Investigating gas permeability in shales using multi-physics and multi-scale network modeling
A. Mehmani1, M. Prodanović1, and F. Javadpour1
1 Department of Petroleum and Geosystems Engineering, The University of Texas at Austin, Texas, USA
1Bureau of Economic Geology, The University of Texas at Austin, Texas, USA
Shale reservoirs, and specifically their permeability, are still poorly understood. Recent high resolution imaging studies have shown that pores in shale are mostly in disconnected regions filled with organic material with sizes in the range of 3-100nm. The pore structure differs fundamentally from clastic structures in that the impervious regions are made up of fibers instead of grains, and the fraction of sorbed gas on the surfaces is comparable to the fraction in the open pore spaces.
We present a novel adaptation of pore network modeling tools for the shale challenge. The network has two interconnected length scales, one representing nanopores, and the other micron-size pathways that connect the nanopore regions. Nanoscale gas flux depends on absolute pressure (as opposed to pressure difference only), incorporates Knudsen diffusion and the equations account for gas slip on the nanopore walls. Microscale gas flux is linearly dependent on pressure difference. The combined flow model is nonlinear, and thus increases the computational complexity.
We find that at lower pressures, slippage effects and Knudsen diffusion are more pronounced and enhance the apparent gas permeability. When desorption effects are incorporated permeability is enhanced due to widening pathways. Different fractions of nanopores vs. micropores as well as the interconnectivity of the two networks, however, can in some cases counter-balance this effect.