The French underground research laboratory in Bure as a precursor for deep geological repositories
Published:January 01, 2008
- PDF LinkChapter PDF
Jacques Delay, André Lesavre, Yannick Wileveau, 2008. "The French underground research laboratory in Bure as a precursor for deep geological repositories", Deep Geologic Repositories, Norbert T. Rempe
Download citation file:
Since 1999, the French National Radioactive Waste Management Agency (Agence nationale pour la gestion des déchets radioactifs [ANDRA]) has been carrying out investigations at its Meuse/Haute-Marne laboratory site to study the possibility of implementing an underground repository for high-level and long-lived waste. The geological formation under consideration consists of a stiff clay rock located at a depth of 500 m.
The purpose of this paper is to describe the role of the underground research laboratory and the associated experiments under way within the overall approach of designing a waste repository. The legislative framework indicates that laboratory investigations must focus on the properties of the host rock in its natural state and on its behavior under chemical or mechanical disturbances. Hence, the underground research laboratory is neither a pilot repository project, since the work site of the future repository will be located somewhere else, nor a methodological laboratory, since the experimental and geological-survey methods being used are considered reliable. The purpose of the work conducted by ANDRA is to provide the basic information required to design a safe repository. The research is meant to determine the capability of the clay formation to delay the migration of radionuclides and their release in the biosphere; to propose a repository architecture and building methods for underground structures (excavation and support modes); and to study the physical and chemical disturbances induced by the construction of underground structures and to propose solutions to reduce disturbances and to restore the closest conditions possible to the undisturbed state after shutdown. The Meuse/Haute-Marne underground research laboratory is also used to verify in situ the relevance of the concepts developed to ensure the containment of radionuclides over long time scales and, consequently, to assess the performance of the construction and support techniques that may be applied at an industrial scale in order to implement a repository and the specific techniques capable of reducing the mechanical and chemical disturbances induced by excavation activities.
This paper presents the work and investigations carried out in the underground research laboratory to determine which construction and instrumentation arrangements could be used to implement and operate a waste repository.
France defined its approach to research for radioactive waste disposal by an Act of Parliament in December 1991 that prescribed three alternative strategies: long-term surface storage, radionuclide transmutation, and deep geological disposal. This act also created the National Radioactive Waste Management Agency (Agence nationale pour la gestion des déchets radioactifs [ANDRA]) to manage the program and to conduct investigations for a deep geological repository for high-level and long-lived waste. In agreement with the act, ANDRA reported its findings on potential waste-management solutions to the government at the end of 2005 (ANDRA, 2005a).
In November 1999, ANDRA began building an underground research laboratory in eastern France, on a site located near the village of Bure, ∼300 km east of Paris. The site (Fig. 1) straddles the border between the Meuse and Haute-Marne Departments, a border that also separates the regions known as Champagne-Ardenne and Lorraine.
The research activities of the underground research laboratory are dedicated to reversible, deep geological disposal of high-level and long-lived waste in an argillaceous host rock. As in the case of underground research facilities for radioactive waste studies in other countries, the Meuse/Haute-Marne underground research laboratory supports several major goals, including: scientific characterization of the geological environment, establishment of a disposal concept based on the properties of the host rock, and understanding of excavation and operational effects, as well as the development of a scientific basis for the quantitative assessment of geological containment.
First, we present the geological context of the underground research laboratory and the scientific objectives of the survey program, followed by a description of some of the design elements selected by ANDRA for the repository, as shown in the Dossier 2005 (ANDRA, 2005a) Lastly, the work carried out in the underground research laboratory is covered through a technological overview of the excavation and lining methods used for laboratory shafts and drifts and through a scientific overview of the thermal characterization of the rock with a view to determining the spacing between disposal cells for exothermal waste. Both overviews help to establish the specifications of a reversible repository (ANDRA, 2005b, 2005c).
Underground Research Laboratory Site Overview
The target horizon for the underground research laboratory was a 130-m-thick layer of argillaceous rocks between ∼420 and 550 m below the surface. Stratigraphically, the depositional period straddles the Callovian and Oxfordian subdivisions of the Middle to Upper Jurassic. Argillaceous rocks contain a mix of clay minerals and clay-sized fractions of other compositions. The clays, constituting 40%–45% of the Callovian-Oxfordian argillaceous rocks, isolate groundwaters. Silica- and carbonate-rich sedimentary components reinforce the rock and promote stability for the underground construction.
