Tight Gas Sands-Permeability, Pore Structure, and Clay

Abstract

Summary Gas permeability has been measured on a suite of cores from the Spirit River tight gas sand of western Alberta and on two samples from the Cotton Valley formation of east Texas. Using nitrogen as the mobile fluid, we have measured permeability as a function of partial water saturation at in-situ levels of pore pressure and confining pressure. Samples from both locations show strong dependence of permeability (k) on effective pressure and on the degree of water or brine saturation. The validity of Darcy's law in the micro darcy range has been verified in a dry Spirit River sample. Extensive thin-section, x-ray diffraction, and scanning electron microscope (SEM) studies have been conducted. The primary clays in Spirit River and Cotton Valley cores are chlorite and illite. In one sample we measured k vs. saturation first with distilled water and then with a 2% KCl brine solution and saw no significant change in permeability behavior. By observing the effects of pressure, partial saturation, and salinity on permeability in these samples, we can deduce several important characteristics of the pore structure and can evaluate the relative importance of clay content. Introduction Conventional well logging and well stimulation methods have had to be modified substantially to be successful in tight sandstones. It is now clear that there are some fundamental differences in rock/water/gas interactions in these formations as compared with "normal" gas reservoirs. These differences result primarily from significant pore structure alterations as the rock undergoes compaction and diagenesis. In this paper the effect of tight sandstone pore structure on gas permeability is presented in terms of the response to effective pressure and partial water saturation. Hopefully, some of these results will facilitate the development of well evaluation and stimulation techniques. We measure nitrogen permeability with a pulse-decay technique that allows independent control of pore pressure, confining pressure, and partial saturation. The core sample is about 2 in. (5 cm) in diameter and is 2.4 to 2.8 in. (6 to 7 cm) long. It is jacketed, placed in a confining pressure vessel, and subjected to hydrostatic loading of up to 1.000 bar (100 MPa). The pore fluid is nitrogen at pressures of up to 700 bar (70 MPa). Permeability is measured by applying a 1 -bar (0.1 -MPa) pore pressure differential across the sample and by observing the rate of pressure decay as gas flows through the core. This pulse decay can be used to calculate permeability. A complete description of the experimental method is given in Ref. 1. We have performed several tests to ensure the accuracy and repeatability of all measurements, and one such test is described in this paper. The primary advantage of the pulse-decay method is the speed and simplicity of data acquisition as compared with steady-state experiments. Sample Characterization Ten of the samples tested were vertical cores from the Spirit River formation in the Deep Basin of western Alberta. The other two samples were horizontal cores from the Cotton Valley formation of east Texas. The mineralogy of these samples as determined from thin section analysis is shown in Table 1. Samples are designated by a number corresponding to depth in feet preceded by "SR" for Spirit River and "CV" for Cotton Valley. SEM and X-ray diffraction studies indicate that chlorite and illite in approximately equal proportions are the major clay constituents in Cotton Valley samples and that illite is the major clay in Spirit River cores. The porosities shown in Table 1 were determined by careful weighing of dry and fully saturated samples. JPT P. 2708^