Hydro-fracture in the laboratory (NERC): Linking fracture networks to permeability and seismicity using rock physics as a laboratory tool

This project presents a new method for simulating hydro-fracture in a controlled laboratory environment in which key parameters of stress, anisotropy and fluid activity (temperature/acidity) may be measured and controlled in order to test current hypotheses and theories on fluid-driven fracture mechanics. The method is essentially a variation of the classic thick-walled cylinder test using an over-pressurized bore to fracture an outer shell (specimen) in tension. Unlike previous versions, the entire assembly is encased in a rubber jacket fitted with ports for Acoustic Emission to monitor fracture development and measure the fluid through the freshly generated tensile fracture. The assembly is mounted in a triaxial vessel allowing a range of pressure/temperatures to ~4km reservoir depths to be simulated.

Two hypotheses are tested: (1), stress and anisotropy has a key influence on the fracture pattern produced and, (2), the swarm of AE location follow a diffusion pattern. The study concludes that anisotropy exerts a significant control on fracture characteristics and the pressures required to achieve failure. In general, the hydro-mechanical characteristics of the fracture process in Nash Point Shale are similar for both normal and parallel bedding: maximum fluid injection pressure is followed by a rapid fluid pressure decay and an oscillation phase. However, when testing samples with normal bedding orientation, significantly higher fluid pressures were required to initiate hydraulic fracture normal to bedding. With increasing confining pressure the ratio of the required fluid pressure between Short-Transverse and Divider orientation did not change over the tested pressure range, suggesting that the effect of anisotropy does not diminishes at elevated pressures. From radial deformation behaviour and AE activity in samples with normal bedding, it may additionally be concluded that once the fracture reaches the outer sample edge (radially, divider orientation), propagation still continues in the vertically sense, but in the arrester orientation.