Permeability Estimation in Tight Gas Sands Using NMR — A New Interpretive Methodology

Prince, Christopher M.¹

¹PTS Laboratories, Inc, Houston, TX.

One of the challenges of working with tight sands and shales is the measurement of permeability. One of the most difficult tasks can be acquiring a sample that has not been damaged by the coring and sampling process. Shales are fragile samples. The simple act of drilling a core plug can induce fracturing. Most of the methods used to measure permeability depend upon fluid flow through the sample and any sample damage, be it fracturing due to unloading or sample preparation, will invalidate the measurement even at high confining pressures. T2 spectral characteristics are routinely used to infer petrophysical characteristics in both clastic and carbonate reservoir rocks. It would be advantageous to be able to apply those techniques to low-permeability samples. . NMR spectral methodologies have distinct advantages in that they can be applied to damaged samples, and they can be applied to well logs eliminating the need for a core sample. The major problem is that the methodologies available were deduced from sandstones and carbonates with permeabilities in the milliDarcy range, not those in the micro- and nano-Darcy range. The reason for this breakdown may be associated with the method of interpretation of the spectrum.

We present an alternative interpretation, relating the permeability to the modal T2 time, not the geometric mean. The modal T2 time represents the most common pore size. The utility of these methods arises from the relatively fixed aspect ratio between pore size and throat size in sandstones. Averaging approximately 2:1, the stability of this relationship means that pore size can proxy for throat size in any estimation equation. It is an interpretation which honors the petrologic characteristics of the pore network and, as a methodology, it relatively fast, inexpensive, and it is one that can be applied to well logs.

As a demonstration of the efficacy of this technique, the results from an analysis of two data sets are presented. The first is a set of seven shale samples and the second is a set of 30 shale and sand samples from a tight gas unit. Subsamples were taken of each plug for MICP analysis from which permeability was derived using Swanson’s air permeability method. Permeabilities were derived from the MICP data and compared with the log of the modal T2 time. The resulting correlation was 0.93. The results were surprising, not only because of their strength, but because the technique worked well in both sands and shales.

AAPG Search and Discover Article #90100©2009 AAPG International Conference and Exhibition 15-18 November 2009, Rio de Janeiro, Brazil