Image-based core-scale real gas apparent permeability from pore-scale experimental data in shale reservoirs

This paper presents a new upscaling method to derive the core-scale apparent gas permeability from an improved pore-scale permeability model and experimental data, with more rigorous incorporation of varying gas storage/transport mechanisms in nano/micro pores. First, in use of SEM images of a gas-rich shale field example in Sichuan Basin from our lab, pore network models of inorganic-matter (IOM) and organic-matter (OM) are characterized by using a digital-core technique. Next, an improved pore-scale real gas apparent permeability is modeled rigorously for both IOM/OM, respectively, with 1) bulk gas transport, gas adsorption, surface diffusion, pore-size confined phase behavior, and stress-dependent rock properties and 2) an additional reduction in inorganic pore sizes by water film adhered on pore surfaces. Core-scale permeability is then derived by assembling the permeabilities of stochastically distributed IOM/OM patches with different pore network models properties using the Monte Carlo sampling method. The new core-scale permeability model is validated by pulse-decay permeability experiment. Moreover, the representative elementary volume (REV) size is determined by analyzing the relative standard deviation of apparent gas permeability in cases with different sample sizes. The contributions of different gas transport mechanisms are discussed, and the impacts of stress-dependence for several field examples (i.e., Sichuan, Pierre and Barnett Basins) and water film with varying relative humidity (RH) on core-scale apparent permeability are analyzed. This work provides an effective approach to determine the core-scale shale permeability by directly using pore-scale experimental data, which is a common challenge in the unconventional resources.