrated over the chemical composition, the structure of the pore network and the time scale, into their differential counterparts. Into this category typically fall, say, the apparent diffusivities of ionic species. To serve this objective, a two-dimensional unsteady molecular-level electrokinetic transport model for ionic species in bentonite clays will be developed. The model incorporates additional features to the conventional Gouy-Chapman (GC) theory: ionic hydration, dielectric saturation, and the volume exclusion of ions. The governing equations for the flow of electrolyte solution through the pores are solved by an iterative numerical scheme to relate the characteristics of the flow to the characteristics of the pores and to the composition of the external solution in contact with the clay. The pore geometry of the clay is modelled as an array of non-interconnected tortuous channels with no parallel or serial-type non-uniformities along the pathway. This roughly corresponds to the picture of clay particles of infinite extent aligned in parallel and spaced apart by a constant distance. The model aims to simulate and interpret equilibrium and transport experiments for bentonite clays containing different types of background electrolytes at various compactions. Specifically, emphasis is placed on quantifying the extent of co-ion exclusion and understanding the postulated surface diffusion mechanism on the basis of the well-established electric double-layer (EDL) theory. This contribution presents and discusses some preliminary results, based on the modified Boltzmann statistics, for the equilibrating part of the model.

A model has been developed that can simulate in some detail the chemistry of the glass dissolution process. To test the performance of the model, calculations were carried out to predict the dissolution behaviour of a commercial borosilicate glass. The model could reproduce accurately the behaviour of major elements released from the glass, although in the case of silica, ‘sorption’ at the glass surface was required in order to achieve good agreement with experimental data. Secondary reaction products sepiolite, montmorillonite, analcime and goethite were predicted to form. Further calculations were carried out to simulate the release of radioelements (caesium, strontium, uranium, plutonium) from within the dissolving glass. The precipitation of insoluble solids was the only mechanism included in the model to retain the radioelements at the glass surface. For plutonium, there was good agreement between model predictions and available experimental data. For caesium and strontium, the model underestimated the amount of retention at the glass surface and additional retention mechanisms, such as coprecipitation or ion exchange reactions involving newly-formed clay-type minerals at the glass surface must be invoked to explain experimental observation.

A geochemical model has been developed to describe the interactions of the main elements released from the COGEMA R7T7 SON68 glass with the Boom Clay environment and predict the formation of secondary phases which may influence the glass corrosion. It is based on two submodels: a model describing the interaction between the clay and its porewater, and a model for glass corrosion in groundwater. According to the calculations, no secondary minerals are predicted to precipitate when glass dissolves in the Boom Clay. The oncentrations of the main elements released from the glass are controlled by stable phases present in the Boom Clay (a.o. Si, Al) or by cation exchange reactions (e.g. Na). The concentration of silica in solution can be maintained below the glass silicon solubility limit. Therefore, the Boom Clay can suppress the saturation of silica, explaining the enhancing effect of Boom Clay on glass corrosion. The influence of the presence of clay on the other components of the glass is also discussed.

36Cl data from the Exploratory Studies Facility (ESF) at Yucca Mountain, Nevada, reported by LANL researchers [1,2,3,4] provide important empirical constraints on the velocity and pathways for water flow under natural conditions in the unsaturated zone at Yucca Mountain. It has been concluded that some data exhibiting exceptionally high 36Cl/Cl ratios (i.e., greater than 1250×10−15 [3]) indicate unambiguous bomb pulse contamination and fast groundwater travel times to depths of the proposed waste emplacement horizon. Several lines of evidence indicate that some ESF samples with 36Cl/Cl ratios greater than 900×10−15 to 1000×10−15 are also affected by bomb pulse contamination. The distribution of 36Cl/Cl data can be well represented by a mixture of two normally distributed populations, which are hypothesized to represent uncontaminated and bomb pulse contaminated sets. The transition between these groups occurs at 36Cl/Cl between 600×10−15and 1000× 10−15, and the hypothesized contaminated samvles comprise 20 to 25 percent of the total. Shapiro-Wilk statistics confirm that when samples with 36Cl/Cl ratios greater than about 1000×10−15 are included, the distribution deviates from normality. Variation in the Earth's magnetic field, which is a primary mechanism for variation in natural 36Cl production, also appears to be normally distributed, supporting the hypothesis that natural, relatively undecayed 36Cl/Cl ratios are normally distributed. Further evidence that samples with 36Cl/Cl greater than 950×10−15 to 1000×10−15 are likely to have a bomb pulse component is provided by the spatial correlation of these samples with those containing unambiguous contamination at values greater than 1250×10−15. Zones in which elevated 36Cl/Cl ratios are localized constitute about 23 percent of the length of the ESF at the level of the proposed waste emplacement horizon.

