Hydro-mechanical Response to Gas Transfer of Deep Argillaceous Host Rocks for Radioactive Waste Disposal

During recent decades, argillaceous sedimentary formations have been studied as potential host formations for the geological disposal of long-living and heat-emitting radioactive waste—Boom Clay in Belgium and Opalinus Clay and Brown Dogger in Switzerland. A significant issue in the long-term performance of these potential host rocks concerns the generation and transport of gases. The pressure resulting from the generation of gas in an almost impermeable geological medium in the near field of a repository will increase. Under high gas pressures, the mechanical and hydraulic properties of the host rock are expected to change significantly. Preferential gas pathways may develop which exploit material heterogeneity, anisotropy (bedding planes), rock discontinuities, or interfaces between the different components of the repository, and may eventually lead to the release of the produced gases. Gas flow through these clayey rocks is investigated on the basis of laboratory work. Priority has been given to studying the volume change response of these initially water-saturated materials through relatively fast and controlled volume-rate gas injections. The effect of the gas injection rate, the confining pressure and the bedding orientation on the gas transport properties have been studied with particular attention paid to the coupling with strain behaviour. The results have shown features common to the three formations concerning the gas transfer process through preferential pathways, despite their initially differential properties.

Seismic anisotropy of Opalinus Clay: tomographic investigations using the infrastructure of an underground rock laboratory (URL)

Seismic anisotropy and attenuation make claystone formations difficult to characterize. On the other hand, in many geotechnical environments, precise knowledge of structure and elastic properties of clay formations is needed. In crystalline and rock salt underground structures, high-resolution seismic tomography and reflection imaging have proven a useful tool for structural and mechanical characterization at the scale of underground infrastructure (several deca- to hundreds of meters). This study investigates the applicability of seismic tomography for the characterization of claystone formations from an underground rock laboratory under challenging on-site conditions including anisotropy, strong attenuation and restricted acquisition geometry. The seismic tomographic survey was part of a pilot experiment in the Opalinus Clay of the Mont Terri Rock Laboratory, using 3-component geophones and rock anchors, which are installed 2 m within the rock on two levels, thus suppressing effects caused by the excavation damage zone. As a source, a pneumatic impact source was used. The survey covers two different facies types (shaly and carbonate-rich sandy), for which the elliptical anisotropy is calculated for direct ray paths by fitting an ellipse to the separated data for each facies. The tomographic inversion was done with a code providing a good grid control and enabling to take the seismic anisotropy into account. A-priori anisotropy can be attributed to the grid points, taking various facies types or other heterogeneities into account. Tomographic results, compared to computations using an isotropic velocity model, show that results are significantly enhanced by considering the anisotropy and demonstrate the ability of the approach to characterize heterogeneities of geological structures between the galleries of the rock laboratory.

Microscopic and spectroscopic investigations of uranium(VI) reduction by <i>Desulfosporosinus hippei</i> DSM 8344