The underground research laboratory location lies in the eastern portion of the Paris Basin, which covers a major portion of northern France (Fig. 2). The beds are nearly flat-lying and have a slight dip of less than 1.5° westward toward the center of the basin. The deep-water depositional environment of the Callovian-Oxfordian argillaceous rocks created a homogeneous layer that is continuous over most of the Paris Basin.
The stratigraphy of the underground research laboratory region consists of Jurassic limestones, marls, and argillaceous rocks (Fig. 3). The stratigraphy of the site is made up of alternating limestone-rich and clay-rich units. The major overlying limestone units are the Tithonian Barrois limestone, which forms a surface veneer over the site, and the Oxfordian limestones from ∼150 to 400 m deep. Between the Barrois and Oxfordian limestones, there is a 150-m-thick sequence of mixed Kimmeridgian argillaceous rocks, marls, and limestones. Underlying the Callovian-Oxfordian argillaceous rocks, the Bathonian and Bajocian Dogger limestones and dolomitic limestones can be found (Vigneron et al., 2004).
Approach to Repository Design Within A Geological Formation
The purpose of long-term management for high-level and long-lived waste is to protect human beings and the environment against the risk associated with that waste over a period of several hundreds of thousands of years. The solution investigated by ANDRA consists of isolating the waste within a deep geological formation to prevent the dissemination of the radionuclides it contains. Passive isolation is involved in the sense that it requires no maintenance or monitoring for periods that may extend over several hundreds of thousands of years.
To fulfill the requirements for both operational and long-term safety of a disposal facility, ANDRA developed two different approaches in line with the phases of the repository's lifetime (ANDRA, 2005c): the first one involves operational safety and is similar to a conventional approach—it focuses on repository-specific issues and helps to assess building measures within the framework of a reversible operation; the second approach relates to the safety assessment during the postclosure monitoring phase in order to estimate the robustness of the repository with regard to the long-term isolation of radioactive waste from the biosphere.
To successfully carry out both approaches, ANDRA conducted a functional analysis that attributed safety functions to each significant repository component (geological formation, waste packages, potential engineered barrier) and listed those functions under the various specification requirements for the components to be built (thickness of the containment envelopes of disposal packages, size of disposal cells, drift plugs, etc.). Those functions and the behavior of those components were assessed through repository-evolution scenarios.
The design of the repository is regulated by the safety approach that will lead to the size and specifications of containment barriers with a view to: (1) preventing water circulation (impermeability of the geological environment and sealing of repository structures); (2) immobilizing radionuclides at the package level by creating or maintaining favorable physicochemical conditions for that retention; and (3) retarding and mitigating any potential migration of radionuclides outside the disposal cells.
Hence, in the framework of the Meuse/Haute-Marne project, investigations have led to the proposal of a repository with the following features (ANDRA, 2005b):
It is located at the center of the layer to maximize the thickness of the impermeable geological formation and to ensure the best containment possible.
The disposal areas for each waste category are compartmentalized to reduce intrusion risks or failure consequences. The different waste categories are emplaced in separate disposal areas to simplify their safety assessments and to ensure the thermal independence of the different areas. According to repository-design studies, the temperature of the argillite is limited to 90 °C to prevent the alteration of its retention properties.
The structures have a simple geometry with circular profiles that are considered most stable and are lined in such a way as to remain stable for 100 yr.
The materials coming in contact with the rock, whether they are natural (clay rock) or manmade (concrete, steel, plugging or backfill materials), help to maintain the physicochemical conditions that retard package alteration and degradation to limit the premature release of radionu-clides into the biosphere.
Figure 4 shows a block diagram of the repository in such a formation.
French authorities requested that ANDRA study the option of reversibility for the repository. That requirement was taken into account in the design. Reversibility implies a human presence, a monitoring system, as well as maintenance activities throughout the entire reversibility period involved (OECD/NEA, 2001). None of these measures may compromise the long-term safety of the repository.
In ANDRA's project, reversibility must be understood as the possibility to manage the repository with flexibility by identifying the main stages in the repository's lifetime, starting from its construction until its final shutdown, and by proposing at every step the freedom of decision for future generations to decide whether they should continue with the ongoing process. Such flexibility in management also allows the repository design to evolve and minimizes delays during construction.
Repository reversibility stipulates that waste-package retrieval will remain possible without compromising operational and post-closure safety. Reversibility will be made much easier if durable structures and materials are selected and if construction and handling concepts, as well as operational conditions, are kept simple.
