CHARACTERIZATION OF PORE SIZE DISTRIBUTIONS USING NON-NEWTONIAN FLUIDS

The pore structure plays a primary role in single and multiphase flow in porous medium systems. Accurate determination of pore size distributions traditionally requires expensive, ex situ methods, such as x-ray microcomputed tomography (μCT) or MRI. Recently, an alternate approach was developed that uses data from simple, Darcy-like flow experiments conducted with non-Newtonian fluids as input for a numerical model that provides a distribution of effective pore radii and their contribution to total flow. Earlier work used virtual media to validate this method; here we apply the method to real sands to demonstrate its accuracy. We conducted a series of one-dimensional column experiments with three sands with varying grain size distributions and a polydisperse sand/glass bead mixture. For each media, flow experiments were conducted at controlled flow rates with water and six non-Newtonian fluids, and the resulting pressure gradient was measured with a pressure transducer. The data from these experiments were used to generate pore size distributions from the numerical model. The resulting pore size distributions were validated in several ways, including direct comparison with μCT imaging and comparison between simulated saturated water flow and drainage curves using model-generated pore radii and experimental data. All validation approaches indicated good agreement between modeled and experimental results, with errors of 2-10% for single-phase flow and accuracy comparable to μCT-measured radius distributions for the drainage curves. The method shows promise for providing a simple, cheap approach for determination of pore size distributions, with future development aimed at allowing in situ, field applications.