Transport of Multicomponent Hydrocarbon Mixtures in Shale Organic Matter by Molecular Simulations
J Collell and G Galliero and R Vermorel and P Ungerer and M Yiannourakou and F Montel and M Pujol, JOURNAL OF PHYSICAL CHEMISTRY C, 119, 22587-22595 (2015).
During the past decade, gas recovered from shale reservoirs has jumped from 2 to 40% of natural gas production in the United States. However, in response to the drop of gas prices, the oil and gas industry has set its sights on the oil-prone shale plays, potentially more lucrative. This shift from dry to condensate-rich gas has raised the need for a better understanding of the transport of hydrocarbon mixtures through organic-rich shale reservoirs. At the micrometer scale, hydrocarbons in shales are mostly located in amorphous microporous nodules of organic matter, the so-called kerogen, dispersed in an heterogeneous mineral matrix. In such multiscale materials, a wide range of physical mechanisms might affect the composition of the recovered hydrocarbon mixtures. More specifically, kerogen nodules are likely to act as selective barriers due to their amorphous microporous structure. In this work, we study the transport of hydrocarbon mixtures through kerogen by means of molecular simulations. We performed molecular dynamics simulations of hydrocarbons permeating through a molecular model representative of oil-prone type II kerogen. Our results show that the permeation mechanisms through this type of material is purely diffusive. Consequently, we have computed the Onsager's species-specific transport coefficients of a typical condensate-rich gas mixture within kerogen. Interestingly, we have observed that the transport coefficients matrix can be reasonably approximated by its diagonal terms, the so-called Onsager's autocorrelation coefficients. Inspired by the classical Rouse model of polymer dynamics and surface diffusion theory, we propose a simple scaling law to predict the transport coefficient of linear alkanes in the mixture. In good agreement with our simulations results, the Onsager's autocorrelation coefficients scale linearly with the adsorption loading and inversely with the alkane chain length. We believe our results and predictions are applicable to other materials, such as carbon-based synthetic microporous membranes, with structural properties close to that of kerogen.
Return to Publications page