Closely associated with reversible management of the repository, an observation and monitoring program will be implemented to verify that operation of the repository complies with the forecasts (IAEA, 2001). Analysis of various repository parameters provides useful safety-related information (OECD/NEA, 2001). If needed, monitoring arrangements provide an opportunity, through an adaptive-staging approach, to draw experience feedback on the reliability of the forecasts concerning the behavior of the structures and to propose other operational modes (DOE, 2002).
Properties of the Host Rock Studied Through Underground Research Laboratory Experiments
The geological environment must be able to act as a radionu-clide barrier over the long term. It is therefore essential to know the physicochemical characteristics of the investigated rock in order to verify that all safety-related functions are adequately fulfilled. That knowledge also helps to determine building and operational conditions for the repository.
Since water circulation is likely to be the main factor that will not only be able to alter the packages, but also to dissolve and transport the radionuclides, it must be kept low. The Callovian-Oxfordian clay rock consists of clay materials, carbonates, and quartz. Although such a mineralogical assemblage and the size of its components form a relatively porous rock (10%–18% porosity), the radius of the pores is smaller than 0.1 µm. That pore size, together with the low connectivity of the pore network, generates very low permeability values. Furthermore, radionu-clides are dissolved in water as ions that move and come in contact with clay minerals. Positive ions (cations) may settle on those surfaces or between flakes and be retained by retention or sorption. Due to the small size of the pores and pore thresholds, negative ions may be driven off from the surface of the flakes, which causes their migration to be slowed down by anionic exclusion. Those very slow movements lead, at the scale of the formation, to the creation of fluids at chemical equilibrium with the constituent minerals of the rock.
It is also important to verify that chemical or physical disturbances induced upon the environment do not modify those properties. Construction and operation of a repository will induce the penetration of external elements into the argillite: the introduction of air will oxidize certain components, such as pyrite and organic materials. In addition, certain elements result from the degradation of repository materials, such as concrete and cement. Degrading concrete releases alkaline elements that will raise the pH of water, thus leading to the dissolution of quartz and clay materials and the precipitation of calcite and other silicate minerals, such as zeolites (Savage, 1998). However, the mass ratio between the minerals of the host formation and foreign elements that penetrate it or are generated in it is so high that the extension of chemical disturbances induced by a repository is limited to a few decimeters: in such cases, the clay rock is said to have a buffer effect. Lastly, among the disturbances to be taken into account, thermal disturbances have been studied through an experiment combining the characterization of the thermal properties of the rock and their associated hydromechanical couplings (Lebon et al., 2001).
The mechanical properties of the rock must allow for the construction of large structures under realistic economic conditions. The structure-building techniques must be able to limit as much as possible the disturbances that are likely to alter containment properties, particularly in the case of permeability variations in the immediate vicinity of the drifts. Tests and measurements on samples, as well as excavations, conducted in the Meuse/Haute-Marne underground research laboratory show that the upper part of the host rock is hardened and is only prone to small and slow deformations; those features allow the use of conventional techniques (drill-and-blast and pneumatic hammer) to build underground structures.
Those favorable characteristics are due to the presence of quartz and carbonates that ensure high compressive strength values. Nevertheless, although the rock is not subject to deformations, a fractured zone and a disturbed zone have developed in the lower part of the layer with richer clay content. Fractured zones consist mainly of shear fractures that form on the face and develop into a chevron pattern below the drift floor (Mertens et al., 2004). These fractures are quite spectacular on the face and on the walls when the excavation work follows the direction of major horizontal stress. The initial conditions of those fractures and disturbed zones, as well as their extensions, depend on the orientation of in situ stresses, the excavation method, and the supporting means as work progresses. Nevertheless, it was confirmed that permeability levels at Bure remain low (lower than 10−12 m/s) beyond the fractured zone.
Repository Construction and Operational Options
In the Dossier 2005 (ANDRA, 2005a), ANDRA presented various options for building and operating a repository. These options are compatible with safety and reversibility designs, but at this stage, they only represent a technical overview of what might be a disposal facility, and they are likely to evolve. There remains significant engineering work to be performed before achieving the industrial-development stage of the project.
