Mechanical Anisotropy of Unconventional Shale – Build the Correct Relationship between Static and Dynamic Properties

Abstract

Abstract Most of unconventional shales are mechanically anisotropic and usually treated as transversely isotropic (TIV) media. Full anisotropy characterization of gas shale samples traditionally requires laboratory tests on several plugs cut along different orientations to the bedding. In practice, very often, this is untenable – due to the scarcity of shale plugs, experimental challenges, cost consideration and other factors. As alternate, ultrasonic measurements are only performed along the longitudinal axis of the sample, where the apparent dynamic properties are determined using isotropic media equations, and then used to build a correlation model between static and dynamic properties. However, such approximations could result in significant difference in reservoir rock mechanical properties determined from downhole acoustic logging measurements when the simplified apparent dynamic/static correlation model is used. The paper performs ultrasonic measurements, velocity anisotropy analysis, and triaxial compression testing on a series of shale samples cut in two orientations – parallel and perpendicular to the bedding. The static Young's moduli measured along the bedding are not always higher than those measured perpendicular to the bedding. For the samples tested, horizontal plugs generally have higher Young's moduli than vertical plugs in the range of low modulus, while lower than vertical plugs in the high modulus range. Generally, horizontal plugs have relatively higher static Poisson's ratios than vertical plugs. Both the dynamic and apparent dynamic Young's moduli are higher than the static moduli. There exist strong correlations between the dynamic/apparent dynamic and static moduli. However, the apparent dynamic moduli seem to have a relative better correlation with static values than the dynamic ones. In addition, the apparent dynamic moduli are higher than dynamic ones when they are measured perpendicular to the bedding, while they are almost the same when measured parallel to the bedding. The discrepancy observed between the dynamic (measured from velocity anisotropy analysis) and apparent dynamic (using the isotropic model) confirms the importance of distinguishing and carefully selecting between these values when building static-dynamic relations for log-core calibration. The method of performing velocity anisotropy analysis on one horizontal plug only enables one to obtain the complete dynamic properties for the transversely isotropic media, which could considerably simplify the anisotropic measurements, save core material, and make much more geomechanical data available for shale well development.