Fractures and self-healing within the excavation disturbed zone in clays

In-situ test 4 is quite similar to in-situ test 2 but for Boom Clay. The natural closure of a borehole is studied. Permeability and possibly seismic measurements were performed to follow the evolution of flow properties through the borehole. The testing is performed in the following way: - Drilling of a test borehole - Installation of the test equipment - Monitoring of the borehole convergence (creep) and of the EDZ - Hydro testing to evaluate transmissivity along borehole We can conclude from in-situ test 4 that long-term seismic and AE measurements are able to reveal disturbances and detect the evolution of sealing/healing processes with time in Boom Clay by means of measuring the associated variations in seismic velocities, damping of amplitudes and frequencies of the seismic signals. Collapsing of an uncased borehole entails a decompression of the surrounding clay, which leads to an increase of deviatoric stresses and a decrease of pore water pressure. Both effects can lead to a decrease of seismic velocities and amplitudes, which was observed in the present seismic measurements. After the decompression phase, the stresses as well as the pore pressure tend to recover. Due to on-going reconsolidation, the acoustic transmissibility of seismic waves increases again in Boom Clay. The increasing trend could be observed rather in the seismic amplitudes than seismic velocities. Experiments on a reinstalled fractured clay core demonstrated that seismic parameters recover and fractures seal within several weeks or months.

Performance Assessment is concerned with the potential radiological consequences of the disposal system. Hence, the analysis of the evolution of the repository system, of which the EDZ is a part, is an essential part of PA. The EDZ is initiated during the repository construction. Its behaviour is a dynamic problem, dependent on changing conditions that vary from open-drift period, to initial closure period, to the entire heating-cooling cycle of the decaying waste. Other factors concern the even longer-term issues of chemical reactions and biological activities. The EDZ originates in the redistribution of stress that results from excavation and/or excavation method. Fracturing of the rock cannot be avoided in practice. However it is possible to limit the extents of the fractures by controlling the convergence. During the consolidation phase sealing occurs; the open fractures close progressively. During the exploitation phase, sealing can be slow or even inhibited due to the ventilation of the galleries which may prevent a full saturation of the host-rock. Open fractures favour the oxidation of the host rock. However, this oxidation phenomena has only been observed along fracture planes. Within the clay matrix, no evidences of oxidation were found. After the closure of the disposal galleries, the system will progressively resaturate and suction will disappear. The backfilled galleries will be saturated with water. Favourable conditions to sealing will be recovered. In case of indurated clay host rock bentonite backfill material generating swelling pressure are considered to accelerate the sealing process. Consequently, it can be assumed that the remaining open fractures will seal and reducing chemical conditions will progressively be restored within the EDZ. Once saturation is attained, the thermal load may cause an increase of stresses because of thermal expansion of solids and pore water and heating-related reduction of strength. This process could further increase the size of the EDZ but in turn, heating may also increase creep rates and thus accelerate the closure of open fractures and reduce the influence of EDZ. Peak temperatures within the EDZ should normally be reached after a few tens of years. A subsequent slow cooling phase will follow. Finally, EDZ and engineered barriers will chemically interact. Safety assessment calculations for Opalinus Clay and Boom Clay showed little impact of the EDZ on radionuclide release even for very conservative assumptions. The different experiments realised in the frame of SELFRAC show that the effective hydraulic conductivity of the EDZ in the considered rocks will be lower than 10-10m/s within several years as soon as the bentonite backfill of the emplacements drifts become fully saturated and the expected swelling pressure will be build up. The hydraulic parameters of natural fractures in Opalinus Clay indicate that the long-term conductivity of the EDZ is even lower. Consequently, the maximum hydraulic conductivity of the EDZ is expected to be approximately one order of magnitude higher than the value of intact rock, which is in the case of Mt. Terri in the overall range of 2x10-12 to 2x10-14m/s (Heitzmann, 2004). Moreover the assumption that the presence of an EDZ provides a fast path for the escape of solutes from the canister to the geosphere and biosphere is now recognized to be an oversimplification. It may be true that the EDZ is, at least over a period of time, a zone of relatively high permeability, but whether flow can take advantage of it to transport solute to the accessible environment requires an evaluation of the total flow system. Thus, if the high-permeability zone is surrounded by low-permeability regions, or the hydraulic gradient is sufficiently low, there will be an insufficient supply of flowing water in the EDZ to negatively impact the repository performance. Consequently EDZ should not be considered as a critical issue for the performance assessment of radioactive waste repositories in argillaceous formations. For the safety cases, however, a profound scientific knowledge on the subject is necessary. The results of the SELFRAC project undoubtedly strengthen the sound scientific background. Nevertheless there are some important remaining questions, mostly related to the response of the EDZ to gasses and chemical and thermal changes through time. The understanding of these phenomena is also necessary to support future safety cases.