To ensure the link between surface and underground installations, it was suggested on the basis of the design studies to sink four separate shafts, each with its own function, including: downward transfer of packages; downward/upward transport of personnel; equipment transfer and sludge removal; and ventilation. These would vary between 6 and 12 m in diameter and they would be fitted with mine-tested devices. The shafts would be grouped within the same zone to prevent any hydraulic gradients and, consequently, any circulation among them during the post-closure monitoring phase. That type of shaft layout is similar to the solutions adopted at the Waste Isolation Pilot Plant (WIPP), United States (DOE, 2004), and at Gorleben, Germany (Biurrun and Haverkamp, 2004).
The proposed architecture for underground installations is horizontal with a small vertical extension to ensure the thickest protective layer above and below the repository. Structures are located on a single level. Separate disposal areas have been designed according to the nature of the waste involved, and the repository is divided in subunits. Hence, the different waste categories (B, C, and possibly spent fuel, as the case may be) are sequestered in separate areas subdivided into modules and cells. In the type-B zone, a module corresponds to a single cell, whereas in the type-C zone, a module consists of several hundreds of cells. Modules are built as disposal requirements develop (Fig. 5).
Every subunit is built according to a dead-end geometry and may be sealed from the other subunits. Those arrangements reduce interactions among disposal areas. The modular concept ensures flexible management in the development of the repository, thus allowing for evolutions in geometries and techniques.
Access drifts ensure the link between shafts and disposal drifts. They are specific and dedicated to the transfer of packages and building/sealing materials, as well as to ventilation. Their organization and management are designed to coordinate the various operations involved during the construction, disposal, and sealing of cells and drifts.
Cells, drifts, and shafts are sealed with special swellingclay devices as soon as the shutdown decision has been made. These plugs are designed to prevent water circulation and tend to restore the impermeability of the formation. Disposal areas are backfilled in order to improve the mechanical stability of the overall repository.
Repository operation is dominated by the principle of reversibility. The repository is designed as a series of successive steps to be reached, but with no preset deadline in each case. The transition from one step to the next is neither final nor imposed by a predetermined operational scheme. On the contrary, each step involves various options: return to a previous stage, status quo, or a different solution. The key steps identified in the disposal and shutdown processes are illustrated in Figure 6, as follows:
Once a package has been deposited in a disposal cell, it remains directly accessible for potential retrieval.
Once the disposal cell is sealed, only the top of the cell remains directly accessible, since the cell has been plugged with swelling clay.
Once the module is sealed after all the cells of the disposal module have been backfilled, type-B waste is processed the same way as in the previous step (a single cell), but type-C waste requires the internal drifts of the module to be backfilled.
Once a disposal area is sealed, drifts leading to the disposal area remain accessible.
Once the repository is shut down, shafts are backfilled and sealed. After those conditions have been implemented, the postclosure monitoring phase may start for a period to be specified.
The sealing of the different cells, modules, and drifts is accompanied by the installation of monitoring and observation devices.
Construction of the Underground Research Laboratory and Lessons Learned For A Repository
Two 500-m-deep shafts provide access from the surface to the argillite host rock. The main shaft has a 5 m diameter and allows access for personnel and equipment, material extraction, and ventilation. The 4 m auxiliary shaft, located 100 m away from the main shaft, serves the ventilation system and provides not only mine safety, but also a second access for lowering equipment. As in other underground research laboratories, detailed monitoring of hydraulic, chemical, and mechanical responses makes the construction phase an important part of the experimental program for research. From the shafts, the laboratory has two levels of access drifts at depths of 445 and 490 m (Fig. 7). The upper drift has a simple T-shaped configuration with a total length of ∼45 m. It provides access to boreholes for monitoring shaft-sinking effects through the argillaceous host rock (Lebon et al., 2001).
The several hundred meters of drifts at the 490 m level constitute the key experimental level of the laboratory. Experimental zones are located in a specific area in order to allow the construction and the drift-outfitting work to take place at the same time (Fig. 8).
The choice of a suitable shaft-sinking method was limited to drilling and blasting. This approach was chosen over shaft-drilling methods for several reasons, including the lack of experience with shafts as large as that of the laboratory. To save time, raise-boring was ruled out because both shafts were being sunk in parallel from the surface. The most important consideration, however, was the need to conduct scientific activities and observations in the shaft during construction, which would have been very difficult in a shaft-drilling operation.
