Towards validation of porous media models using NMR imaging and image-analysis techniques

Abstract

The relevance of a numerical model to a “real system” can only be determined by the physical measurement of the system. This investigation demonstrates that, in conjunction with image analysis methods, NMR microscope data, with determined confidence limits, can be acquired from fluid flowing in porous media. Using a “time of flight” stimulated echo sequence, flow-sensitive proton spin density data were collected from a flow cell filled with glass beads. The use of image-analysis software allows extraction of information concerning the porosity and flow behaviour of the media. The results presented here are of a preliminary nature. With the use of more discriminating experiments, estimates of local velocities, irreducible saturation, and permeability will be possible in porous media, even where more than one fluid is involved. As this technique matures it will provide a direct method of validating computational models of systems such as catalytic reactors, or oil/gas reservoir rock, which are not yet well understood.