Clay formations are potential host rocks for the long-term storage of high-level radioactive waste in a deep geological repository. Bentonites are supposed to serve as backfill material, not only for a final disposal site in clay formations but also in crystalline rock. For a long-term safety assessment, various aspects must be taken into account. Besides geological, geochemical and geophysical considerations, naturally occurring microorganisms also play a crucial part in the environment of such a repository. In the event of a worst-case scenario when water enters the disposal site, they can interact with the radionuclides and change for example the chemical speciation or the oxidation state (Lloyd et al., 2002). Desulfosporosinus spp. are an important representative of anaerobic, sulfate-reducing microorganisms, which are present in clay formations as well as in bentonites. Various studies have shown that they play a major role in the microbial communities of these surroundings (Bagnoud et al., 2016; Matschiavelli et al., 2019). A closely related microorganism to the isolated species is Desulfosporosinus hippei DSM 8344, which was originally found in permafrost soil (Vatsurina et al., 2008). This bacterium was used to investigate its interactions with uranium(VI) especially regarding the reduction to the less mobile uranium(IV). Time-dependent reduction experiments in artificial Opalinus Clay pore water (Wersin et al., 2011) (100 µM uranium(VI), pH 5.5) showed the removal of about 80 % of the uranium(VI) from the supernatants within 48 h. Corresponding UV/Vis measurements of the dissolved cell pellets exhibited an increasing proportion of uranium(IV) in the cell-bound uranium. Calculations with the inclusion of extinction coefficients led to a ratio of 39 % uranium(IV) after 1 week. Therefore, a combined sorption-reduction process is a possible interaction mechanism. Time-resolved laser-induced luminescence spectroscopy verified the presence of two uranium(VI) species in the supernatant. A comparison with reference spectra led to an assignment to a uranyl(VI) lactate and a uranyl(VI) carbonate complex. The species distribution showed a decrease of the proportion of the lactate species with time, whereas the proportion of the carbonate species remained almost constant. Uranium aggregates are formed on the cell surface during the process, as determined by transmission electron microscopy (TEM). Furthermore, uranium occurs inside and outside the cells as well as vesicles containing uranium. These findings help to close existing gaps in a comprehensive safeguard concept for a repository for high-level radioactive waste in clay rock. Moreover, this study provides new insights into the interactions of sulfate-reducing microorganisms with uranium(VI).

Impact of the heterogeneity of the sandy Opalinus clay facies at the Mont Terri underground research laboratory on radionuclide migration

The assessment of the safety of a deep geological repository (DGR) for high-level radioactive wastes over assessment time scales of up to 1 million years requires an in-depth understanding of the multi-scale coupled processes that affect the repository system evolution over time, to reduce uncertainties and conservatism in safety analyses. This is in particular required with respect to the challenges of a comparative assessment of different repository concepts in different host rocks within the process of a site selection for a DGR for heat-generating radioactive wastes in Germany. The collaborative project “Integrity of nuclear waste repository systems – Cross-scale system understanding and analysis (iCross)” conducted jointly by five research centres of the Helmholtz Association and co-funded by the Initiative and Networking Fund of the Helmholtz Association and the Federal Ministry of Education and Research (BMBF) has been initiated with the overall objective to improve the understanding of coupled thermal-hydraulic-mechanical-chemical-(micro)biological (THMCB) processes and to develop simulation tools that enable a holistic close to reality description of the long-term evolution of the repository system. Geological formations, such as those foreseen as potential host rocks for DGRs, and their surroundings are heterogeneous on various length scales ranging from nanometers to kilometers. Therefore, the aim of this work in the context of iCross is to evaluate the effects of mineralogical, geochemical and microstructural heterogeneities of repository host rocks on radionuclide transport in the repository far field, using the sandy facies of the Opalinus clay (SF-OPA) from the Mont Terri underground research laboratory (St. Ursanne, Switzerland) as an example. Here, we address in particular the migration behaviour of Ra-226 as an important radionuclide to be considered in safety cases for deep geological disposal of spent nuclear fuel. To assess the impact of the heterogeneities in SF-OPA on radionuclide transport, a complementary approach combining microstructural characterisation methods, experimental techniques for the determination of transport parameters of the rock matrix and the mobility of Ra-226 with innovative developments in reactive transport modelling on the pore and continuum scales was pursued. One of the results was that although the limited clay content in SF-OPA decreases the total amount of Ra bound to the illite phase, the solid solutions of sulphate and carbonate compensate for this and provide a major fixation mechanism.