The selected shaft-sinking method used a multistage platform with many automated and controlled operations for safety and efficiency purposes. The platform supported all shaft-construction operations, including drilling and blasting, mucking, and application of the concrete liner. The platform was suspended from the surface by four hoist cables. Due to the height of the platform, it only went into operation once the initial 30 m of each shaft was sunk. Once the preliminary construction phase had been completed, the platform continued to operate with its normal cycle of operations. Blast patterns were first monitored and adjusted for effectiveness when the shaft construction started in the limestone, then they were adapted later when more argillaceous rocks were reached. The excavation-damaged zone (EDZ) was monitored through boreholes and geophysical measurements. Due to the homogeneity of the formation, the mucking-shot modalities were hardly ever modified until the completion of the excavation work at depths of 509 and 505 m for the main and auxiliary shafts, respectively.
In the case of repository shafts, the classical drill-and-blast technique is also being investigated. The sinking equipment will be adapted to the foreseeable diameters ranging from 6.5 to 11.5 m, and the breaking will be carried out by blasting or by using a road header. This technique helps to control excavation operations as they progress, the state of disturbances on the rock wall, as well as any potential lithological variations. In addition, it helps to ensure proper support and lining as work progresses.
Wall Lining and Support
In the case of the underground research laboratory, the support system was installed directly and immediately after excavation. It consisted of bolts and wire mesh covered with concrete to prevent spalls. In the Callovian-Oxfordian clay rocks, the zones dedicated to scientific measurements were not bolted but were fitted with sliding arches in order to install the required equipment in the boreholes. Shaft deformations were monitored in five measurement zones distributed over the entire height of the clay formation. Deformations remained within a range of a few centimeters and never exceeded 18 and 45 mm along the major and minor horizontal stresses, respectively. The anisotropy of the deformations corresponds to the anisotropy of the state of stress.
The final lining consisted of concrete poured in 3 m sections at a time. The thickness of the concrete ring was ∼30 and 45 cm in the carbonates and argillaceous rocks, respectively. The installation of that ring prevented any deformation of the shaft (Fig. 9). In addition, the stress within that lining was recorded by vibrating wires that were installed while the concrete ring was being poured.
At the current stage of the studies, the repository shafts are also lined with a concrete ring with a diameter ranging from 50 cm in the limestone formations to 1.35 m in the clay formations.
The access shafts to the underground research laboratory were cased and cemented over their first 20 m to prevent any disturbances in surface aquifers. However, the shafts were also fitted every 6 m with water rings designed to collect waters from the intersected formations in order not only to reduce the thickness of the lining of the laboratory shafts, but also to observe the hydraulic effects of the excavation operations on the Oxfordian limestone formations. It should also be noted that the current pumping rates of the shafts are in the order of 8 and 5 L/min (Fig. 10) for the main and the auxiliary shafts, respectively.
Waste repositories are designed to provide waste isolation. The construction and operation approaches must not compromise the natural isolation provided by the host rock. Hence, an important part of an underground laboratory must be the scientific evaluation of construction methods to minimize the effects of the construction and operation. The primary choices of the construction method for the drifts include controlled blasting, mechanical excavation (pneumatic hammer or road header), and a water-assisted road header.
Because controlled blasting is well established as an effective excavation method, it is used as the reference for planning and design. Blasting is a flexible method of construction that allows freedom in shaping the size of underground rooms. The drawbacks of blasting are the ability to control the blast damages; however, an important part of the experimental program of the underground research laboratory is the measurement of excavation damages and the development of suitable means to minimize excavation-induced effects.
Mechanical excavation using road headers is less expensive and better controlled than blasting. It involves a rotary or percussion cutter mounted on the end of a hydraulic arm. The method is also very flexible in shaping and sizing underground openings for experiments. Mechanical cutting technologies have been used in other underground research laboratories, such as Yucca Mountain (Nevada, USA), the Waste Isolation Pilot Plant (New Mexico, USA), and the Mont Terri Laboratory (Jura Canton, Switzerland) (Thury and Bossart, 1999). One disadvantage of mechanical excavation is dust generation. That problem may be partly resolved by using jet or water-assisted cutters.
Due to the building requirements of the Bure underground research laboratory, the drill-and-blast method was applied at a depth of 445 m, and the pneumatic-hammer method was used to open drifts at a depth of 490 m. Supports consisted of 2.4-m-long bolts and sliding arches. Down at 490 m, the floor was reinforced with bolts. A concrete lining sprayed over wire mesh prevented spalls. Since the drifts are specifically dedicated to the observation and the study of the behavior of the structures, they were not covered by a concrete ring (Fig. 11). Since one of the purposes of the geomechanical measurements is to assess potential convergences and stresses on a final lining, specific zones have been implemented for the measurements of convergences on the supports. The observed convergences depend on the orientation of the drifts and on the excavation and support methods being used. After one year, the measured convergences are on the order of ∼10 cm, and deferred deformations (creeping) are observed.
