Transport of gas in extremely-low permeability shale-gas reservoirs involves complex processes of absorption, adsorption, and pore-confinement effect in nanopores; significant deviations occur from Darcy-type flow; and gas properties such as real gas deviation factor and viscosity are significantly altered compared to conventional reservoir conditions. This paper presents a physically rigorous modeling of shale gas transport by considering the various effects of importance in nanopores to derive the proper equations of gas storage and transport, and to demonstrate various applications of practical interest. First, previous approaches are critically reviewed to delineate their outstanding features and shortcomings. Then, a non-Darcy gas transfer equation, comprehensive gas storage model, and quantification of the relevant parameters including permeability are developed. Next, the improved model is used to simulate gas transport in laboratory tests conducted under near-real shale-gas reservoir conditions. Improved non-Darcy nanopore gas storage and flow model describes the shale gas transport properly and can be used satisfactorily in shale-gas reservoir simulation.
Although the theory of gas transport through extremely narrow flow paths in porous media have been reasonably well established, the analyses of experimental data have not been quite successful judging by the results reported in the literature. For example, Javadpour (2009) had to adjust the values of three empirical parameters to be able to achieve a matching of experimental data. Darabi et al. (2012) applied the apparent permeability function (APF) concept which was originally formulated by Ertekin et al. (1986). Darabi et al. (2012) also have three adjustable parameters. Unique determination of these three adjustable parameter values is questionable. Because of the error as pointed out in this paper, the model given by Javadpour (2009) did not match the measured data. The simulation results presented by Roy et al. (2003) and Veltzke and Thöming (2012) also deviate significantly from their own experimental data. These papers attempted to determine the values of their adjustable parameters using only one set of experimental data. Civan et al. (2012) explained that "one must run a minimum number of tests that is more than the number of adjustable parameters with the same system but conducted under different conditions to achieve uniqueness."
