A comparison of shale permeability coefficients derived using multiple non-steady-state measurement techniques: Examples from the Duvernay Formation, Alberta (Canada)

Abstract

Matrix permeability, while an important control on fluid flow in unconventional reservoirs, is difficult to measure in the laboratory. There are now multiple methods for laboratory determination of permeability for shales, but little consensus on the appropriate method for permeability measurement. Each technique is based on different physical principles and utilizes reservoir samples at different scales. The combination of sample size and preparation and measurement conditions can lead to a wide range in permeability estimates, creating confusion for recipients of the data. In this work, we compare different non-steady state methods for determination of gas permeability in low-permeability Canadian shales and provide insight into the causes of permeability variation. Further, we analyze and discuss the effects of different controlling factors including porosity, pore-fluid content, mineralogy and effective stress on permeability.Gas permeability measurements were conducted on low-permeability (shale) samples from the Duvernay Formation (Alberta, Canada) using three different methods: profile (probe), pulse-decay and crushed-rock permeability techniques. The analyzed samples differ in total organic carbon (TOC) content, pore network characteristics (porosity, pore size distribution), pore-fluid content (“as-received” and cleaned/dried) and mineralogy. Profile (probe) and crushed-rock permeability measurements were performed on samples in the “as-received” and cleaned/dried conditions. Pulse-decay measurements were conducted on samples in the cleaned/dried state. Helium pycnometry/expansion measurements were performed using “as-received” and cleaned/dried samples under unconfined and controlled “in situ” effective stress conditions.Permeability values derived for the Duvernay samples are strongly dependent on measurement technique, sample size and sample conditions. In the “as-received” state, profile (probe) permeability values, both uncorrected (3.7×10−4–2.7×10−2mD) and corrected (1.5×10−5–5.7×10−4mD) for “in-situ” stress, are consistently higher than crushed-rock (3.7×10−7–5.9×10−6mD) permeability values. Similarly, in the cleaned/dried state, profile (probe) permeability values, both uncorrected (1.9×10−2–1.2mD) and corrected (5.8×10−5–1.4×10−2mD) for “in-situ” stress, are consistently higher than pulse-decay (8.4×10−5–7.6×10−4mD) and crushed-rock (3.8×10−5–1.1×10−3mD) permeability values. In the cleaned/dried state, pulse-decay permeability values (8.4×10−5–7.6×10−4mD) are comparable with crushed-rock (3.8×10−5–1.1×10−3mD) permeability values. Profile (probe) and crushed-rock permeability values measured on cleaned/dried samples are approximately two orders of magnitude higher than those measured on “as-received” samples.The observed discrepancies between permeability values derived from the various non-steady-state techniques is rationalized in terms of sample size, treatment, stress-state and physical principals of measurement. The combined use of these methods is however recommended to provide insight into the controls of sample heterogeneity at sub-cm scales, which is particularly important for shale samples.