Concerning the disposal drifts, the current method being used is a road header. The characteristics of this method are well adapted to the nature of the clay and the geometry of the different structures. The proposed support system consists of bolts and concrete-covered wire mesh. The proposed lining consists of concrete either cast directly on site or brought in as precast arches.
Thermal Specifications of the Repository: Acquisition of Elements by A Specific Experiment (Ter Experiment)
The heat released by high-level waste is an essential element for the sizing of the repository, since it is particularly important to minimize damages to the clay and to continue to operate under conditions in which THM (thermo-hydro-mechanical) phenomena are well known and controllable. Hence, it is appropriate to maintain temperatures below 100 °C within the host rock. Beyond that limit, the current state of knowledge does not provide reliable forecasts on the complex mechanisms to be generated and the effects of water-pore vaporization. In addition, it is important to prevent mineralogical transformations and to limit thermomechanical disturbances during the most intense phase. The acquired information on smectites shows that those clays are transformed only when a significant energy, combining both the length of the thermal period and the temperature being reached, is exerted upon them. According to the findings of those studies, a temperature increase of 70 °C over a period of 10,000 yr would only induce very limited irreversible transformations on the Callovian-Oxfordian clay rocks. As a precaution, it was therefore decided that a maximum temperature of 90 °C would be adopted for specification purposes, and it was verified that the temperature would go back down to 70 °C in the repository after 1000 yr (ANDRA, 2005b).
To verify the sizing hypotheses of the repository concept, a thermal experiment (TER) was designed and set up in 2005 in order to specifically study THM processes at temperatures below 100 °C in the rock.
The first objective of the TER experiment was to identify the thermal properties of the rock at the meter-scale and to compare them with the values obtained on samples measuring a few centimeters in length. Discrepancies on the order of 20%–30% were observed between the scales. They were considered to be associated with the size of the samples, the testing methodology being used, and the modifications of the rock when it was removed from its initial environment (loss of mechanical containment, fracturing, desaturation, etc.) (Wileveau, 2005).
The second objective of the TER experiment was to estimate the mechanical and hydraulic responses to temperature increase. The specifications of the experiment included heating the argillaceous rocks over one year in order to reach a temperature of 100 °C on the rock wall. A heat probe including an inflatable rubber membrane was used to ensure thermal conduction toward the rock mass. Based on the preliminary sizing calculations for the test, the probe was installed more than 6 m into the drift wall to prevent disturbances due to excavation operations and desaturation of the drifts. The heating borehole was horizontal and oriented along the major horizontal stress, σH, to prevent significant deformations or spalls from the rock wall. Nevertheless, in spite of common precautions being taken during air-drilling activities in the borehole, bursts were observed during video loggings (Fig. 12) and led to the decision to case the borehole in order not to damage the membrane of the heat probe.
During the heating process, the THM parameters were monitored by appropriate instruments that were designed according to precise specifications based on the particular properties of Callovian-Oxfordian argillaceous rocks (low permeability and water saturation). These specifications were developed at the Mont Terri methodological laboratory, Switzerland, through a similar thermal experiment (called HE-D) from 2003 to 2005 (Kull et al., 2007).
In the case of the heat probe used for the TER experiment, the resistances were internal, embedded within a resin-graphite mixture, and covered with a special grease to ensure conduction. Although one of the packer membranes of the HE-D heat probe deflated during the test (about three months after the test had started), the same inflation system was used for the TER experiment, except that the design was adapted to the existing diameter (100 mm instead of 300 mm). The same peripheral instrumentation was also applied, taking into account the favorable results achieved through the Mont Terri experiment. Hence, PT100 platinum probes were used to measure temperature; water-pressure measurement chambers were installed at the bottom of 20-mm-diameter boreholes, thus reducing the volume to a few cubic centimeters. Deformations were measured by devices brought into the borehole, when needed, to measure displacements between the sealed rings located at every meter along the borehole. More innovative techniques using “optical fiber” Bragg-grating sensors were not used in the TER experiment because of their delicate implementation and their lack of precision with regard to the specifications for such a test (Dewynter-Marty et al., 2005).
