Specific surface area: A reliable predictor of creep and stress relaxation in gas shales

In recent years, short-term creep parameters determined in the laboratory from cylindrical gas shale samples subjected to triaxial (in-situ) stress conditions have been used successfully to infer long-term deformation and stress relaxation at the reservoir scale across geologic time scales. Due to the viscoelastic formalism, both the laboratory creep response and field-scale stress relaxation can be modeled with power law functions of time involving the elastic compliance of the shale B, the time-dependence exponent n, and the amount of total strain ∊. Gas shales often exhibit a high specific surface area associated with their high content in clay minerals and/or total organic carbon (TOC). The low-pressure nitrogen adsorption technique can be used advantageously to estimate specific surface area (SN2); i.e., it is a relatively fast and cost-effective measurement conducted on powdered samples of shale material. A robust global empirical correlation between gas shale creep parameters and SN2 emerges from the analysis of laboratory data collected from multiple gas shale formations in Australia (the prospective Goldwyer Formation) and the United States (Barnett, Haynesville, and Eagle Ford formations), and spanning a broad range of clay content, organic matter, maturity, and porosity values. This data set also shows that the summed fractions of clay minerals, TOC, and porosity, the so-called weak phase fraction, correlates nearly as well with primary creep parameters. The weak phase fraction can also be estimated from faster and more cost-effective measurements or from well logs. To evaluate its predictive capacity, the key correlation between SN2 and creep parameters is used in a case study to predict the magnitude of present-day least principal stress Shmin across six depth intervals/lithologic layers in a prolific unconventional shale formation in the northeastern United States. Several Shmin measurements are available for verification, and our approach successfully captures the observed layered variation of stress with depth.