A comparative study of shale oil transport behavior in graphene and kerogen nanopores with various roughness via molecular dynamics simulations

Understanding the transport behavior of molecules within nanopores is crucial in various fields, including nanofluidic, bioengineering and geophysics. To elucidate oil flow within shale organic nanopores, the simplified carbon nanopores (e.g., graphene and CNTs) and realistic kerogen nanopores have been extensively applied. The variation in pore structure models leads to divergent findings in oil flow behavior. This study employs molecular dynamics simulations to compare alkane flows within graphene and kerogen nano-slits with similar surface roughness. We observe that using graphene nanopores with ultra-smooth surfaces may introduce errors of several orders of magnitudes in flux calculations. Conversely, when accounting for similar roughness, the static characteristics and flow behaviors in graphene and kerogen slit pores are largely comparable. The relative error of flux and the flow velocity could be lower than 6% when using graphene to replace the realistic kerogen. Alkane molecules exhibit a parabolic velocity profile, with a no-slip condition adjacent to the rough surfaces. In addition, when applying the Hagen-Poiseullie equation for alkane flow in rough nanopores, Gibbs dividing surface is recommended for defining the pore radius, reducing the error of flux predicted to below 10%. Our work fulfills the gap in accessing the suitability of using graphene as a model for oil transport in shale organic nanopores, and provides valuable insights and guidance for future studies on nanofluidic in shale, including molecular simulations, nanofluidic experiment designs, and multi-scale modeling.