Anaerobic corrosion of carbon steel in compacted bentonite exposed to natural Opalinus clay porewater: Bentonite alteration study

Carbon steel is a potential canister material for the disposal of high-level radioactive waste in deep geological repositories in clays and clay rocks. Bentonite is considered as a potential backfill material for those multi-barrier systems. To predict the long-term performance and for safety assessment the knowledge of canister corrosion behavior is important. The corrosion products formed and mineralogically altered bentonite at the canister/bentonite interface can potentially provide an additional barrier against radionuclide migration. In-situ corrosion experiments were performed at the Mont Terri underground research laboratory. Coupons of carbon steel were embedded in Volclay MX-80 bentonite with controlled densities, installed in a borehole under simulated repository and anaerobic conditions and exposed to natural Opalinus clay porewater for a period up to 5.5 years (Smart et al., 2017). In the present study, the bentonite layer at the canister/bentonite interface was characterized by complementary microscopic and spectroscopic techniques (XPS, SEM-EDX, µXANES) under anoxic conditions. The interface revealed reddish-brown staining up to 2 mm depth into the bentonite in the zone adjacent to the steel in all three obtained samples. The XPS analysis revealed formation of sulfides at the interface consisting of iron and other trace metals present in the steel. The SEM-EDX analyses of the interface (embedded cross-cut with steel removed) showed different degrees of calcium enrichment in the bentonite adjacent to the metal for various samples. The µXRF analysis performed on the bentonite at the interface showed a scarce or distinct calcium-enriched rim up to 100 µm into the bentonite and iron-enriched rim depending on the sample (one sample in Fig. 1), while µXANES analysis revealed formation of iron silicate compounds in the reacted reddish-brown zone. The iron appears to displace calcium from the interlayer sites in montmorillonite. The calcium then precipitates at the interface as calcite. The extent of this process seems to be strongly related to the bentonite density. The steel coupon was removed prior to embedding, with the location marked as resin in Fig. 1. A line scan from the edge towards the bulk bentonite did not indicate any systematic gradient in the Fe2+/3+ ratio. The formation of mixed Fe2+/3+ silicate compounds appears to be heterogeneous. This work contributes to an increasing understanding of steel corrosion mechanisms in clay, which can improve the robustness of canister lifetime predictions.

Claystone formations in Germany: what we (don't) know about them and how we can change this

Deep geological formations are considered for safe long-term disposal of high-level radioactive waste. Such a repository would be requested to prevent radionuclides from entering the biosphere for a period of 1 million years (StandAG, 2017). Consequently, a holistic characterization including lithological, mineralogical, geochemical, hydrological, structural and geomechanical properties of any potential repository-hosting rock formation is required. Nine claystone formations have been identified as “sub-areas” within the German site-selection procedure (BGE, 2020). The area covered by these formations comprises about half of the total area considered as being qualified for further exploration. However, despite its relevance to act as a geological barrier for, e.g. hydrocarbons or radionuclides, the characterization of clay-rich formations at depths exceeding 300 m in Germany has attained substantially less attention than economically more relevant units hosted by, e.g. sandstones or rock salt, which have been intensively explored. The BGR project BASTION aims at contributing to characterizing these claystone formations and emphasizes properties relevant to host a repository for nuclear waste. Investigations comprise (micro)structural/petrographic, mineralogical, geochemical, geophysical, hydraulic and thermomechanical analyses. In project phase I (2013–2019), claystones deposited in Northern Germany during the Lower Cretaceous were studied. These rocks belong to the fourth largest sub-area hosting claystones. Two of the main foci were to explore variations in lithology, mineralogy and geochemistry, and to identify deformation mechanisms (natural and artificial) by microstructural analyses. Although rocks appeared to be quite homogeneous on the 10–100 m scale, the results revealed distinct structural and sedimentary heterogeneities on the meter scale affecting fracture density. Another sub-area located in Southern Germany hosts the Opalinus Clay Formation (OPA). This up to 150 m thick claystone formation was deposited during the Middle Jurassic (Franz and Nitsch, 2009). Owing to its self-sealing capacity and ability to retain fluids, it is supposed to host the nuclear waste repository of Switzerland (Bossart et al., 2017). The OPA is quite well understood in terms of its lithology and (bio)stratigraphy, and there have been mineralogical, hydrological and petrophysical analyses, mostly documented in university theses a few decades old and sometimes difficult to access. However, it is questionable to what extent these investigations reflect the situation at depths relevant for the site-selection procedure. Well-documented data on the OPA and its properties relevant for nuclear waste disposal are available via the Swiss site-selection procedure (Bossart et al., 2017). However, as there remain substantial questions regarding the nature of the German portions of the OPA (e.g. spatial distribution of lithology, mineralogy, microstructures) at depths greater than a few decameters, it is unclear to what degree insights obtained in the Swiss site-selection procedure also account for Germany. Therefore, phase II of BASTION, which began in 2020, aims to use the multidisciplinary approach developed during phase I to characterize properties of the OPA relevant for the save long-term disposal of nuclear waste by identifying and quantifying structural and rheological heterogeneities. This will constitute important input for numerical models in any long-term safety assessment.

