HR: 08:55h
AN: H41H-04    [Abstracts]
TI: Pore-Scale Hydrodynamics in Sandstone as a Function of Porosity
AU: * White, J A
EM: joshua.white@stanford.edu
AF: Sandia National Laboratories, MS 0750, Albuquerque, NM 87185-0750 United States
AU: Fredrich, J T
EM: fredrich@sandia.gov
AF: Sandia National Laboratories, MS 0750, Albuquerque, NM 87185-0750 United States
AB: Coupled 3D imaging - direct numerical simulation of fluid flow in porous media has typically used the computationally expensive but highly accurate lattice Boltzmann (LB) method to calculate macroscopic permeability. However, LB simulations provide a rich description beyond this zero-order measure of the system, offering insights into the underlying pore-scale hydrodynamics that cannot be measured in a laboratory core-scale experiment. In this work, we use high-resolution 3D image data coupled with massively parallel LB simulations to investigate pore-scale hydrodynamics in sandstones. The image data, corresponding to 1-3 mm³ volumes, were acquired at resolutions of 3.34 and 1.67 micron during synchrotron computed microtomography experiments performed at the APS. As previously reported, permeabilities calculated from the simulations for sandstones with porosities varying from ~5 to 35% agree with laboratory core-scale experiments over several orders of magnitude range in permeability. In this work, visualization of the microscopic velocity field is used to examine flow geometry, and suggests the existence of preferential flow paths at low porosity. Next, we examine the microscopic distribution of kinetic energy resulting from the macroscopic (1D) fluid forcing. The velocity components that are not aligned with the macroscopic forcing are analyzed to provide a quantitative measure of flow tortuosity. Inefficient out-of-plane and recirculating flow consumes proportionally more of the energy as porosity is reduced. At low porosity, 48% of the flow energy is expended in this manner, dropping to 31% at high porosity. Further, regions of backwards flow are observed, corresponding to 70% of the pore space at low porosity vs 3% at high porosity. The distribution of kinetic energy as a function of nondimensional velocity reveals a unimodal distribution at high porosity, but bimodal distribution at low porosity, with the secondary peak corresponding to backward flow. To further examine pore space efficiency, we define a dead zone metric as the volume of pore space carrying less than 1% of the total kinetic energy. At low porosity, 74% of the pore space is dead zone, as compared to 55% and 35% at intermediate and high porosity, respectively. At low porosity, 10% of the pore space carries 95% of the kinetic energy, whereas 46% of the pore space carries 95% of the kinetic energy at high porosity. We compare these results to Poisseuille flow, and simulations performed on simulated periodic simple, body centered, and face centered cubic porous media.
DE: 1859 Rocks: physical properties
DE: 4430 Complex systems
DE: 5112 Microstructure
DE: 5114 Permeability and porosity
DE: 5139 Transport properties
SC: Hydrology [H]
MN: Fall Meeting 2005