From process to system understanding with multi-disciplinary investigation methods: set-up and first results of the CD-A experiment (Mont Terri rock laboratory)

A potential repository site for high-level radioactive waste should ensure the highest possible safety level over a period of one million years. In addition to design issues, demonstrating the integrity of the barrier is essential as it ensures the long-term containment of radioactive waste. Therefore, a multi-disciplinary approach is necessary for the characterization of the surrounding rock and for the understanding of the occurring physical processes. For site selection, however, the understanding of the respective system is essential as well: Do fault zones exist in the relevant area? Are these active and relevant for interpreting system behavior? What is the role of the existing heterogeneities of the claystone and how do these site-dependent conditions influence the physical effects? To answer these questions, the site-selection procedure requires underground exploration, which includes geophysical and geological investigations on milli- to decameter scales. Their results serve as the basis for numerical modelling. This combined, multi-disciplinary interpretation requires extensive knowledge of the various methods, their capabilities, limitations, and areas of application. In the cyclic deformation (CD-A) experiment in the Mont Terri rock laboratory, the hydraulic–mechanical effects due to excavation and the climatic conditions within the rock laboratory are investigated in two niches in the Opalinus Clay. The twin niches differ mainly with regard to the relative humidity inside them, but are also characterized by different boundary conditions such as existing fault zones, the technical construction of the neighboring gallery, etc. In order to gain insights into the relevance of the individual influences, comparative studies are being carried out on both niches. The presented results provide a first insight into the initial experimental years of the CD-A long-term experiment and illustrate the benefits of multi-disciplinary investigations in terms of system understanding and the scale dependency of physical effects. Amongst other effects, the assessment of the impact of heterogeneities on the deformation behavior and the evolution of pore water pressure is very complex and benefits from geological interpretation and measurements of for example deformation, water content, and pore pressure. The numerical modeling allows statements about the interaction of different processes and thus enables an interpretation of the overall system, taking into account the knowledge gained by the multi-disciplinary investigation.

Large-scale testing of a sandwich shaft-sealing system at the Mont Terri rock laboratory

Shaft-sealing systems for nuclear waste repositories are constructed to limit fluid inflow from the adjacent rock during the early stage after closure of the repository and to delay the release of possibly contaminated fluids from the repository at later stages. Current German concepts of shaft seals contain the hydraulic sandwich sealing system as a component of the lower seal in host rock (Kudla and Herold, 2021). The KIT-developed sandwich sealing system consists of alternating sealing segments (DS) of bentonite and equipotential segments (ES) that are characterized by a high hydraulic conductivity. Within the ES, fluid is evenly distributed over the cross section of the seal. Water bypassing the seal via the excavation-damaged zone or penetrating the seal inhomogeneously is contained, and a more homogeneous hydration and swelling of the DS is obtained. The functionality of such a system was proven in laboratory and semi-technical-scale experiments (Schuhmann et al., 2009). After a joint international pre-project (Emmerich et al., 2019) dedicated to the planning of a large-scale in situ test that demonstrates the feasibility and effectiveness of the sandwich shaft-sealing system in interaction with the host rock, the large-scale experiment was launched at the Mont Terri rock laboratory in July 2019 with partners from Germany, Switzerland, Spain, UK, and Canada. It consists of two experimental shafts of 1.18 m diameter and 10–12.6 m depth, constructed using a core drilling technique with a custom-made drill rig in a new niche in the sandy facies of the Opalinus Clay. The seal in shaft 1 consists of four DS (calcigel) of 1 m thickness and five ES (fine-grained quartz sand), each 30 cm thick (Fig. 1). Shaft sinking began in August 2020 and was completed in November 2020. In the following months, the sealing system and instrumentation of shaft 1 were installed. The sealing system is saturated from a pressure chamber located at the shaft bottom via an inclined lateral feeding borehole. Hydration of the system started in May 2021. Shaft 2 will host a slightly modified system emplaced 1–1.5 years later, in order to integrate experience obtained during the early operation phase of shaft 1. In contrast to shaft 1, the excavation-damaged zone around shaft 2 will have had time to develop. The seals and the surrounding rock are intensely monitored. Measurements in the rock (geophysics, pore pressure, and total stress) were started between August 2019 and March 2020. Characterization of the excavation-damaged zone along the wall of shaft 1 was performed by geophysical and surface packer measurements prior to seal emplacement. Measurements inside the shaft comprise water content, relative humidity, and temperature, pore pressure, stress, and displacements. The in situ work is backed by laboratory testing and model simulation. Data and experience obtained to date will be presented. The sandwich experiment is funded by the German Federal Ministry for Economic Affairs and Energy under contract 02E11799.

