PH4: A 250M deep borehole in Tournemire for Assessing the reliability of Chloride-, Helium and water stable isotopes profiles in the TOARCIAN/DOMERIAN shales

Titre du congrès :Clays in natural and engineered barriers for radioactive waste confinement Ville du congrès :Lille Date du congrès :17/09/2007

Introduction One method to obtain information about fluid flow and solute transport in porous media of low hydraulic conductivity and water content is based on the study of natural-tracer distribution in pore-water. Such an approach has been previously applied to the Tournemire Toarcian/Domerian shales and, based on the profiles of the water stable isotopes and chloride in the pore water, it was concluded that diffusion might be the main transport mechanism in these rocks (Patriarche et al., 2004). However, the low water content (2 to 5 %), the small pore size (few nm) and the charged clay surfaces of such rocks make the determination of such natural tracer concentrations difficult and measurements are easily subjected to artefacts that deviate the measured data from that actually present in situ. For example, a comparison of different vacuum distillation and isotope exchange methods to determine the pore water isotopic composition revealed that the vacuum distillation method resulted in systematically lower, geologically difficult explicable, 18O and 2H values, making the previous studies based on this technique questionable (Altinier et al., 2007). The present study aims at assessing the reliability of various techniques used for the determination of natural–tracers in pore water by the comparison of several natural tracer profiles (chloride, water stable isotopes and helium) with the tracer data obtained from adjacent samples from the same borehole. Experimental concept Drillcore samples were collected from a 250m-deep vertical borehole drilled in winter 2006/2007 from the tunnel down to the lower calcareous aquifer and penetrated through 220m of Toarcian/Domerian argillite and marls. The borehole was air-drilled throughout in order to prevent contamination of the natural tracers by drilling fluid. The water stable isotope contents in the pore water have been approached by two methods. At intervals of 10 m, core material was recovered and conditioned on-site (removal of the core rim, crushing, placing in a vapour-tight container together with an isotopically spiked solution) for the vapour exchange method for water isotopes. From the same core piece less than 10cm apart, another sub-sample was prepared for radial diffusion experiments (see Savoye et al., 2006a & 2006b for details). The supernatant solution of these experiments was analysed for its isotopic composition as well as for its Cl concentration. In addition, this method also allows the derivation of the geochemical porosity of the rock and its diffusion parameters. For the determination of pore-water helium concentrations, about 10 freshly drilled rock samples were conditioned on-site to octahedral cylinders by dry-cutting the core rim and were immediately transferred in He-tight stainless steel containers. The containers were repeatedly flushed with nitrogen and evacuated, to remove excess air and finally left under a slight under pressure. Complete outgassing of He is attained after several weeks to a few months and He will be analysed by mass spectrometry. Possible contamination of the samples with air is monitored by the simultaneous measurement of the Ne isotope ratio on the recovered gas. Other core samples were dedicated to on-site petrophysical measurements such as the gravimetric determination of the water content by drying at 150°C to stable weight conditions and still other sample were preserved in aluminium bags under vacuum for their mineralogy. Results Although some of the experiments take months until their full equilibration and thus the analytical work has not yet been completed, some first results are already available: The hydrogen isotope ratios of the pore-water derived by the vapour exchange technique describe a profile with a “classical” V-shape: from 2H values of about -45‰ V-SMOW at the tunnel level, deuterium contents regularly increase up to 2H values of -30‰ in the lower Toarcian and drop down again to -50‰ in the lower Carixian aquifer. A discrepancy of about 20‰ is observed between the present 2H values and those obtained by Patriarche et al. (2004) for a part of the present profile. This difference persists along the entire stratigraphic column and must be attributed to the artefacts linked to the applied vacuum distillation technique, as already highlighted by Altinier et al. (2007). Similar is observed for the 18O profile showing the same vertical trend the 2H values, but less accentuated. Although the data acquisition is still in progress, these first findings already show the importance of having a) an improved understanding of artefacts introduce by sample conditioning and methods applied, b) a control on the obtained tracer concentrations by applying complementary methods, and c) a complete multi-tracer profile that extends to the bounding aquifer what will allow a much better constrained modelling of the tracer data.