Borehole sealing
After hydration of the pre-compacted blocks of FoCa and Serrata clay, the permeability in the borehole seals was measured. A hydraulic conductivity of 4.3 10-13 ms-1 and 5.5 10-13 ms-1 was obtained for FoCa clay and Serrata clay, respectively. These values are 10 times lower than the hydraulic conductivity of undisturbed Boom Clay and are in good agreement with the laboratory measurements at the corresponding dry density. To test the gas sealing ability, a gas injection was carried out. Gas was injected in the host clay formation from a filter close to the FoCa seal compartment at the end of the borehole. The gas breakthrough occurred at a pressure level of about 3.1 MPa, i.e. a pressure equal to the radial total stress measured in the FoCa seal. The breakthrough has been detected on a filter in contact with Boom Clay, closer to the gallery. The pressure sensors connected to the filters in the FoCa seal show a very weak reaction to the breakthrough.

Shaft sealing

The relative humidity sensors inside the seal shows that the seal is near saturation after 4 years of artificial injection.

figure 1: Pore water pressure evolution in the filters of a radial piezometer. Before the removal of the lining a constant pore water pressure is observed. A sudden decrease in pore water pressure is noticed due to the removal of the lining. A slow re-increase of pore water pressure is later on noticed. This increase goes progressively from the furthest filter to the filter closest to the seal - host rock interface

Figure 2: Detail of figure 1 showing the sudden decrease in pore water pressure related to the removal of the lining, followed by a small increase in pore water pressure.

The installed piezometers provide quantitative data on the EDZ evolution around the seal (Fig. 1). The radial piezometers show, after installation and before removal of the lining, a quite constant pore water pressure, illustrating an equilibrium state. As a result of the removal of the lining a sudden decrease in the pore water pressure is noticed. The pore water pressure drops to a value lower than the detection limit of the measurement set-up. A more detailed view (Fig. 2) of the pore water pressure evolution shows a pressure increase following this sudden decrease. The time of occurrence for this sudden increase is related with the distance to the experimental shaft wall. The pressure evolution after this step and the curve of the air atmospheric pressure are qualitatively the same. This comparison indicates that at this time the sensors connected to the filters of the piezometer measure the atmospheric pressure. This means that there is a direct connection between the filter and the atmosphere. In other words, fractures have been developed through the host rock due to the decompression of the host rock around the shaft where the lining has been removed. The instrumentation was limited to 1 m around the shaft, showing that fractures occur at least up to 1 m distance from the shaft/clay interface.

Borehole sealing
The permeability of the seals is a factor 10 lower than the permeability of the host rock formation and the behaviour of the seals under gas injection conditions confirm the feasibility of producing borehole seals. The experimental results have demonstrated that it is possible to seal with a bentonite material the excavation disturbed zone around a borehole.

Shaft sealing
The hydration of the sealing material takes much longer than originally foreseen. After 4 years of artificial hydration, several observations indicate a near saturation of the seal. The removal of the concrete lining has a strong effect on the EDZ. Fractures were formed around the shaft, causing many clay blocks to fall from the host rock wall. It is impossible to distinguish between re-activation of existing fractures (caused during excavation) and newly developed fractures caused by removal of the lining. The fractures in connection with atmospheric pressure occur up to at least 1 m depth in the host rock, but are limited to a cylindrical zone where the lining has been removed. The self-healing of the fractures is a slow process. However, it should be taken into account that the speed of self-healing can not be deduced from this RESEAL experiment, as here it is largely influenced by the hydration of the bentonite seal. The available data clearly show that the self-healing is a progressive process starting in the host rock and evolving towards the interface host rock - lining.