Upscaling the Geomechanical Properties of Shale Formations from Drill Cuttings

The geomechanical properties of reservoirs—important for formation stimulation—are often determined from triaxial tests on core plugs. Recovery of these samples is difficult because shales are mechanically unstable and usually disintegrate. To predict the elastic properties of shales at the core scale, we propose two models based on different measurements of drill cuttings. First, a conceptual model is proposed to predict the elastic properties of shales at the core scale from nanoindentations. Nanoindentation measures the local (small-scale) mechanical properties of a shale, whose correlation to the core- and block-scale properties is unclear because of shale’s heterogeneous structure. The proposed model accounts for the effective stiffness of small-scale constitutive minerals at a large scale. It is applied to samples recovered from the Woodford shale and the results are promising. Second, a hierarchical model is developed to predict the elastic properties of shales at the core scale by accounting for the mineralogy, porosity, pore structure, and grain-size distribution. The hierarchical model entails two-scale simulations. At the small (grain) scale, a physically representative element is developed to capture the elastic deformation of a solid grain with a known porosity and pore structure. At the large (core) scale, the model is built based on the volume fractions of the minerals and representative elements characterized at the small scale. The minerals are randomly distributed in the core-scale model. The finite element method is used to compute the elastic moduli. The proposed models are applied to the New Albany, Rocky Mountain Siliceous, Lower Bakken, and Barnett shales. The predicted elastic moduli capture the measured values parallel to the bedding plane. The hierarchical model allows us to determine the anisotropic elastic properties of shale formations at the core scale with additional consideration of the microfabric structures of the matrix. The model is implemented for Lower Bakken, Barnett, and Haynesville shales. Independent lab measurements corroborate the predicted anisotropic elastic moduli at the core scale. The two models have major applications in characterizing shale formations at the core scale from drill cuttings and can potentially be used for real-time analyses in a mobile field laboratory.