The Role of Geological Models and Uncertainties in Safety Assessments

Safety assessments in nuclear waste management typically include the analysis of thermo-mechanical (TM) coupled processes. The TM behavior of the host rock is, amongst other aspects, dependent on the prevalent geological geometry. This study aims to evaluate the impact of uncertainties in geometry on the TM rock behavior. It is one of the very first studies aiming to bring uncertainties of structural geological models and numerical simulations together.To analyze the influence of geological geometries, a simplified model of the region around the Mont Terri rock laboratory was created. A 3-D structural geological model was set up and uncertainties of the lithological contacts were quantified by means of stochastic simulations, resulting in an ensemble of 89 model realizations. These realizations were transformed to a 2-D numerical model. In this numerical model, TM simulations were computed over a simulation time of 500 years, employing the Finite Element Method. To simulate a heat source of nuclear waste the lower edge of the model was set to 100°C.The results of these simulations show mean temperature variations of 90.89 °C and 92.70 ° after 500 years, with a maximum stress varying between 0.02 MPa and 0.16 MPa of elastic shear energy density and according mean cumulative displacements ranging from 30 cm to 38 cm. The presented results indicate that different model geometries and differences in material properties lead to noticeable variabilities of the TM behavior of claystone. However in this case, these variabilities would not significantly affect the integrity of the rock.

Investigations based on biaxial tiltmeter array and uniaxial hydrostatic levelling system at the Mont Terri rock laboratory

<p>In order to enable investigations and further comprehensive understanding of dynamical processes, it is clear one has to identify all relevant parameters and aim to record them all under best conditions concerning e.g. resolution, coverage in space, and in many cases on a multitude of scales in time. Obviously, it is also difficult to satisfy all these constrains in full. Especially scientific long-term observations often suffer the lack of necessary lasting commitment; secure funding, continual high quality maintenance, protected environment, or sufficient planning stability. Fortunately, the Swiss Mont Terri rock laboratory, with its history of now 25 years of forefront scientific expertise, a long-standing fruitful cooperation formed by the partners of the consortium and in consequence thereof state-of-the-art results obtained through 100 completed individual experiments and 45 additional experiments actually ongoing, ensures the conditions listed above.</p><p>Based on this favorable prospect, a now growing tiltmeter array is established at the underground laboratory. The instruments are embedded in several multidisciplinary experiments, dedicated to numerous, different scientific questions. Starting in April 2019, the first two platform tiltmeters became operational. Less than two years later, ten biaxial instruments are quasi-continuously monitoring deformation at various locations within the galleries and niches at Mont Terri. The envisioned, increasing spatial coverage of the network will facilitate geodetic observations of the underground rock laboratory as a whole and of its subregions as well.</p><p>Already in September 2012, a 50 m long hydrostatic levelling system (HLS) was installed along a gallery in the underground laboratory to detect displacements across an active geological fault zone. The combination of both, i.e. the uniaxial, integral deformations data provided by HLS together with the array of biaxial, point measurements acquired by the tiltmeters offers a unique concerted opportunity to achieve detailed deformation data in a large underground rock laboratory and to survey the associated dynamical processes occurring on timeframes covering seconds to decades.</p>

In situ measurements at the Mont Terri rock laboratory to study argillaceous rocks

<p>Safe as well as sensible economic uses of the subsurface demand both the comprehensive knowledge of the present state of a system and the understanding of the relevant dynamical processes. In order to facilitate these requirements, adequate characterisation, sufficient monitoring, and conclusive experiments have to be performed. Following this directive, the German Federal Institute for Geosciences and Natural Resources (BGR) has developed, adapted, and successfully employed methods to prospect Opalinus Clay in the Swiss Mont Terri rock laboratory. These methods encompass geoscientific in situ characterisations as well as investigation techniques as part of long-term monitoring programmes from the complementing fields of e.g. micro-seismics, Electrical Resistivity Tomography, micro-structural petrography, geohydrology, and Nuclear Magnetic Resonance. With this expertise, BGR has contributed numerous experiments, which are embedded and coordinated in the long-standing and fruitful cooperation with the partners of the Mont Terri Consortium.</p><p>The knowledge gain, based on now almost 25 years of BGR's engagement in the Mont Terri Project, offers comparison and evaluation of different, complementing methods determining present values and their evolution in time of e.g. moisture, saturation, pressure, deformation, the characterisation of parameter variability, and localisation of heterogeneities. It provides information allowing for programme optimisation of in situ measuring methods concerning penetration, resolution, effort, time, or feasibility. Therefore, the research results can be used for decision-making to refine investigation endevours in regards to specific demands of a certain site or a particular scientific problem not only for Opalinus Clay but also other claystone formations, and in some cases even for non-argillaceous rocks.</p>

