Natural fracture shear slip can induce an immediate increase in permeability. This permeability increase is the goal of shear stimulation via fluid injection. However, shear slip could breach caprock seals that are key to containing fluids in the subsurface and this slip can be seismic, causing damage to buildings and infrastructure. To predict the effect of slip on subsurface flow and transport, models require relationships that link fracture properties to porosity and permeability. Here, we combine measurements and theory to provide empirical relationships that could enable predictive estimation of fracture permeability after slip. The proposed relations link fracture length, shear displacement, fracture apertures, hydraulic apertures, and fracture permeability. We consider both fracture creation and the re-activation of existing fractures that have been infilled by minerals. Our hope is that these relationships are useful to the community to ensure safer and more successful subsurface fluid injection and production operations.
Shear fractures in the subsurface can dominate fluid flow by more than six orders of magnitude higher permeability than the adjacent rock matrix (Frash et al., 2017). These natural fractures are a challenge to characterize in-situ (National Research Council, 1996) and are subject to time-dependent processes (Meneffee et al., 2020) and stress-dependent processes (Bandis et al., 1983). Fluid injection can increase pore pressure to stimulate fractures by shear slip. This helps to increase the accessed reservoir volume for oil and gas production (McClure and Horne, 2013) or to interconnect wells for enhanced geothermal energy production (Tester et al., 2006). However, shear slip also brings the risk of induced seismicity (Ellsworth, 2013). A relationship is needed to predict the possible range of fracture permeability and the nominal values of permeability for subsurface flow and transport modelling.
Numerous lab studies have investigated the mechanical and hydraulic properties of shear fractures (Witherspoon et al., 1980; Bauer et al., 2016; Ye and Ghassemi, 2018; Frash et al., 2019a; Proctor et al., 2019; Li et al., 2020). Many of these studies preferred saw-cut rock specimens to improve experiment control and data interpretation. However, the small variances that are typically observed for smooth fractures do not adequately reflect the six or more orders of magnitude variability of permeability that is observed for natural fractures and fractures in the field. It has become crucial to capture the fuller uncertainty of natural fracture properties in predictive flow models in order to provide confidence in the safety of future field developments and the success of stimulation designs.