Dependence of Methane Transport on Pore Informatics in the Amorphous Nanoporous Kerogen Matrix.

Abstract

Fluid transport in kerogen is mainly diffusion-driven, while its dependence on pore informatics is still poorly understood. It is challenging for experiments to identify the effect of pore informatics (such as pore connectivity and tortuosity) on fluid transport therein. Therefore, in this work, we use molecular dynamics simulations to study methane transport behaviors in amorphous kerogen matrices with broad pore properties. The pore properties including porosity, pore connectivity, pore size, and diffusive tortuosity are characterized. Next, self-diffusion coefficients in the connected pores (DeffS) and in the total pores without distinguishing its connectivity (DtotS) are calculated in all the kerogen matrices based on the free volume theory. We find that both DeffS and DtotS exponentially decreases with methane loading with two controlled parameters: fitting constant αeff and DeffS(0) (DeffS at infinitely small loading) for DeffS and fitting constant αtot and DtotS(0) (DtotS at infinitely small loading) for DtotS. However, in the kerogen models with relatively low pore connectivity, αeff and αtot as well as DeffS(0) and DtotS(0) can be quite different, inducing the different estimations of DeffS and DtotS. Since methane in the unconnected pores does not contribute to the actual transport, it is important to recognize connected pores when evaluating the fluid transport in kerogen. On the other hand, DeffS(0) strongly depends on the effective limiting pore size (rlim_eff) of the dominant flow path and effective diffusive tortuosity (τeff), in which DeffS(0) linearly increases with (rlim_eff/τeff)2. We also find that αeff is a multivariable function of ϕeff, τeff, and rlim_eff, but their generalized relation requires more data to obtain. This work provides important insights into fluid transport in kerogen based on the kerogen pore informatics, which are important to shale gas development.