(46d) Energy Optimization of Nanochannel Fluid Extraction with Various Channel Geometries

Many factors influence the extraction of hydrocarbons from shale and other subsurface sources. During secondary oil recovery, an injection well floods the underground reservoir with inexpensive fluids like seawater. Occasionally, a chemically engineered working fluid is created using surfactants, and injected instead. Attempts have been made to computationally determine the role salt content and surfactants play in this process. None of these studies directly targeted the relative wettability between water and hydrocarbons, while simultaneously accounting for their nanoconfinement within solid channels of varying surface morphologies. DPD is a coarse-grained, particle-based computational model capable of performing mesoscale simulations. Thus, DPD can give us insights into hydrocarbon recovery at tens of nanometer length scales and for tens of nanosecond time scales. This cannot be accomplished with molecular dynamics models using current computational resource capabilities. We perform many-body dissipative particle dynamics (mDPD) simulations to study fluid flow in channels with sizes ranging from 4.5 nm to 13.5 nm. Model parameters are rigorously calibrated to handle the realistic working fluid, water, and source fluid, heptane (C₇H₁₆), confined in amorphous silica (aSiO₂) channel walls. Different pressure levels are applied to the water as pressure driven boundary conditions, and heptane is extracted from the channel. The mDPD attractive parameter is varied in order to change water from a wetting fluid to a relatively non-wetting fluid. Based on the study of channel geometries (i.e., flat, convergent-divergent, and pocket designs), we find a relationship between increased surface roughness and permeability reduction. We will present an average power exerted on the water during initial channel flushing, as well as the maximum amount of heptane recovered, as a function of the water wettability and surface roughness. An optimization algorithm will be proposed to determine the most efficient heptane extraction forcing term in each case.