Sorptive permeability loss determined from strain-based analysis of tightly constrained experiments on shale

Abstract

We build a model to determine sorptive permeability loss from observed total permeability evolution by considering adsorption and poromechanical expansion as parallel processes. The model is cross verified with a separately derived strain-based model for sorptive permeability evolution. Both models are compared to laboratory data and are shown to have excellent agreement. We isolate the sorptive strain from the total strain measured during serial injection of sorptive then non-sorptive gas species. Our model predicts that injection of a non-sorptive gas into a shale saturated with a sorptive gas causes permeability to approximately double in both the bedding-parallel and bedding-perpendicular directions. We preform nitrogen floods at constant stress, pore pressure, and temperature to observe the sorptive permeability recovery absent other confounding effects. We confirm that in shales the component of permeability evolution due to sorptive swelling can be isolated from effective stress effects. Laboratory results show a 206% and a 234% permeability increase in the bedding-perpendicular and bedding-parallel directions, respectively, as a result of nitrogen flooding. This counters and indeed dominates over any sorptive permeability loss that despite a modest fraction of organic material (∼1–2%) is amplified by the low pore density. We find that sorptive permeability evolution in shales is controlled by the sorptive strain, pore density, and pore geometry.