Fracture transmissivity as a function of normal and shear stress: First results in Opalinus Clay

• Published in: Physics and chemistry of the earth. Parts A/B/C 2011, Vol.36 (17), p.1960-1971
• Main Authors: Cuss, Robert J., Milodowski, Antoni, Harrington, Jon F.
• Format: Article
• Language: English
• Subjects: Argillaceous rock; Fracture flow; Normal stress; Opalinus Clay; Shear; Time dependency
• ISSN: 1474-7065; 1873-5193
• DOI: 10.1016/j.pce.2011.07.080

► Fracture transmissivity reduces with increased normal load in Opalinus Clay. ► At constant normal load transmissivity was seen to change whilst the fracture sheared. ► Fracture flow from a central injection point was seen to be complex and not simple radial flow. Fracture transmissivity has been investigated along an idealised fracture for the influence of normal stress and for the transient behaviour during a slow shear experiment. A linear trend for the relationship between effective stress and transmissivity has been proposed for normal loads between 1 and 5 MPa; as effective stress increases transmissivity decreases. Transmissivity was very low throughout the complete spectrum of effective stresses examined and was close to the permeability for intact Opalinus Clay, suggesting that the fracture had effectively closed. During active shearing at a constant normal load, fracture transmissivity was seen to initially reduce, probably due to clear smearing. A series of flux events were seen, with transmissivity increasing by a factor of four. Some of the flux events corresponded with dilation, whilst others did not. This suggests that the opening flow paths were localised and did not result in bulk dilatancy. During the course of the shear test the sample formed its own series of fractures and a complex pattern of deformation occurred along the fracture surface to a depth of less than 1 mm. The impression of the end of the injection hole clearly shows that the block underwent at least 5 mm of the total 6 mm of shear displacement. The injection of fluorescein showed that flow along the fracture was not uniformly radial, as one might expect for such an experimental geometry. At the time of injection there were a number of dominant flow features, mainly in the direction of shear and only perpendicular on one side of the fracture surface. Flow occurred along the original fracture surface as well as the newly formed shear surface, indicating multiple pathways in a complex manner. The evolution of fracture transmissivity is very complex, even along initially planar surfaces. Fracture transmissivity has been seen to be a function of normal stress and porewater pressure, and has also been seen to be a dynamic feature during shear.