The evolution of porewater chemistry in bentonite was studied in solution-bentonite interaction experiments under anaerobic conditions at room temperature. The parameters varied were the bentonite density, bentonite-to-water ratio (b/w), ionic strength of the solution, and the composition of bentonite. At the end of the experiment, the equilibrating solution and the porewater squeezed out of the bentonite samples were analysed. This paper presents the preliminary experimental results of these ongoing studies. The evolution of porewater chemistry was determined by the dissolving components initially present in the bentonite together with the ions coming with water from the surroundings. Ion-exchange processes occured between the exchangeable cations of montmorillonite and the cations in the water. The obtained concentrations in the external solution and porewater strongly depended on the b/w used. The concentrations in the squeezed porewaters were clearly lower than in the external waters and decreased with increasing density during squeezing.

We evaluate the charge distribution on the surface of a model cluster which simulates a 2:1 type clay mineral and the binding forces between the model cluster surface and water molecules with semi-empirical molecular orbital method (MOPAC). It is found that the negative charge is not localized on the oxygen atoms of the model cluster surface. We calculate the binding forces acting on the water molecules in model cluster-water system and compare them with the binding forces of solid and liquid phases of water. We qualitatively discuss the binding force acting on the water molecule directly sorbed on the model cluster surface and the binding force acting on the water molecule which is not sorbed on the model cluster surface.

The initial stage of some HLW disposal systems will be characterised by a large thermal pulse in the near-field environment, due to the heat of the radioactivity decay. This will lead to the development of a transient spatial thermal gradient between the hot canister and the cold geological medium, which could significantly affect the composition and the elemental distribution within the near-field environment. A coupled experimental and modelling work is presented in order to determine the influence of a thermal gradient on water-rock interaction processes. First experiments with a simulated nuclear glass evidenced mass transfer processes leading to chemical differentiation in the solid phases between the hot and the cold end of the system. The relevance of these experimental results to the case of a HLW disposal is strongly supported by in-situ experiments at Stripa, in which a realistic EBS under thermal gradient developed exactly the same mass transfers.

In order to understand the driving force of these processes, we tried to model simplified experiments by using a mixing cell geochemical model built upon the geochemical code EQ3/EQ6. The discrepancies between modelling and experiments indicate the existence of coupled processes involving irreversible precipitation.

Finally, thermal gradients were applied in nuclear glass-clay interaction experiments to enhance elemental migrations. The main results are: (i) a re-crystallisation of the initial clay toward a more silicic one through incorporation of elements released by the glass, (ii) a strong influence of clay chemistry on the nuclear glass secondary phases.

The microstructure of clays controls their transport properties. This is concluded from comparing microstructural parameter data with the hydraulic conductivity and the ion diffusive transport capacity. Illitic clays contain a number of interacting open voids with a high flow capacity while natural smectite-rich clays are more homogeneous with smaller voids and a lower hydraulic conductivity than illitic clays with the same density. Artificially prepared smectitic clays, like those proposed for embedding canisters with highly radioactive waste, have a higher conductivity than natural clays with the same smectite content because the microstructural homogeneity of the artificial clays is less good.

The anion diffusive transport capacity of smectite-rich clays with high density is much lower than that of clays with low density in contrast to the cation diffusive capacity. This is explained by using quantitative microstructural data.

Since cementitious materials used in repositories will be in contact with water, one of the most important tasks is to assess their long-term performance while they are being degraded very slowly by leaching. The authors have proposed an electrochemical acceleration test method and clarified its applicability. In this method, a specimen is placed between two glass vessels containing water. An anode and a cathode connected to a DC power source provide a potential gradient across the specimen. Ca2+ ions in the pore solution move rapidly to the cathode side, and thus hasten cement hydrate dissolution.