A new gallery was realised to extend HADES. In-situ test 3 consists in studying the evolution in function of time of the hydro-mechanical properties within the EDZ around the gallery during a period of three years after its construction. For this purpose two multi piezometers were installed around the connecting gallery so that we can follow the evolution of the flow properties with time and self-boring pressuremeter tests were performed to follow the evolution of the mechanical properties. We can conclude from in-situ test 3 that gallery construction in Boom Clay at an important depth (223 m in the case of the URF HADES) will always induce fractures. Fracturation is caused by stress redistribution, which is inherent in tunnelling. Fracture extent can however be limited by using appropriate excavation and lining techniques. It was observed that these fractures seal but only heal partially. In-situ test 3 has shown that no interconnected fracture network exists beyond (at most) a few decimetres into the host rock. Furthermore, hydraulic conductivity around the gallery seems not to be influenced by fractures. Higher k-values (up to about a factor 2) are observed up to 6-8 m and are probably caused by the lower level of effective stress in that zone. With time, k tends to become lower (when measured on a vertical piezometer) although further measurements are needed to confirm this. Unlike k, pore pressure is influenced up to an important extent by the presence of the gallery. The anisotropic aspect of pore pressure distribution can be explained by stress anisotropy (on the shorter term) and by anisotropic hydraulic conductivity of the host rock (on the longer term). Material parameters are only influenced slightly and up to a limited extent (2-3m).

The most important parameter for performance assessment to be measured in the EDZ is the axial transmissivity along tunnels, as this preferential flow path could act as a shortcut for radionuclides. All tests performed so far are designed to measure the local effect of fractures, but not the change of the axial transmissivity, which is controlled by the interconnectivity of single fractures. A 1:1 scale test of the axial transmissivity along tunnels is a very difficult and expensive task. Therefore, it is planned to investigate the axial transmissivity on a much smaller scale along a borehole. Previous laboratory and in situ tests have indicated that under special boundary conditions borehole instabilities will occur in Mont Terri clay. The main reasons for borehole instabilities are the orientation of the boreholes with respect to the direction of the material anisotropy (transversely isotropic material) or bedding, the stress field and the chemistry of the borehole fluid. The most stable conditions prevail for boreholes drilled with air perpendicular to the bedding, while the worst case results when the axis of the borehole is oriented parallel to bedding and water penetrates into the borehole. During the in situ test, a borehole is air drilled parallel to bedding. The borehole will be instrumented with a double-packer system to perform the testing. One packer is a combined hydrotesting / dilatometer tool which is able to isolate a borehole interval with controlled pressure and at the same time to measure the deformation of the borehole wall as a function of inflation pressure. The test is performed in the following way: - Drilling of a test borehole with air - Installation of the test equipment and filling of the borehole with synthetic formation water - Inflation of the dilatometer up to the minimum pressure required to get the displacement sensors in contact with the borehole wall - Monitoring of the borehole convergence (creep) and of the EDZ - Hydro testing to evaluate transmissivity along borehole (packer by-pass) - Increase of dilatometer pressure in several steps and monitoring of the associated creep (closure of EDZ) and change in transmissivity per pressure step It is shown that an EDZ with a distinct higher transmissivity than surrounding the rock mass develops in short time around a borehole and that this transmissivity can be reduced as a function of the stepwise increase of the inflation pressure of the dilatometer from 0.5 to 5 MPa. The observed hydraulic properties of the EDZ around the borehole appear to be controlled by the load applied to the borehole wall. When the load applied to the borehole wall reaches 5 MPa, a reduction of the transmissivity of several orders of magnitude to about E-13 m2/s of the axial flow path along the dilatometer is observed.