Coupled HM modeling assists in designing CO2 long-term periodic injection experiment (CO2LPIE) in Mont Terri rock laboratory

<p>We are performing a series of coupled hydro-mechanical (HM) simulations to model CO<sub>2</sub> flow through Opalinus Clay at the Mont Terri rock laboratory in the CO<sub>2</sub> Long-term Periodic Injection Experiment (CO<sub>2</sub>LPIE). CO<sub>2</sub>LPIE aims at inter-disciplinary investigations of the caprock sealing capacity in geologic CO<sub>2</sub> storage in a highly monitored environment at the underground laboratory scale. Numerical modeling allows us to gain knowledge on the dynamic processes resulting from CO<sub>2</sub> periodic injection and to assist the experimental design. The cyclic injection parameters (i.e., the period and the amplitude) have to be optimized for the field experiment and therefore different values are taken into account. Opalinus Clay is a claystone with nanoDarcy permeability that contains well developed bedding planes responsible for its anisotropic HM behavior. The hydraulic anisotropy is defined by a permeability parallel to the bedding planes being three times the one perpendicular to it. Additionally, the drained Young’s modulus is measured to be 1.7 GPa parallel and 2.1 GPa perpendicular to bedding. Excavation reports by swisstopo document a SSE-dip of 45° for the bedding planes at the experiment location. CO<sub>2</sub> injection generates a mean overpressure of 1 MPa into the brine that propagates into the formation. The differential pressure between CO<sub>2</sub> and formation water, i.e., capillary pressure, is lower than the entry pressure and thus, CO<sub>2</sub> diffuses through the pores but does not advect in free phase. The liquid overpressure distribution is distorted by the hydraulic anisotropy, preferentially advancing along the bedding planes, as the associated permeability is higher than the one perpendicular to the bedding. The pore pressure buildup induces a poromechanical stress increase and an expansion of the rock that leads to a permeability enhancement of up to two orders of magnitude. The cyclic stimulation propagates trough the domain faster and with a lag time and an attenuation, both of which increase with distance from the source with, their values being dependent on permeability, porosity and stiffness of the rock. As a result of the model orthotropy, the attenuation and the lag time change with direction, i.e. they are higher in the direction perpendicular to the bedding and lower in the direction parallel to the bedding. Given the very low permeability of Opalinus Clay, the overpressure generated requires a long time to diffuse into the rock. Furthermore, the amplitude attenuation dissipates quite rapidly, so monitoring wells should be placed as close to the injection well as possible. The study of amplitude attenuation and time lag is necessary to determine how they can be utilized to evaluate the evolution of the HM properties as the rock is altered by the acidic nature of CO<sub>2</sub>-brine mixture Comparison between field data and numerical simulations will be a useful asset to fill the gap.</p>

3D Predictive HM-modeling in the heterogeneous Opalinus Clay of the Mont Terri rock laboratory and validation with monitoring data from a mine-by test

<p>The Mont Terri rock laboratory began in 1996 with 8 niches, followed by a research tunnel in 1998. Since then the laboratory has been expanded every 10 years, mainly in the shaly facies of the Opalinus Clay. In March 2018, south of the existing laboratory, the Mont Terri Project Partners initiated another extension «Gallery 18» of the Mont Terri rock laboratory mainly located in sandy facies of the Opalinus Clay. In October 2019 the extension was finished, resulting in more than 500 m of additional galleries and niches for new experiments. In the frame of this extension, for the first time a heterogeneous mine-by test, comprising a sheet of sandy facies and carbonate-rich sandy facies sandwiched between shaly facies was conducted in the rock laboratory. This so-called MB-A experiment (hydro-mechanical characterization of the sandy facies before and during excavation) consists of two lateral niches for instrumentation and monitoring and a test gallery of 30 m length oriented perpendicular to the latter. The instrumentation based on 26 boreholes with lengths up to 40 m consists of pore pressure transducers, extensometers, inclinometers and stress monitoring stations. It was finished several months before excavation of the test section was started in order to assure equilibration close to the initial conditions. Excavation of the test gallery running parallel to bedding strike was carried out in May 2019 in 20 days.</p><p>Elastic predictive modeling is performed in 3D to estimate the hydro-mechanical behavior of the rock mass during a sequential excavation according to effective daily advances and as-is sensor locations. The modeling results are compared with monitoring data. The calculation predicts a rotation of the early time near-field pore pressure reduction from perpendicular to parallel to bedding for late times. In general, monitored peak pore water pressures were higher than predicted, with a remarkable phase shift depending on distance and spatial position with respect to the drift. Monitored deformations were clearly underestimated with the elastic calculation. The overall behavior of the excavation in the sandy facies was unexpectedly not so different from former excavations in shaly facies.</p><p>A parametric study was performed to assess key parameters of potential effects of excavation on the hydromechanical responses of the excavation. It is concluded that adapted constitutive laws are needed in order to properly predict the hydromechanical response in stiffer claystone, such as for instance the sandy and carbonate-rich sandy facies of the Opalinus Clay.</p>