In this study, we obtained two types of almost homogeneously degraded specimens by controlling cumulative quantity of dissolved Ca2+ ions. In the first type, there was no Ca(OH)2and the C-S-H phases underwent insignificant alteration, whereas in the second type, degradation of the C-S-H phases occurred.

As a preliminary evaluation of cementitious materials, diffusion coefficients of tritiated water were measured for nondegraded and degraded specimens. Diffusion coefficients were increased by degradation and closely corresponded to changes in porosity.

FEBEX (Full-scale Engineered Barrier EXperiment) is a demonstration and research project dealing with the bentonite engineered barrier designed for sealing and containment of waste in a high level radioactive waste repository. In the FEBEX experiment the bentonite barrier will be subject to heating and hydration during at least three years. One of the novel aspects of FEBEX is the use of artificial tracers, which are intended to provide additional insight into the hydrodynamic and geochemical behavior of the bentonite buffer. Here the results of the numerical prediction of tracers migration are presented. Numerical modeling of the experiments has allowed the identification of the most relevant parameters and will be useful for the design of post-morten analyses.

Designs for repositories for the underground disposal of radioactive waste throughout the world are generally based on the ‘multiple barrier’ concept. This comprises engineered barriers that are the introduced features of the disposal facility, and natural barriers provided by the host geology. In the near field of the repository, the engineered barriers can provide both physical and chemical containment. This paper is mainly concerned with the deep disposal of intermediatelevel and low-level waste. In particular, it reviews current understanding concerning the evolution of the chemical and physical barriers for the disposal concept developed by United Kingdom Nirex Limited. Examples of work from underlying research programmes that support this understanding are discussed. The paper also discusses the representation of the engineered barriers in performance assessment calculations.

The combined influence of gamma radiation (10 Gy/h) and high temperature (150 °C) on the corrosion of the promising HLW container material Ti99.8-Pd was investigated in an MgCl2-rich brine, and the corrosion surface films formed were characterized by XPS. For comparison, specimens without irradiation were also examined.

Under the test conditions used, the alloy Ti99.8-Pd is resistant to local corrosion and its general corrosion is negligible low. The thin corrosion films formed on the surface of unirradiated specimens and in the crevices of specimens exposed to radiation consist of TiO2. However, outside the crevices of irradiated specimens, a surface layer consisting of Mg (main component of the brine) and Si (impurity in the brine) oxide is built up over the TiO2 layer. Comparable TiO2 layer thicknesses (30 - 65 nm, depending on the experimental conditions) are found for unirradiated and irradiated laboratory specimens. The TiO2 layer formed on the in-situ corrosion specimens (33 nm / 5.3 years) is thinner than that of the laboratory specimens (58 nm / 191 days) indicating less aggressive conditions in the field experiments. In view of these results, the alloy Ti99.8-Pd continues to be considered as a strong candidate container material and will be further investigated.

The transport of six U(VI) species, HCO31− and CO32−, through the bentonite bed was modelled as diffusion in pore water combined with sorption/desorption on the surface of bentonite particles. The studied system consisted of the Czech commercial sodium bentonite SABENYL and synthetic granitic water spiked with 233U(VI), where equilibrium distribution of 233U was experimentally determined at first. The transport was simulated by means of eight dynamic, 1 D- partial differential equations while the equilibrium sorption/desorption was assumed. Three types of surface complexation models, namely the Constant Capacitance Model (CCM), Diffuse Double Layer Model (DLM) and the so-called Chemical Equilibrium Model (CEM), were used for the description of sorption/desorption interaction processes. The characteristic parameters of the individual models were obtained from evaluation of experimental data. The results of the simulation were expressed as path-length dependent concentrations of total U(VI), total carbonates and of individual migration species, as path-length dependent pH and uranium Kd, values at a given time and as time dependent concentrations of the total U(VI) at a given migration distance. It has been found that the transport of U(VI) through the bentonite bed is significantly influenced first of all by the value of pH and by total carbonates concentration. The important length dependent changes of the pH and Kd values were determined.