Figure 13 shows the layout of the experimental system for the TER experiment. It consists of a heating borehole, three boreholes for the thermometry, two boreholes to measure deformations, and five small-diameter boreholes to monitor pore pressures. Their geographical distribution in the underground research laboratory is shown in Figure 14.
The experimental zone was first equipped with pressure and temperature boreholes. The pressure readjustment was monitored over two months. The reduced volume of the chambers made possible the rapid saturation of the boreholes again. Next, the borehole designed to host the heat probe was drilled. Figure 15 shows the evolution of hydraulic pressures during the reequilibration phase and the coring effects of the heating borehole. On that graph, hydraulic pressures stabilized at various values (between 22 and 36 bars), whereas the hydraulic pressure recorded by probes that have been installed for many years in undisturbed environments is on the order of 45 bars (Delay and Cruchaudet, 2004). Since pore-pressure sensors are positioned at ∼7 m from the drift wall, it is possible to infer that the selected zone for the thermal experiment is influenced by the layout of the drifts surrounding the TER-experiment zone (Fig. 14). Coring effects on pressure as the coring machine passes are uneven. A vertical pressure drop of close to 17 bars and a horizontal pressure increase of ∼20 bars indicate that the vertical stress in the investigated zone is lower than the horizontal stress, which is not consistent with the current state of knowledge concerning the tensor of natural stresses at such a depth in the Callovian-Oxfordian formation, where σv ≅ σh ≅ 12.5 MPa (Wileveau et al., 2007). That measurement therefore reinforces the hypothesis of a heavily mechanically disturbed excavation-damaged zone with a higher vertical stress in the pillar formed by the three adjacent drifts.
Hence, it is necessary to find a rock-behavior law that would integrate hydromechanical coupling and drift desaturation in order to provide sound interpretations of the thermal test and reliable parameters for THM couplings. Once the initial state of stress is ascertained, it will be possible to correctly analyze the thermal effects. The heating period started in 2006 and is scheduled to last over one or two years.
This example shows that experiments in underground research laboratories are not limited to the acquisition of physical properties in situ, but they are also able to detect situations that are likely to occur in a repository, and then they may sensibly modify the repository's design and operation.
The experimental program of the underground research laboratory addresses two major issues: it demonstrates the natural isolation capability of argillite and the feasibility of constructing and operating a repository without compromising those isolation properties. The rationale for argillite disposal is twofold. First, the disposal concepts in argillaceous rock formations hypothesize that groundwater does not move by advection in some clay-rich rocks, and radionuclides will be transported by molecular diffusion with strong chemical retention. The second rationale is that damage to the rock caused by excavation and repository operations may be controlled and possibly reversed by the self-healing properties of clays under mechanical loading or geochemical interactions.
The experiments carried out at the French underground research laboratory involved the following: a thorough understanding of the relevant aspects of the geological, hydrological, and geochemical settings; determination of the hydraulic properties of the argillaceous rocks over a range of scales, using direct hydraulic measurements and inferences from coupled chemical and mechanical behaviors; monitoring of construction and the controlled excavation of underground openings in order to characterize the excavation-damaged zone; and testing of methodologies for excavation-damaged zone mitigation and remediation.
Through its construction and experimental activities, the laboratory has helped ANDRA to develop a concrete approach with a view to proposing suitable architectures and management methods for a repository. Thanks to the lessons learned from the different technical and scientific means implemented, it is possible to characterize the feasibility of such a facility within the Callovian-Oxfordian formation by providing adequate long-term safety guarantees.
Future work in the laboratory will include the life-size construction of the different components of a disposal facility, such as the cells or the plugs for cells and drifts. Once specifications are set, it will be possible to draw a concrete preliminary design that will integrate the specific characteristics of the selected zone for the implementation of the repository.
Figures & Tables
Deep Geologic Repositories
Deep Geologic Repositories reviews the success stories of underground waste isolation. It focuses on repositories that did, do, and will permanently and safely isolate dangerous materials from the near-surface biosphere. Complementary topics address the isolation capability of average crustal rock, investigations at one representative underground research laboratory, and the geologic preservation of fission products from Precambrian nuclear reactors. An international cast of contributors presents proven practical solutions to a formerly confounding issue in environmental and engineering geology: What do we do with wastes that retain their dangerous characteristics in human terms forever? The principal answer: Recycling into the lithosphere by “reverse” mining.