Alteration of Bentonite Reacted with Cementitious Materials for 5 and 10 Years in the Mont Terri Rock Laboratory (CI Experiment)

The cement–clay interaction (CI) experiment was carried out at the Mont Terri rock laboratory to complement the current knowledge on the influence that cementitious materials have on Opalinus Clay (OPA) and bentonite (MX). Drill cores including the interface of OPA, concrete (LAC = low-alkali binder, and OPC = ordinary Portland cement), and MX, which interacted for 4.9 and 10 years, were successfully retrieved after drilling, and detailed analyses were performed to evaluate potential mineralogical changes. The saturated compacted bentonites in core samples were divided into ten slices, profiling bentonite in the direction towards the interface, to evaluate the extent and spatial variation of the mineralogical alteration of bentonite. Regarding the mineral compositions of bentonite, cristobalite was dissolved within a range of 10 mm from the interface in both LAC-MX and OPC-MX, while calcite precipitated near the interface for OPC-MX. In LAC-MX and OPC-MX, secondary products containing Mg (e.g., M-S-H) also precipitated within 20 mm of the interface. These alterations of bentonite developed during the first 4.9 years, with very limited progress observed for the subsequent 5 years. Detectable changes in the mineralogical nature of montmorillonite (i.e., the formation of illite or beidellite, increase in layer charge) did not occur during the 10 years of interaction.

Active seismic monitoring of CO2-saturated brine injection into a fault (CS-D experiment in the Mont Terri Rock Laboratory)

<p>Confirming the permanent containment is a key challenge for the storage of CO<sub>2 </sub>in deep underground reservoirs. Faults in the cap rock of such reservoirs are potential flow paths for the CO<sub>2</sub> to escape. Our decametre-scale experiment at the Mont Terri Rock Laboratory aims to better understand mechanisms of CO<sub>2</sub> leakage trough a fault, and to test strategies to monitor the propagation of CO<sub>2</sub>-saturated water through faults.</p><p>Two boreholes were drilled through the main fault in Mont Terri with packer-intervals dedicated to fluid-injection and hydraulic/geochemical monitoring. Another five boreholes in the close surrounding were equipped with various instruments for geotechnical and geophysical observations. During the first phase of the experiment, the hydraulic response of the fault was characterized with injections of formation water in a step-up mode at pressures up to 6.0 MPa. The second phase, which was still on-going at the time of the abstract submission, consists of a long-term (several months) injection of CO<sub>2</sub>-saturated formation water at a constant head of 4.5 MPa, which is below the fault opening pressure. All injection activities were monitored with active seismic measurements, along with a comprehensive set of hydraulic-, mechanical-, geochemical- and other geophysical surveys. We will present the active seismic imaging results from the step-up injection test and compare them with the other surveys. Additionally, preliminary results will be shown acquired during the long-term injection of CO<sub>2</sub>-saturated formation water into the fault.</p>

Tiltmeter Measurements at the Underground Rock Laboratory in Mont Terri

<p>Mont Terri rock laboratory, located in the Swiss Jurassic Mountains, was established with the focus on the investigation and analysys of the properties of argillaceous formations. The scope of Opalinus Clay as a safe, potential option for nuclear waste disposal was broaden, as the behaviour of claystone is of high interest also in the context of caprocks, and hence, for many dynamical processes in the subsurfaces. Extensive research has been performed already for more than 20 years by the partners of the Mont Terri Consortium. These close cooperations cover a broad range of scientific aspects using numerical modelling, laboratory studies, and last not least in-situ experiments. Here, included in the long-term monitoring programme, new investigations apply tiltmeters. Since April 2019, platform tiltmeters have been installed at various locations within the galleries and niches of Mont Terri. The biaxial instruments have resolutions of 1 nrad and 0.1 µrad, respectively (Applied Geomechanics and Lippmann Geophysikalische Messgeräte). The tilt measurements are embedded within various experiments contributing to specific, multiparametrical studies. However, the growing tilt network as a whole will also provide novel information of the rock laboratory. The different time-scales of interest include long-term observations of yearly and decadal variability. So far tilt signals were identified due to excavations during the recent enlargement of the laboratory, earthquake activity (Albania), and local effects. First results of these quasi-continuous recordings will be presented.</p>