Prediction of permeability change at high ambient stresses via the isotropic Skempton coefficient B

For gas, oil and water exploration reservoir permeability as a function of effective stress is one of the most important hydraulic parameters. Estimation of permeability, especially in deep reservoirs, is very difficult and time-consuming. Therefore, permeability is often estimated in laboratory experiments under simulated in-situ conditions. Under these experimental conditions with a flow across the sample, many effects lead to changes in permeability. Besides the flow paths reduction as a function of effective pressure, plugging of the sample and filters by fines migration or rust and a swelling of the clay content can occur, which results in a decrease in permeability. All these non-mechanical effects are time dependent and affect the permeability measurements, hence a separation of all these influences is hard to achieve. To avoid these problems we estimated the permeability pressure dependence with the isotropic Skempton coefficient. The Skempton coefficient is defined as undrained pore pressure change due to ambient stress changes B=dpu/dσm. We could show that a heterogeneous deformation of pore space geometry led to a decrease of the Skempton coefficient with increasing confining pressure. The mechanisms which influence the Skempton coefficient are similar to the behavior of the sandstone sample during the permeability measurements. In both cases we consider a change in pore pressure and an adjacent equalization across the flow channels at the micro-scale. These flow channels change their geometry depending on the applied stresses. Therefore, the reduction of the Skempton coefficient should be comparable to the reduction of permeability. To validate this assumption we present experiments on Lower Permian sandstone (Rotliegend) samples from the NE German Basin and compared Skempton coefficient and permeability measurements to find a coherence of both rock properties. Applying this relation of Skempton coefficient and permeability, we can predict rock permeability at high ambient stresses based on a single measurement at atmospheric pressure.