Transport in tight material enlightened by process tomography

The analysis of fluid transport through tight barrier materials poses two major challenges: (i) Long equilibration periods require long minimum experiment durations, and (ii) the fluid transport frequently results in complex pattern formation. Measuring times that are too short may feign transport rates that are too low; intact homogeneous samples are often missing problematic features, e.g. fractures. Both issues are detected and analyzed by using process tomography techniques, thereby providing an improved understanding of transport processes in complex materials. We thus continuously develop and apply the positron emission tomography (PET) method for geomaterials (Kulenkampff et al., 2016). This is able to trace very low concentrations of β+-emitting radionuclides during their passage through drill cores of barrier material with reasonable resolution (1 mm) and over variable periods (hours to years). The method yields time-resolved quantitative tomographic images of the tracer concentration (e.g. https://doi.org/10.5281/zenodo.166509), in contrast to input-output experiments like common permeability measurements, diffusion cells or break-through curves. Our current research includes the analysis of diffusive transport in heterogeneous shales (sandy facies of the Opalinus Clay) (BMBF and HGF iCross project), the reactive flow in fracture-filling materials of crystalline rocks (Eurad FUTURE project) and transport in engineered barriers and the contact zone (Euratom Cebama, Eurad Magic, as well as MgO and Stroefun BMWi projects). The efforts combine flow field tomography, structural imaging and reactive transport modelling to improve process understanding and to provide a bridge from the molecular to the macroscopic scale. The benefits include: Insight into temporal stability and spatial heterogeneity of the observed transport process Parameterization of local velocity distribution and effective volume as well as comparability with pore-scale model simulations Ability to quantify multiple internal transport rates without the need to register the delayed output signal Transparent and palpable visualization of processes hidden in the opaque material The method requires specific constraints of the experimental setup (size, fluid pressure, temperature). Nevertheless, it provides unique insight into reactive transport processes observed in potential materials for nuclear waste management.

Diffusive transport of uranium and americium through clay rock down to ultra-trace levels

Clay rock is a potential host rock for the final disposal of nuclear waste in deep geological formations. In the scenario of ground water intrusion into the nuclear waste repository and subsequent corrosion of canisters and waste, possibly released actinides, such as uranium (U) and americium (Am), would be transported through the engineered barrier and clay host rock mainly by diffusion. Actinides are known to exhibit low solubility and are strongly sorbing under the reducing conditions of deep geological formations. Diffusion experiments are therefore difficult to perform due to analytical constraints. To our knowledge, the diffusion of U in clay rock has not been investigated below concentrations of 10-7molkg-1 clay (e.g., Joseph et al., 2013) and for Am, no diffusion experiments have been performed in a clay rock, considered suitable as host rock, such as Opalinus clay (OPA). This work aimed at the investigation of the diffusive behavior of U and Am down to ultra-trace concentrations (≪10-7molkg-1) in OPA. Laboratory-scale diffusion experiments were conducted with samples of OPA, obtained from the Mont Terri underground laboratory, Switzerland for up to 240 d. The OPA samples were placed in contact with synthetic pore water (I=0.22molL-1, pH=7.24), spiked with 5×10-9molL-1 of 233U and 243Am, respectively. After termination of the experiment, the OPA samples were segmented into thin layers of 20–400 µm. The obtained clay segments were analyzed for the 233U and 243Am content with accelerator mass spectrometry. Concentration profiles were determined down to 10-12molkg-1 clay for 233U and 5×10-10molkg-1 clay for 243Am. U showed a typical diffusion profile which can be interpreted by applying Fick's laws, while the Am profile was more complex, requiring further investigation of transport modes. These results will provide relevant insights into the transport behavior of U and Am in OPA over a wide range of concentrations down to ultra-trace levels.

