The influence of particle size, microfractures, and pressure decay on measuring the permeability of crushed shale samples

Measurement of matrix permeability is essential for predicting, evaluating, and modeling the performance of shale reservoirs. However, the repeatability and accuracy of these measurements can be questioned because procedures have not been standardized. As a result, permeability measurements from the same sample by different laboratories can vary by orders of magnitude. Microfracturing related to changes in stress during core retrieval and crushing during sample preparation is thought to be a significant source of error. Different interpretations of pressure decay curves could also account for inconsistent permeability values. The goals of this research were to analyze relationships among crushed particle size, microfractures, and matrix permeability, as well as to evaluate the ways in which pressure decay curves are interpreted and to determine the sample mass best suited for analysis. Crushed rock pressure-decay measurements of particles of different sizes were obtained using a shale matrix permeameter, and permeability was estimated by curve fitting using Core Laboratories software and by other methods that assess the geometry and evolution of pressure decay curves. Results indicate that the relationships between permeability and particle size vary considerably when determined by different methods. Analysis of pressure decay curves reveals three distinct segments. The early segment is characterized by hyperbolic decay, whereas the late segment is characterized by exponential decay. A third segment records a pseudo-steady state where pressure has declined to the extent that decay can no longer be characterized. Decay was measured for about 2000 s; most decay curves stabilize within 500 s, and data collected beyond 500 s are dominated by noise associated with the pseudo-steady state and are beyond the resolution of the apparatus used. Analysis of early hyperbolic curves yields permeability values one to two orders of magnitude greater than whole curve analysis. The hyperbolic pressure decay segment appears to be influenced by microfractures and other large pores near the surface of samples, whereas the late time segment and whole curve correlate more strongly with the microporous to nanoporous rock matrix. Also, permeability values derived from whole curve analysis are sensitive to measurement duration, and different values are obtained when permeability is determined from different time windows. SEM images of all particle sizes analyzed reveal microfractures with diameters ranging from 60 to 1020 nm, but no correlation was found between microfracture aperture and particle size. The optimal sample mass used in our shale permeameter is 50-100 g, which facilitates resolution of the major elements of the decay curve. Optimal particle sizes are between 1.0 and 1.4 mm.