Technical Report NTB 03-07

Diffusion of HTO, 36Cl-, 125I- and 22Na+ in Opalinus Clay: Effect of confining pressure, sample orientation, sample depth and temperature

The Opalinus Clay (OPA) formation in the Zürcher Weinland is a potential host rock for a repository for spent fuel, vitrified high-level waste and long-lived intermediate-level waste in Switzerland. Owing to its small hydraulic conductivity (10-14 - 10-13 m·s-1), it is expected that transport of solutes will be dominated by diffusion. This study addresses the diffusion of tritiated water (HTO), 36Cl-125I-­ and 22Na+ through Opalinus Clay samples. The samples were collected in the Mont Terri Underground Rock Laboratory, where the OPA formation is located at a depth between -200 and -300 m below the surface, and in the deep borehole in Benken (Zürcher Weinland), where the OPA layer is located at a depth between -539 and -652 m.

Effective diffusion coefficients (De), rock capacity factors (α) and diffusion-accessible porosities (ε) were measured using the through-diffusion technique. Transport (diffusion) was measured both perpendicular and parallel to the bedding. Special cells that allowed the application of an axial confining pressure were designed. The pressures applied ranged from 1 to 5 MPa for Mont Terri samples and between 4 and 15 MPa for Benken samples, the upper values representing the in-situ confining pressure at both locations. The test solutions used in the experiments were synthetic Opalinus Clay pore water, which has Na and Cl as main components (Mont Terri: I = 0.39 M; Benken: I = 0.20 M).

Pressure only had a small effect on the value of the effective diffusion coefficients. In the case of Mont Terri samples, increasing the pressure from 1 to 5 MPa resulted in a decrease of the effective diffusion coefficient of 20% for HTO, 27% for 36Cl-, 29% for 125I-­­ and 17 % for 22Na+ In the case of Benken samples, increasing the pressure from 4 to 15 MPa resulted in a decrease of De of 17% for HTO, 22% for 36Cl-, 32% for 125I-­ and 17 % for 22Na+. Moreover, the effective diffusion coefficients for 36Clare smaller than for HTO, which is consistent with an effect arising from anion exclusion. This ion exclusion effect is smaller in samples from Mont Terri than in samples from Benken, which can be explained by the higher ionic strength of the Mont Terri water used in the experiments. The diffusion of 22Na+ is similar to that of HTO in the case of Mont Terri OPA. For Benken OPA, the De value of 22Na+ is a factor of 2 higher than that of HTO. This last observation cannot be explained so far but is comparable to experimental data from ANDRA (1999) on Callovo-Oxfordian claystones from the Meuse/Haute Marne site.

125I is retarded with respect to 36Cl-. This is caused by a weak sorption of 125I-­ on the Opalinus Clay. The distribution coefficients, calculated from the rock capacity factor under the assumption that the diffusion-accessible porosity of 125I- is the same as for 36Cl-, range between 0.01 and 0.02 cm3·g-1. The effective diffusion coefficients of 125I-­ are comparable with those of 36Cl-.

Out-diffusion data of HTO, 36Cland 22Na+ are in good agreement with the through­diffusion data. In the case of 125I-­ the agreement is less. The flux calculated with De and α derived from through-diffusion measurements is smaller than the observed flux. This indicates that other (unknown) processes are taking place.

The diffusion coefficients measured in this study on Mont Terri samples are in good agreement with recent measurements of three other laboratories, within the framework of a laboratory comparison exercise. The values of the diffusion-accessible porosities, however, show a larger degree of scatter, indicating that through-diffusion is not the method of choice for obtaining reliable porosity values.

Diffusion parallel to the bedding is higher than diffusion perpendicular to the bedding. The effective diffusion coefficient for diffusion parallel to bedding is a factor of 4 – 6 larger than for diffusion perpendicular to the bedding. This is due to the layered structure of the Opalinus Clay, resulting in a smaller tortuosity factor for diffusion along the bedding planes. The observed effect was similar for HTO, 36Cl- and 22Na+. This anisotropy is more pronounced for the Opalinus Clay from Benken than for Mont Terri, indicating that the clay platelets are more preferentially oriented in the case of Benken OPA.

The temperature dependence of diffusion of HTO in OPA is of an Arrhenius type. The activation energy (22 kJ·mol-1), however, is larger than for diffusion in bulk water (18 kJ·mol-1). This indicates that confined water in the narrow pores of the Opalinus Clay has partly a different structure.

Diffusion measurements with HTO on OPA samples from different depths showed that the effective diffusion coefficients for diffusion perpendicular to the bedding decrease with increasing depth. The difference between the top and the bottom of the OPA layer, however, is not more than a factor of 1.5. For diffusion parallel to the bedding, no difference between top and bottom could be observed. It can be concluded that the Opalinus Clay layer is very homogeneous with respect to its diffusion properties.

The effective diffusion coefficient measured for the HTO in OPA is in good agreement with values measured in other sedimentary rocks and can be related to the porosity using Archie’s Law with exponent m=2.5.