Stress-induced permeability anisotropy in sandstone under true triaxial loading

Abstract

The change in elastic properties of rocks (wave velocity, elasticity modulus) as a function of stress orientation is a well-documented phenomenon known as stress-induced anisotropy. Rigorous studies of rock samples loaded in a Conventional Triaxial Apparatus (CTA, ) have revealed crucial insights on the magnitude of anisotropy with increasing differential stress, the extent to which it is recovered after unloading, and its driving micromechanical mechanisms. However, how stress-induced anisotropy affects the direction of fluid flow within saturated rocks subjected to anisotropic stresses is still not well understood. This is likely a result of CTAs being limited to permeability measurements only in the direction of the axis, impeding the quantification of the anisotropy of permeability as a function of differential stress. Furthermore, these instruments neglect the effect of the intermediate principal stress axis in stress-induced anisotropy, complicating the extrapolation of results to crustal conditions.A novel True Triaxial Apparatus (TTA), developed in the UCL Rock & Ice Physics Laboratory, allows us to subject cubic rock samples (50 mm edge length) to truly triaxial stresses (). This apparatus is equipped with a confining pressure system and independent pore pressure intensifiers that permit permeability measurements along the three principal stress axes using the steady-state flow method. To quantify the anisotropy in permeability driven by truly triaxial stress states, a cubic sample of high porosity, quasi-isotropic, Darley Dale sandstone (DDS), was introduced in the TTA and subjected to anisotropic stresses whilst permeability measurements were made along the three axes for each stress state. Differential stresses were kept below the onset of dilatancy for this rock type to ensure that deformation remained in the poro-elastic regime. Preliminary results indicate that permeability variation between stress axes can reach up to one order of magnitude at differential stresses as low as 15 MPa. These initial findings suggest that stress-induced permeability anisotropy can be significant. Further experiments are needed to fully describe the anisotropy of permeability under a wider spectrum of stress states and to ensure the reproducibility of results, and this is currently a work in progress.