A review and reporting of the state of the art on basis of existing theoretical studies and in situ observations realised in underground laboratories has been performed. A clear distinction has been proposed within the SELFRAC project, between the Excavation Damaged Zone (EDZ) and Excavation disturbed Zone (EdZ). The definitions were established in the context of underground radioactive waste disposal, establishing a bridge between geomechanics and PA: - The Excavation disturbed Zone (EdZ) is defined as a zone with hydro-mechanical and geochemical modifications, without major changes in flow and transport properties. Within the EdZ there are no negative effects on the long-term safety. - The Excavation Damaged Zone (EDZ) is defined as a zone with hydro-mechanical and geochemical modifications inducing significant changes in flow and transport properties. These changes can, for example, include one or more orders of magnitude increase in flow permeability. These definitions were internationally discussed and accepted at the conference "Impact of the excavation Disturbed or Damaged Zone on the performance of radioactive waste geological repositories" held in Luxembourg 3-5 November 2003. However the definitions should be deepened for each type of rock and each site. In particular, the terms significant and major should be quantified and the time effect should be determined for each particular site.

The modelling of the excavation process and the strain localization prediction along the process has been performed using the finite strain Finite Element Code Lagamine, with the constitutive law CLoE. A special modelling procedure was designed, taking into account the progressive excavation as a process in time, the placement of a lining behind the excavation front, the existence of a over-excavation with respect to the lining dimensions, and the hydromechanical coupling. The results are consistent with the site observations, especially the localization are predicted. Further refinements would be necessary to reduce some discrepancies observed with respect to site data.

In the underground laboratory of Mont Terri, a comprehensive study of the EDZ has been carried out. The observed induced fractures, which generally have been created under local tension, led to an increase of the local permeability (several orders of magnitude) in the vicinity of the tunnel. At several locations, the behaviour of the disturbed region was observed during water circulation in the EDZ. During this re-saturation phase a significant reduction (1-2 orders of magnitude) of the permeability was measured, probably due to swelling of the clay, but the EDZ still showed a higher permeability in comparison to the intact rock mass. This observation is not in agreement with laboratory measurements on core samples in a triaxial pressure cell where a nearly perfect self-sealing of fractured rock samples was observed. This discrepancy could be explained by the effect of normal stress on the fracture planes in the laboratory tests which is not present in situ in the vicinity of the tunnel. Within this in-situ test, the effect of a normal load on the tunnel wall is studied. In a real repository the waste canister is surrounded by a highly compacted bentonite buffer that will swell if water flows in the near field of the disposal tunnel. The swelling pressures expected of a high-level waste repository will be in the order of several MPa and will possibly reach the lithostatic pressure in an equilibrium. This effect is studied under controlled boundary conditions where the load on the EDZ is applied with loading plate equipment in the tunnel. The load plate covers a region in the tunnel where transmissive features have been observed and characterised. The in situ experiments performed in Opalinus Clay show that the effective hydraulic conductivity of the EDZ is expected to be relatively quick, within several years, lower than 10-10m/s as soon as the bentonite backfill of the emplacement drifts becomes fully saturated and the expected swelling pressure will be build up.

Leistungen

Diese Seite teilen

Herunterladen