Characterization of the Pore Structure of Porous Media Using non-Newtonian Fluids

摘要

We demonstrate a simple, cheap method for pore size characterization of porous media that generates a distribution of pore radii for improved flow and transport modeling. The new method for pore structure characterization utilizes recent theoretical developments in non-Newtonian fluids. Numerical evaluations and validations with synthetic porous media showed potential for obtaining a distribution of effective pore radii and their contribution to total flow only by complementing water with non-Newtonian fluids in saturated infiltration experiments. To demonstrate this ability on real sands, a series of one-dimensional column experiments was conducted with varying porous medium packings, including Accusands and a polydisperse sand/glass bead mixture. For each packing, distilled water and varying concentrations of guar and xanthan gum were injected over a range of flow rates and pressure gradients. The model-generated pore radii were compared with pore radius distributions measured by X-ray microcomputed tomography (mu CT), with results demonstrating good agreement between the model and mu CT data. Simulations of saturated water flow and drainage curves using model-generated pore radii compared favorably to experimental data, with errors typically between 2% and 10% for single-phase flow and approaching the error of the mu CT measured radius distributions for the drainage curves.Plain Language Summary Knowledge of pore sizes of porous materials is critical to modeling water flow in the environment. Most pore size measurement methods are expensive and require collecting samples of a limited size for laboratory analysis, thus possibly disturbing the pore structure. Our method shows promise as a simple, cheap approach to measure pore sizes directly in the field. The method involves flowing food-grade fluids that exhibit specific flow properties (non-Newtonian fluids) through soils and using the results as input for the model that provides a pore size distribution. We tested the method on four sands, conducting flow experiments with water and six different non-Newtonian fluids. Model results showed good agreement with direct X-ray measurements of the sands. We also used the pore sizes produced by the model to calculate the flow of water through the sands and compared these results with experimental data. We obtained excellent agreement, with errors on the order of 2-10% for water flow and approaching the error obtained using the X-ray results for the drainage of the material. These results indicate that this simple method provides results nearly as accurate as much more expensive and invasive methods and shows promise for use in the field.