Building the bridge between safety requirements and numerical modeling: an example considering crack development of Opalinus clay in laboratory and field scales

Salt, crystalline and clay formations are under discussion as potential host rocks for storage of heat-generating radioactive waste. Each of these rocks has a different structure and composition, and consequently a different material behavior. The latter needs to be studied and evaluated with respect to the main aim: to find a place to store the waste in a safe and sustainable manner. Several requirements in the context of the safety of a repository need to be fulfilled, concerning the long-term as well as the operational phase. One key point in this matter is the integrity, which refers to retention of the isolating rock zone's containment capabilities. With the focus on some experimental and numerical investigations on the excavation influenced near-field behavior of Opalinus clay (OPA), this contribution aims to illustrate an example for the role of numerical modeling in safety assessment. Once, e.g. anthropogenic action such as excavation starts, the natural state of equilibrium in the formation is disturbed. Trying to restore it, the rock deforms (convergence) and/or releases energy in other ways such as cracking. This could lead to loss of integrity since crack nucleation and propagation can affect the mechanical stability and create paths to transport contaminants. During operation in the excavated rock, environmental changes, e.g. temperature and humidity, further affect its behavior. The understanding of these dynamic phenomena ideally needs to occur at the in situ scale; however, performing an experiment in the spatial and time scales of interest is not always possible. For this reason, the in situ problem needs to be formulated, abstracted and mathematically modeled. The interpretation of the results must take place with simplifying assumptions and complementary laboratory scale experiments can be used to improve understanding of the system. The real problem is approached stepwise, each step associated to the size of the model and its complexity. The gradually obtained knowledge is necessary to achieve a better understanding of the process and to evaluate the capacities and limitations of the models. This contribution aims at showing the basic practical steps for numerical modeling with particular focus on the preparation and interpretation of the models and results, e.g. model calibration, verification and validation. As an example, the OPA at the Mont Terri site is chosen. The material parameters are obtained either experimentally or from the literature. We choose and perform laboratory scale simulations that are related to nearly the same mechanism as in the in situ scale. To have a first impression on the latter, a simplified, large-scale numerical model is prepared. The mechanism in study is drying and wetting, which is associated with shrinkage and swelling. We analyze the pore pressure and stress development in both scales. Thus, hydraulic mechanically coupled approaches are essential. The concept of effective stress is used, which combines the contributions of the solid and fluid phases (gas and liquid). In the current modeling approach, the gas pressure remains constant (atmospheric pressure) and during drying, the liquid pressure induces capillary pressure development and decrease of saturation. The laboratory scale simulation is important to evaluate the model of choice and to assess potential numerical problems. Furthermore, it can be used to perform a sensitivity study of material and numerical parameters. This step is necessary during the development or extension of numerical models as well as to evaluate their applicability on new research questions. The simplified in situ scale numerical model is then extended. In this phase the numerical model is evaluated once again, especially with respect to its complexity. Furthermore, specific questions related to this scale are posed: overall behavior of the rock, influence of the excavation, seasonal and long-term effects. In this contribution we deal with the long-term cyclic deformation (CD-A) experiment. The CD-A experiment has been taking place in the Mont Terri Rock Laboratory since October 2019. It consists of twin niches, a closed and an open niche, subjected to either high air humidity or seasonal humidity changes leading to saturation/desaturation during summer/winter in the OPA, respectively. Several parameters are periodically or continuously measured, including relative air humidity, convergence and crack development. We attempt to transfer the knowledge and numerical models developed in the small scale to the large scale and to evaluate the possibilities and limitations of the chosen approaches by comparing the numerical and experimental results.