Second-Order Correlation of Klinkenberg-Corrected Permeability and its Experimental Verification on Heterogeneously-Stressed Gas Shale

Shale is an extremely tight and fine-grained sedimentary rock with nanometer-scale pore sizes. The nanopore structure within a shale system contributes not only to the low to ultra-low permeability coefficients (10−18 to 10−22 m2), but also to the significant gas slippage effect. The Klinkenberg equation, a first-order correlation, offers a satisfying solution to describe this particular phenomenon for decades. However, in recent years, several scholars and engineers have found that the linear relation from the Klinkenberg equation is invalid for most gas shale reservoirs, and a need for a second-order model is, therefore, proceeding apace. In this regard, the purpose of this study was to develop a second-order approach with experimental verifications. The study involved a derivation of a second-order correlation of the Klinkenberg-corrected permeability, followed by experimental verifications on a cubic shale sample sourced from the Sichuan Basin in southwestern China. We utilized a newly developed multi-functional true triaxial geophysical (TTG) apparatus to carry out permeability measurements with the steady-state method in the presence of heterogeneous stresses. Also discussed were the effects of two gas slippage factors, Klinkenberg-corrected permeability, and heterogeneous stress. Finally, based on the second-order slip theory, we analyzed the deviation of permeability from Darcy flux. The results showed that the apparent permeability increased more rapidly as the pore pressure declined when the pore pressures are relatively low, which is a strong evidence of the gas slippage effect. The second-order model could reasonably match the experimental data, resulting in a lower Klinkenberg-corrected permeability compared with that from the linear Klinkenberg equation. That is, the second-order approach improves the intrinsic permeability estimation of gas shales with the result being closer to the liquid permeability compared with the Klinkenberg approach. Analysis of the experimental data reported that both the first-order slippage factor A and the second-order slippage factor B increased with increasing stress heterogeneity, and that A was likely to be more sensitive to stress heterogeneity compared with B. Interestingly, both A and B first slightly increased and then significantly as the permeability declined. It is recommended that when the shale permeability is below 10−18 m2, the second-order approach should be taken into account. Darcy’s law starts to deviate when Kn > 0.01 and is invalid at high Knudsen numbers. The second-order approach seems to alleviate the problem of overestimation compared with the Klinkenberg approach and is more accurate in permeability evolution.