Water retention properties of Callovo-Oxfordian clay-stone

Document available in extended abstract form only. Many investigations were carried out on the Callovo-Oxfordian (COx) clay-stone that has been selected by the French radioactive waste management agency (ANDRA) as a potential host rock for high level radioactive waste disposal at great depth. Various authors demonstrated the significant water sensitivity of clay-stones. Some cracks generated by desaturation have been observed in the Tournemire URL (France) excavated in clay-stone. By carrying out some ESEM (environmental scanning electron microscope) observations on COx clay-stone samples submitted to cyclic changes in relative humidity, Montes et al. (2004) evidenced the water sensitivity of the clay fraction, with significant successive openings and closures of cracks and pores. Hysteresis effects have been observed in the water retention curve, in the ultrasonic velocity evolution and in the strain changes induced by hydration cycles on COx samples submitted to controlled relative humidities by Pham et al., (2007). Better knowledge of the mechanism of desaturation of the clay-stone is necessary to further understand the changes in hydro-mechanical properties of the excavation damaged zone (EDZ) that becomes significantly de-saturated all around the galleries because of ventilation. In this framework, a detailed study of the water retention properties of the COx clay-stone was carried out i) to complete existing observations with respect to volume changes under suction cycles and ii) to determine the main drying and wetting curves so as to better investigate hysteresis effects in the water retention curve.. Particular attention was paid to the characterization of the initial state in the laboratory (where samples are provided unsaturated due to the combined effect of coring, storage and transportation) and to the volume changes and changes in degree of saturation along the main wetting and drying paths. The determination of the water retention properties was carried out by suction control by vapour control followed by the measurement of the water content once the sample equilibrated. All tests were run on samples of 38 mm diameter and 10 mm high trimmed. At all stages, careful determination of volume changes were carried out by hydrostatic weighing of waxed samples, providing the changes in volume and degree of saturation with suction. Figure 1e clearly shows that the first wetting path from initial state is comprised between the main wetting and main drying path, a well-known feature in unsaturated soils that is confirmed here for the COx clay-stone. Starting from an initial state at w = 7.5% (suction around 15 MPa) the water content at zero suction is between 11 and 12%. The COx sample is able to retain 2.6% water content under a 150 MPa suction. Figure 2, in which volumetric changes are plotted, shows that a significant swelling of around 7% occurred along the wetting path. Figure 2 also evidences a 1.5% shrinkage along the drying path. The changes in degree of saturation with suction are presented in Figure 1d. Starting from the initial state (Sr = 92.6%), it seems that saturation gets close to 100% when reaching a 4.2 MPa suction. Interestingly, the main drying path is identical in this diagram, providing a value of the air entry value of about 6-7 MPa. Note however that this air entry value is that of a swollen sample somewhat different from the natural clay-stone that start de-saturating from an unswollen state along the gallery wall just after excavation. Inspection of Figures 1a and 1c indicates that desaturation starts at a water content of 9.3% after a saturated shrinkage from an initial water content of 12% with a decrease in porosity from 23 to 20.3% (to compare to the initial value of 17.6%). The shrinkage observed along the first drying path at suction larger than initial indicates that the porosity measured in the laboratory on somewhat de-saturated and shrunk samples is not exactly the in-situ natural porosity that should be higher. Actually, the exact porosity is not straightforward to estimate. As suggested by Monfared et al. (2011), it seems that the best way to estimate it in the laboratory to re-saturate the sample under in-situ stress state. The water retention data obtained here provide further understanding on the sensitivity of the COx clay-stone volume changes with respect to changes in water content, confirming significant swelling-shrinkage behaviour. The detailed determination of the drying and wetting paths as presented in Figure 1 will help better characterise the response of the clay-stone submitted to drying-wetting path in the close field. (authors)